<%BANNER%>

Design, Fabrication, and Characterization of Laminated Hydroxyapatite-Polysulfone Composites

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101204_AAAAEX INGEST_TIME 2010-12-05T03:35:31Z PACKAGE UFE0011396_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 5725 DFID F20101204_AADDCJ ORIGIN DEPOSITOR PATH wilson_c_Page_010thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
ee8a84e72946437cb89f5712cb9c1f64
SHA-1
4213464be69e4ffc62946f74e0ce64c23c8a76da
314 F20101204_AADDBU wilson_c_Page_186.txt
2f136fa3b6948db797e4f915b4b37753
5a8ed5ce69679a09eea9e95cfd1596387063fc59
21540 F20101204_AADCXC wilson_c_Page_115.pro
8c8e614d3904590de87bed3017d8c1e8
f8d5103647008c197146dbf48e4194a3c672df6d
31099 F20101204_AADCWO wilson_c_Page_081.pro
d971b9d2da1304139f46a96ba655d3ba
0a3ff3f2c11363566ffbe196524a0fd97c23bd4c
2012 F20101204_AADBSY wilson_c_Page_047.txt
370d2f0d669fcc714d74635b10ea9b5b
e8f89c5fdf40d996108c0744154566adea59c16c
32389 F20101204_AADBUA wilson_c_Page_145.pro
8e3eb85162d1b1a117fe4ef40459b957
2935ff460d30199a30bb9fa095f479dc25a1246b
5764 F20101204_AADBTM wilson_c_Page_122thm.jpg
4e3083197617f7734522cec1ea17502a
4a1acc288faa6c8b549f44f1c05030f2c320e72f
27379 F20101204_AADDCK wilson_c_Page_012.QC.jpg
254125279bcf125135f84361257bfb0c
f619c4b16a3a62f537291ef4a02b43f1cd4b6056
2298 F20101204_AADDBV wilson_c_Page_189.txt
a62541ff03864a2d6deb5099760ac32e
5b09651af45b7658b7bdb945a433d50dc096d077
40189 F20101204_AADCXD wilson_c_Page_118.pro
c6919eb4e2fd051aadf21336cb182db4
5da234539c654da3cf601a459ac142369b76d98e
45211 F20101204_AADCWP wilson_c_Page_082.pro
d81062a114b5f0b7d1d90b534d16e6d0
5856eb50e6deb5959e3d253c4ddccc3192bbcbb4
820122 F20101204_AADBSZ wilson_c_Page_145.jp2
5e53098f887b42704d6c4dc81d4f0512
25123c0a4bce5878c23e02fee01c3dbf43f22899
25271604 F20101204_AADBUB wilson_c_Page_115.tif
4eb73da98276fa923cae9f898a15355b
822561be54d450a95c7b3bb34fd42f9cafa906e8
80233 F20101204_AADBTN wilson_c_Page_005.jpg
1b65d536c5f06f5870f26fab55127bb2
b38b25e5714992bb7cd030cb918990c52f1f0066
785 F20101204_AADDBW wilson_c_Page_191.txt
7406764f311895e55dda851c65558a83
a0a0cad0330e0f658b66cfc9f29f789f5bbbc0bc
40391 F20101204_AADCXE wilson_c_Page_119.pro
9dc330aecb8ae0c2c51fedd4c0967bba
c48652f3bf724f382aee36f4b99d6837810a2f8c
46191 F20101204_AADCWQ wilson_c_Page_083.pro
596a5ee511bfa533970c4c63480efa06
485a91e096b0e77a71067f8d73b4dc12f9ffda02
4504 F20101204_AADBUC wilson_c_Page_142thm.jpg
d4ec728592d966ca8cca4d18c824e1e1
52f09242aecb4f27e1dfdb602aafd8ebd86d7d5f
1053954 F20101204_AADBTO wilson_c_Page_078.tif
e4c4f5036e4cb1f0f9919677bba7ab0b
91ad177fd1117ec754f816584293017944d45493
5595 F20101204_AADDDA wilson_c_Page_028thm.jpg
54b56eedc60ae1bbf01472d7faaba7fc
8035b035d34f1208e88f566a4636360427677fbf
6739 F20101204_AADDCL wilson_c_Page_012thm.jpg
a4bb09f21d96bf3074caaaf2a900cbe0
516cdf97369f1d43df526882c14fc6030e2e2a03
1926 F20101204_AADDBX wilson_c_Page_001thm.jpg
4a835a7ac2e93440ec51242c283a9acb
75a89dc01d67ff7191b1fd4fca649259c173c1de
45101 F20101204_AADCXF wilson_c_Page_122.pro
048c1ded6f44db4a83fb96891f6bacaa
ae8362599c4476f4f6fdd768d26c350af7467b7f
45042 F20101204_AADCWR wilson_c_Page_085.pro
2ced0d1e022e7c179e45bb64247486ba
2322ff7cfa0ec68e10268d679fe9b3a4f803e7eb
77493 F20101204_AADBUD wilson_c_Page_137.jpg
2141f9428491d40850734171c3d0f813
3ed6828e1b8844ffba6e6d40a4b1bf3e30f0d935
1831 F20101204_AADBTP wilson_c_Page_027.txt
b45ded5fab7d298ee10e62016914378c
31c5ee46839d5e3216d170b86626dbdeeccbd7c5
25759 F20101204_AADDDB wilson_c_Page_029.QC.jpg
65733ace8a42e85bbd223c7a79c536a4
4f962f8d1a4ac1a22bd14cdbaf800a284ce4a234
6859 F20101204_AADDCM wilson_c_Page_013thm.jpg
501952d0300400970f7445e34cc8e589
d48208356fd2f4040711cea925141243d33f9395
5698591 F20101204_AADDBY wilson_c.pdf
d95ec15f08c44fadb1d1c1dbd5919ce0
fb087edd902b6c23dd42ded26a4ef04937ca35dc
13880 F20101204_AADCXG wilson_c_Page_127.pro
5d97517e7a1ea2073ea28db7b9bbb01b
89cbefb460cf9e16a106a1fe92dd931c47b1a086
42350 F20101204_AADCWS wilson_c_Page_086.pro
4f921c7acd63551a1bcfb4c1a62d73fc
4f8d5215accd062b98a689afc3a52eef27c18377
106452 F20101204_AADBUE wilson_c_Page_036.jp2
bbc5442c72ebcf5d02a7f2fe78da0333
215611d941e9abda69705c0b92b3010871547da5
1955 F20101204_AADBTQ wilson_c_Page_172.txt
cf29975728d12f4a48b791433a469b4e
6e5c4cf3baf42be9cf93034eaec488e8d4673c1f
6124 F20101204_AADDDC wilson_c_Page_029thm.jpg
c2a524567c8b2554d9870e2555c10c84
0474e96b247cbf44735e62f5f85400cab79af801
28397 F20101204_AADDCN wilson_c_Page_014.QC.jpg
28d27ff2c743d09d77ee2dfe26ac29b9
4aa013c03076d0185acf2f08734d9fa848cbdc41
220726 F20101204_AADDBZ UFE0011396_00001.mets FULL
ca1dc00f211c8fd14136d5ca017af3db
0f1d0d9aa88a57a139808d004dcc2a92accf6990
12992 F20101204_AADCXH wilson_c_Page_128.pro
4b48bd052b5b815a1ff8a7489a1ea112
69c75d0690fcc70fe0c0249cbb08c2154ae84792
42661 F20101204_AADCWT wilson_c_Page_090.pro
7f98cd315a22179a544cb7239755fb4f
e51ab472fb49d3706c92a3bb609a4895b61c4f1e
49814 F20101204_AADCAA wilson_c_Page_048.pro
2b0a5da9744271cf0a49626c984a16d4
22544ed3a30e42b88d289c35f1c220dd4b02fd3b
F20101204_AADBUF wilson_c_Page_182.tif
5b3af238e1e1bb4f4460a4fd26d2deff
106fe168a10cc01dc36e80e626ea2d51e7ee0920
545897 F20101204_AADBTR wilson_c_Page_088.jp2
a708db8df90b269bb4d66cf462d4d7b6
8ad159e32ca90abaa952ad179ed84db6f0ae12eb
22917 F20101204_AADDDD wilson_c_Page_031.QC.jpg
c7048b08e50db519cad37a68ab4711cd
1b43902845da888d9a2cd17c64d4cedf699ad19a
7027 F20101204_AADDCO wilson_c_Page_015thm.jpg
5019e5f4b7ed97264c2e9c572e6ce4e8
c9f4c2a8da7124b31d580c20d157f9f6008a1e54
33309 F20101204_AADCXI wilson_c_Page_130.pro
b34b82db8b54d89ac5493a0c2d563e24
d69cf9d4e0cf174759e4ab40a781ee48fccac333
45588 F20101204_AADCWU wilson_c_Page_091.pro
4de70f70de403a0a4609f7ef2651ebe7
6335380f620fd12b39afc16d8938b8e06ddc0eca
6396 F20101204_AADBUG wilson_c_Page_174thm.jpg
a36153a5bc15dfbbfe175431f7110a40
d8e5446ef11c86866aec59d192022817f4a06708
80931 F20101204_AADBTS wilson_c_Page_036.jpg
422a0c24a13d413307c4908ead7e5ced
b80d9b821dd2457cd019b6016c65c0d759b6e2c9
F20101204_AADCAB wilson_c_Page_103.tif
ac74649c1fbc4af70a743177a4966105
d78d09e783110022b8639ecbffdc6fb399c86bab
5275 F20101204_AADDDE wilson_c_Page_031thm.jpg
b961e010f7d4efb264202833d90e2401
df1da3ad500b68a16f7c2fb7077f7b7b7465dbd5
6931 F20101204_AADDCP wilson_c_Page_016thm.jpg
0c7270b7e2a6b087282777b921dec3ab
c2ae948c8fb82bf78bc27b95f3308c5b76f3d59d
21279 F20101204_AADCXJ wilson_c_Page_135.pro
afbf018d94d2cc33615cdea225bb252b
8c5542cf6770536e275921872559c1926bdb9d40
22478 F20101204_AADCWV wilson_c_Page_095.pro
df2770fa1d6ac3dae745c21674a4d19e
2d34f81cc628d58bdf80c2cb9b287f1a5d3b32c8
13517 F20101204_AADBUH wilson_c_Page_143.pro
40c7eef6d5690dec87e32c44d86c4d9b
5dfb5be9ac560e4231f5ea7d892f487531227aaf
F20101204_AADBTT wilson_c_Page_124.tif
71f2360658b2cf5d2644225bf9b3f954
9d5c122858d48259d3d7b3e9d7e48ccd6b41ab08
823 F20101204_AADCAC wilson_c_Page_143.txt
2d27f97015ba8df06141b6f6571bbace
59b66d7ed471b96cc9ffeefc04b48578bd85d308
23286 F20101204_AADDDF wilson_c_Page_032.QC.jpg
2eaae946d2a4187a4c7895fcfe5fc795
a68f95553353cf1144794fe17bcc8ece94285e7a
6162 F20101204_AADDCQ wilson_c_Page_017.QC.jpg
da68eaeaa18fd368cb9ef72fd6ba8861
5cd14cebd2bed1ca079966e9fefd9e78e0eff94d
45823 F20101204_AADCXK wilson_c_Page_137.pro
14c0e192eaeefd013ba9a37ab91d3778
34f5e617a172cbacd371d2c00564822282d0b346
30964 F20101204_AADCWW wilson_c_Page_096.pro
0ac9447c53a91f4ad1c7547b58fe1f0c
d98c2fec1922dc7fcf36b9add37c4504b9bdfb76
1320 F20101204_AADBUI wilson_c_Page_113.txt
2e847094253080d16ffb816942648be2
50b1c707d56cdc00532502e5fd2dccbc374efb5a
71991 F20101204_AADBTU wilson_c_Page_117.jpg
2e0e4ec2f912269b05ffbe8124e8882e
519d7e4ec9c85e7485e187bf4f91e03ed8787265
5238 F20101204_AADCAD wilson_c_Page_020thm.jpg
e4a9a2af3c37dfe753adc5a0c33692d6
482b7a25648c91833b071fcf2948d5ac1479f99c
5100 F20101204_AADDDG wilson_c_Page_034thm.jpg
019d92d921791efce317897fb40fed4e
c1f3580017332da3d5f59294034224ef81afdb1a
1725 F20101204_AADDCR wilson_c_Page_017thm.jpg
9d6370d5d218beb3fed3da32e6698a5c
409ee3cfd1603052b5f9b8002be01b2e0815d9f0
7763 F20101204_AADCYA wilson_c_Page_178.pro
c2aa601c5041911ef0194d669a01f616
3fd757297320fde6e59be33d9dee581e6800ba8b
30884 F20101204_AADCXL wilson_c_Page_138.pro
2fba8a95008610d5ae90fce0b3972126
b8dc3f0e42e37e1daf28b0df66afc89c74e206cd
37078 F20101204_AADCWX wilson_c_Page_097.pro
6a738961f9bb4fe8af94024c25cb5368
12e617961d830d88a11b69c249ea488eb22e2560
F20101204_AADBUJ wilson_c_Page_020.tif
27190fe74b8404c20debff288ef45363
7d19649a0b532fb62e0a17760ea47800b2d25e7a
F20101204_AADBTV wilson_c_Page_079.tif
78e65bb908d84855e3c6bf248f89cdf7
4171e799d36ed0dcb20a644f6e8c3179db672eab
25330 F20101204_AADCAE wilson_c_Page_065.pro
f1d154f8e111a201b3cb4a1b31d8591c
1fe8e1b102a53346c853f5c8d8b2710a1b02736f
6016 F20101204_AADDDH wilson_c_Page_036thm.jpg
e9aeba409931a42c437b92cd532cb357
4d94eceeeffe39f64dd78a77f8c01b3bbcc535e7
19290 F20101204_AADDCS wilson_c_Page_018.QC.jpg
380f64bca91f1d51622023204b771413
d319eacde36bdcd67d38543e89f8ce07e6aab6c8
51137 F20101204_AADCYB wilson_c_Page_179.pro
64803914ce6815d812fc6b103c5c1f8d
2a90ed79d1f4f1b8088c55a30e4a21a657a3d09b
9397 F20101204_AADCXM wilson_c_Page_140.pro
cf870df1d31c8cc3fc8cff696ede2ce6
403f4e167a70e397b5028b9906b09d71435e6791
9195 F20101204_AADCWY wilson_c_Page_101.pro
fbb714dbb3b67ab22279d56060d0aa66
29f5915c1574b29ae6d4e4bb02f5e627ba5990c4
1836 F20101204_AADBUK wilson_c_Page_091.txt
5dd1479a86681e180e232276e6212b14
9fa07462cc1269daab05eac33cd652cab531940b
1051929 F20101204_AADBTW wilson_c_Page_005.jp2
50f8c8e8b5d85f92e75a426381bac615
f46457e8dfa15da01e17b38e1fc462231dec6f7f
96490 F20101204_AADCAF wilson_c_Page_028.jp2
e5b499a01ce7aafdf09bba4b16ed84ca
1b109f6cf5c18af5b0e7a425f1a16acb77fb37b3
15695 F20101204_AADDDI wilson_c_Page_038.QC.jpg
3134f13babfbe73b1f18a4c5622195a8
c121537d26da61d509049bb4ba317c777443bb61
21629 F20101204_AADDCT wilson_c_Page_020.QC.jpg
c2a06e6980d5a6917e61b0c77ea55746
f98d6dfed77e74d35b2373270e34d2b4f4f76f2c
37525 F20101204_AADCXN wilson_c_Page_150.pro
946a02622175a28096e3e150d8b1b762
5efd10d7b19667b4c1cee2f86ddc01b1ffa3004f
48144 F20101204_AADCWZ wilson_c_Page_102.pro
35b3a49d9f337cbb5b907a544c898ccb
88b02fb5ced17676e1d76bbc58b91e1b5ea0d97a
26084 F20101204_AADBUL wilson_c_Page_047.QC.jpg
416f3f83c3575d06f6f497927f3b0216
fd3217430873e37263bf0153f366cfd9f9bb794e
22243 F20101204_AADBTX wilson_c_Page_071.QC.jpg
1bd7e4db9c6da529f88c085f075c126d
86b4a950eb319cfb9ecfebb7513abecdd0224b56
F20101204_AADCAG wilson_c_Page_045.tif
57ef0a4f0f575a49e64a33594b0e4e3c
4490d99e5e1f8f80615ee037fe88191b8ea35ec2
4224 F20101204_AADDDJ wilson_c_Page_038thm.jpg
26717558a1850393d97ea388661f17b1
5e11b953927031bb63b1b1bb8f05d59b43570690
6001 F20101204_AADDCU wilson_c_Page_021thm.jpg
f838b2768b69fc758bd7304da1100fcb
ae47f7556c951b62d4bb7921834b18727ab0fb7e
24729 F20101204_AADCYC wilson_c_Page_180.pro
d2e26affe349337f0c36d0c856bbe9dd
0696dffd40327c1b860b56f98880be9ee539d14e
52244 F20101204_AADCXO wilson_c_Page_151.pro
9355a488490666f81b85801f870d3062
8c52e2362b9db6927de691678c91f97eaa1b5079
665687 F20101204_AADBVA wilson_c_Page_089.jp2
a60fb093702f140951b7acc9edfd1634
6c123fab5e8b781295b41176b068881570ce7e91
15110 F20101204_AADBUM wilson_c_Page_113.QC.jpg
7ca4917c98dd8d51226e2bcaf96d85fa
aecc7fe2b181ff66ff5f5d310e728ce86a90c31c
6965 F20101204_AADBTY wilson_c_Page_069.pro
10a0ad907226eb37468ed1fad67d19f8
8168cf5c59c00a68817388000307ed9acdca97f0
F20101204_AADCAH wilson_c_Page_155.tif
56a40202c65e3c35d2cec3e6da9a28c4
98193b4b5c8fc3707121da8074ab12697e92a909
14325 F20101204_AADDDK wilson_c_Page_040.QC.jpg
da155c06ebd38ec516278a5a858ff060
71910dfd82f7d309359b74eab48c9925cf339988
5946 F20101204_AADDCV wilson_c_Page_024thm.jpg
ef6b7d4906f07ae083242da9c26e7455
1f69573e0c8091eea111a35792a95f6b10470a8e
55611 F20101204_AADCYD wilson_c_Page_189.pro
4c79c4180ebce406d0f2d37b8d133a27
1666cddc7ffe14461666cf11b4b9e5cd72d59c39
46814 F20101204_AADCXP wilson_c_Page_152.pro
772e1d6afe9f092065b6d20429b9442f
67f22add609533767a5d2d9dabc59509085c5d1b
17068 F20101204_AADBVB wilson_c_Page_147.QC.jpg
9bba91a905c7ba1f90b4eac5fdf462a1
b5a6e95431c0e53751399e3f48c629bdaa062259
10801 F20101204_AADBUN wilson_c_Page_034.pro
8016dd9657c7074c96a3e28d2cb9189d
7657031b7f568815e3d93e5f61e0d6c54c73b8b2
F20101204_AADBTZ wilson_c_Page_055.tif
5bf88701cb194f3692ee28e1e6cc6a64
b9be791bfa42ff3e8abf3d69edfa309393936438
F20101204_AADCAI wilson_c_Page_048.tif
2ebc73bf69be5d6e01cb6a962ba867a7
49394efca4816deb1b0d5e3a027c5bcf3814cecd
5077 F20101204_AADDDL wilson_c_Page_041thm.jpg
c0c12facfec42c5013a2b503396ac441
829803a1df60a18e547e311ddc1ef4d9a5b94b10
5258 F20101204_AADDCW wilson_c_Page_025thm.jpg
99f4cc0f5fe47a4e2abf8f960160c221
4b1eca80f22bd2018f83f0b39597726ffca3e1b7
57662 F20101204_AADCYE wilson_c_Page_190.pro
09d9ba976a2088f465fab01fa321978a
86116933be913de8e4d552a6ce4a96dcd4c1dd92
10745 F20101204_AADCXQ wilson_c_Page_153.pro
48a4aec0385613023a5a61ba058d7337
d05245cec233e7544ed51661cd1543d2528de32f
1943 F20101204_AADBVC wilson_c_Page_124.txt
f6a7bff1a61137cc03491800620540ef
fa87b6150b6588158331ca7bba3db2113991de95
98822 F20101204_AADBUO wilson_c_Page_026.jp2
9d10ab352aee44a2f73e9f712fa31cc0
cd73396e9980bd85ed2f7c70c57965515bf5fe63
81065 F20101204_AADCAJ wilson_c_Page_068.jpg
894cef426c8331ef87ec502b68e9775c
0fe2426d506b1fe4f720ab447316798f6afd87ab
6065 F20101204_AADDEA wilson_c_Page_060thm.jpg
8a7dd6b9db5094b4bb1bfc5431ef474e
d0b23cf913f876166e8792cdfd7b7ba97369455b
24096 F20101204_AADDCX wilson_c_Page_026.QC.jpg
077273bdf03563d1ceb4d0476b8cd26a
57857d1c21e803ad7b7d889c66973745f155f658
465 F20101204_AADCYF wilson_c_Page_001.txt
0247449a7e2183a08ffa54939218ba11
a3675aadcff0613a5ae1b3a2f0fcd417ffa73b31
29269 F20101204_AADCXR wilson_c_Page_155.pro
5c17140462e36ab39869d7f467f949e9
5b5ffb4b556655130df59543b01a6612dcd5e0b7
43053 F20101204_AADBVD wilson_c_Page_039.pro
9e3cbd271fafda873635cf98bf1ce15d
667145d9aaf44a0d8d03dbad267d58773a8ca9b3
5996 F20101204_AADBUP wilson_c_Page_068thm.jpg
eac7b0e0a94d80a6d32be382518e764d
864d993ba5ca8f4875994b6bca014c44dd8e5ab3
16979 F20101204_AADDEB wilson_c_Page_062.QC.jpg
5dd9145bcc1d7df6e03c2700ffdedae1
8c802983480b633454f3b10dd5e9cd66dbba2d46
19581 F20101204_AADDDM wilson_c_Page_042.QC.jpg
6c800d52314a85d599d5788f7605c845
ea7a8ab3916c76c6ca68fd8a095621045191db94
5724 F20101204_AADDCY wilson_c_Page_026thm.jpg
3fb53cccff4b48d54701bec8acb76632
2a83d56429956ddfb9d50a4dcffe3cb8fa5eb292
1282 F20101204_AADCYG wilson_c_Page_004.txt
7fae3c3339f9e718836157d434b4befc
f1ed6877e45ca59025a38ad455357b13c18d95b1
38503 F20101204_AADCXS wilson_c_Page_156.pro
18f7f5c434f88975725e323512a34f3a
1de27b57a0cbf4fce3069cfd553a93d8d1ff7891
1615 F20101204_AADBVE wilson_c_Page_098.txt
ecee1a0e47ff63c4dff670a51fc06bde
7c69082f6ad1d7fc0c789cae212d05a334c8b6dd
6085 F20101204_AADBUQ wilson_c_Page_120thm.jpg
3133489c46df9d128c45ff32cef744e7
e49d9d99fc096b47eb75d37ef20b2de81f56377a
F20101204_AADCAK wilson_c_Page_105.tif
e83efe99d28bd33493c7049caffaac7d
1dfd266baddc8551c23a5d9cc802e669a0b932b5
4590 F20101204_AADDEC wilson_c_Page_062thm.jpg
aa668df41e768e470bcb2c672e8fc5b2
062fe51d55276c548fc1a99bc58ca4e6628a0a16
5017 F20101204_AADDDN wilson_c_Page_042thm.jpg
b4187c105116d747accc0274c1db76da
78176971c470452328ab6d880b1f3aa1638ef17d
22429 F20101204_AADDCZ wilson_c_Page_028.QC.jpg
3e0907ae64433b298b3eb9d09589b0b0
7bd3834dc1f802b90bfcd572f0de82bb0373de58
2878 F20101204_AADCYH wilson_c_Page_005.txt
9a5c767615ebb42388729e54fbf53136
7dffe0231d6fec5003986c71e73b77ff26e9b6fc
7167 F20101204_AADCXT wilson_c_Page_159.pro
7c36e8ea08520c4369ad3e1691566362
1daf37140d9dfd7af0161853b59b60b3345ff2b2
1857 F20101204_AADBVF wilson_c_Page_152.txt
bcc6e0c4d375cc0962bb74f513f0e4bd
4785fb32763e213026a3ad8ae715f0a758010403
21871 F20101204_AADBUR wilson_c_Page_080.QC.jpg
ed32a5c2300755f8795903b10d22c0ac
aa4df703b351d8802657629ac87c75328642d2e6
4238 F20101204_AADCBA wilson_c_Page_104thm.jpg
52f9b89c32ab0a706f289b83cb2367d3
1d283098887a6cae3f8ef6945da2e4c3b3008faa
F20101204_AADCAL wilson_c_Page_063.tif
ed651d2f2ad1fca4a8ab208a3281bead
5e1cc32a0195644df2b6ba1d784e560cbeeaae9b
22368 F20101204_AADDED wilson_c_Page_063.QC.jpg
1d00f659306a24d1238943476c8487ba
33ac4ef94978ecaff1621ac0d48a584d56b313fb
4592 F20101204_AADDDO wilson_c_Page_043thm.jpg
2eca3206096a540abf383cc1e01c871f
fa91f894fc37d0554e6688f0b593777bb1d4c76b
3765 F20101204_AADCYI wilson_c_Page_006.txt
91cc45bb6783287a876ca0f16b56bc66
118d95784b0b7f2277050235d07522e7dc4df3b4
14263 F20101204_AADCXU wilson_c_Page_161.pro
3614ee404353e119d091267139ffb6ad
a446ca9714d3255a984f7c3344774f10afd86880
70217 F20101204_AADBVG wilson_c_Page_004.jp2
e0c746913f1db33880bd55ec43ba2aec
4d9ffc0f8ab978c6a45eb19c951e4bbb2c3ab186
37680 F20101204_AADBUS wilson_c_Page_161.jpg
0d25a0992283d3cce581ec3e6cb149f4
08fe59f43c2f6f50b87fcb959741d9cf51b3c14c
81627 F20101204_AADCBB wilson_c_Page_156.jpg
5b7b3b8fe741b2ed899d90f118ed7230
9e8df1d7748260edc4de0b8b9d918e08cfe1430e
17099 F20101204_AADCAM wilson_c_Page_088.pro
34c9df28de2a68974868c7e21b50fbf2
add0efb359ba05c09e00354967509c36e1cbd898
3066 F20101204_AADDEE wilson_c_Page_064thm.jpg
b6599841cfc40d37ea2d6c4b447e5553
5584c4b989048c6d26dcb8fb498026428472e5d5
5159 F20101204_AADDDP wilson_c_Page_044thm.jpg
bfc36da5604630d8c8f1903cd3524270
a83dac5651c799d4f8f432b3ab19d2e73aa047f1
803 F20101204_AADCYJ wilson_c_Page_009.txt
4b25764316e9d2a68d32b0596c0aa861
a86ee11cfc5e6a3de0a8000f24b7c93644b46d5e
8623 F20101204_AADCXV wilson_c_Page_168.pro
3be9db3cdc4e90fcef6fb66a70eb1cc8
c253114fc124cea9c664bca90c2021f54b2fed13
351531 F20101204_AADBVH wilson_c_Page_105.jp2
87ec7ce63e9fa95d44fba4c0be53e997
a769ae76027356896abcb585d4c523fe3963987f
671730 F20101204_AADBUT wilson_c_Page_142.jp2
671de04ad84584141005cc0cc33cb0fc
0eca1acd5d26aa11fea1be4ffc2555a22a03332e
3142 F20101204_AADCBC wilson_c_Page_069thm.jpg
9f3d26c050f3ff2e217f63fcb309e0fa
b5baea995a3d827b98af7bad4e026765fe85ed54
1714 F20101204_AADCAN wilson_c_Page_117.txt
a24451d75188902d1ce23685bfed06d0
114705a5f10369f8bb34e7f02e22828ea8e36cb9
17103 F20101204_AADDEF wilson_c_Page_065.QC.jpg
49b2c40ed24e94195e84c1eb4065fa4d
02e008baa5ab15aaf0b349b1336b6905c2b705aa
5148 F20101204_AADDDQ wilson_c_Page_046thm.jpg
5050561563fbd1d5baf23aca3dc7b0a5
ec4453a258053f5b2357f6ba263aceca235a277f
2282 F20101204_AADCYK wilson_c_Page_010.txt
fd43e77b12f21972f64723617262e457
c5feef500f8f650c8dd32fb7947b7813c14e654d
9183 F20101204_AADCXW wilson_c_Page_169.pro
9cfc0bc053deb4dae1b4455eca6d32d8
d0fa3febdb2454e1df96836bb0d999077f5e9b7a
69543 F20101204_AADBVI wilson_c_Page_037.jpg
59811fd124fce79d2e1d776f0362bb2d
d639d1010927253baba0f72e960a2919bb83469e
78767 F20101204_AADBUU wilson_c_Page_120.jpg
095307ce3e686f9cdf592853d2cbb328
ff30d43566a6b7cef1ae21b87a6a592bd62b6331
20121 F20101204_AADCBD wilson_c_Page_078.QC.jpg
7051a5779966db5c76a3764cca30eae4
156360ddbab647edcfa210d125c724f4e4cd0985
6018 F20101204_AADCAO wilson_c_Page_138thm.jpg
6a0f0caa4529a00b60ce593ae6ce46c2
e03d86ba2cce2832cfc4f6b7e8fb04d72b59c15c
20714 F20101204_AADDEG wilson_c_Page_066.QC.jpg
431518c85dd69290ec40dc9ca71d3e3e
36f095911041097cafcd8bc594c089c20a817578
6444 F20101204_AADDDR wilson_c_Page_047thm.jpg
13e845eea7a396cfe6377e145fdb7e81
9dc4bd2a8d60135bd3f4c279d5de4cb8c207628d
466 F20101204_AADCZA wilson_c_Page_034.txt
840716f79653c8b7a3bdcc04b426e5e0
e222666aaa8dbbdacef5e4944983a08da7534aa8
2884 F20101204_AADCYL wilson_c_Page_011.txt
6e6d86649fd1268cf469c772d493739a
24682c6109aa26c84f2a2cdf2b57f660219a7350
9057 F20101204_AADCXX wilson_c_Page_170.pro
6e91436306d2d619726ab500676d39fc
87faa9dd7206e1f89126bc767a1be82ada61f757
51023 F20101204_AADBVJ wilson_c_Page_047.pro
a306a87e009e841d14b4511df63b41a5
b3cabbd786bb03dc0364ab02035aace31e49e595
11375 F20101204_AADBUV wilson_c_Page_126.pro
70dfb3691349d62eb2208caa19cf7036
97ecd347f8582dd97a68fdbc54cfbfe4b8c0dc60
421388 F20101204_AADCBE wilson_c_Page_149.jp2
380b43ac0868790468423a09c1c4cc97
facce2f5adbfe2cd31f28bd4a902ac78427fbb35
614517 F20101204_AADCAP wilson_c_Page_148.jp2
de743333dee8afb497a92ba206c49100
ca7ef1e0a3aa504b0abea2f2b72935d3933183ab
5826 F20101204_AADDEH wilson_c_Page_067thm.jpg
7bab07e045542210eede23f537df58c2
ede762dee55137ac73c465f78e8d0cb6eb1df9a9
5901 F20101204_AADDDS wilson_c_Page_051thm.jpg
3cb5bf31073bc5d38811e5eba1d9213a
fcd9c6fabcf00ca3d084a7f931b3f6ad813817c9
1162 F20101204_AADCZB wilson_c_Page_035.txt
6386ef2a54b51f0dd671a2b8cf45222a
949153bfed18270ddfb7f50a5aaad5a80d0f6ea8
3032 F20101204_AADCYM wilson_c_Page_015.txt
93319c747bbfe24b7e6f1bc06c400c47
106576447247fcd842ac6545c42ebc1d5effb69f
49573 F20101204_AADCXY wilson_c_Page_172.pro
0c2c249d23212c2c61ecb31a027a77af
4b2c744514df395d481fe2cf7a4b5f286a474d14
494069 F20101204_AADBVK wilson_c_Page_153.jp2
0a06e1468c868a17136191945c4f08bd
71ad2bea1f150f0b8cc1e75a890374c41228da11
818374 F20101204_AADBUW wilson_c_Page_043.jp2
aa53ced88c7a9a9ebea430f6acfedc08
3f0d72ef3259e7fba0c38bd3b89888464fc28344
517806 F20101204_AADCBF wilson_c_Page_017.jp2
78fda5124b8d2f9687c1e484ee8f719e
b9e63dd08a7adf51f4c55da0cf2ec7121fe40b00
85961 F20101204_AADCAQ wilson_c_Page_151.jpg
b2edd4ffd2d8439d20853154ab76573b
1f61dca30f87e4207a555559d7cf286af17de3d8
25180 F20101204_AADDEI wilson_c_Page_068.QC.jpg
a55ff06eebd64da7a55d7f7313181b1d
10239e6f60e9d66736a522375f6867df82501078
4406 F20101204_AADDDT wilson_c_Page_053thm.jpg
7057eb1cb3d333d45e23e3c3b390608b
634c661cd91bacb6e8d9518566efe564758fbb0b
F20101204_AADCZC wilson_c_Page_036.txt
1ad511017700d2a63f3825fd8e3a49d1
a8b9d7a1f050dfe5e89fdaaa4f5e3c2246a4a4c6
2939 F20101204_AADCYN wilson_c_Page_016.txt
ac0b2abfeccd748a4494f1fd49a43e67
8f2c1f57ed9ee9c0eed4b8f5cbc2d56459a5f07e
5050 F20101204_AADCXZ wilson_c_Page_177.pro
9b475eeda64ce6c5dd6f4e1ba2d892f2
819cd12c132ca2767cce396b58f5d9db8b5f031a
38487 F20101204_AADBVL wilson_c_Page_125.pro
1d4152d6fda07bea7f6c5155d6da6037
cf0818c5dc7214edd4cba118327109d37f73ee56
1722 F20101204_AADBUX wilson_c_Page_071.txt
21ed3356994e5eeec83ecba7001d56a6
f362da1e5865329eff51e72efcedbbc5412d4115
97104 F20101204_AADCBG wilson_c_Page_085.jp2
6ed255902fac22aec44bdb5a5acf3f0b
827fa3760682af918e0d37f4d3b29c1a0cc46b11
35485 F20101204_AADBWA wilson_c_Page_141.pro
94071666db801e9757c10c89e986c043
a4e924ab0bc6e9c182a2d494899e672b4e1e02e3
13864 F20101204_AADCAR wilson_c_Page_106.pro
f6b584413fd9b4ac49e2977fbc02ec76
5b0988b01ca124f5a8b40892012d07d4a0af03e5
9545 F20101204_AADDEJ wilson_c_Page_069.QC.jpg
80a0f2fae85b16720b2c53b7a2dc3c03
522aff2193f85a70762e231fa5588dbcec1df72e
23458 F20101204_AADDDU wilson_c_Page_054.QC.jpg
9d820026447e2d379d859da812ca066f
273efa390ddfd5fc9bb155751c8f0b0bae919ad0
523 F20101204_AADCYO wilson_c_Page_017.txt
2e060149a98d5d4c43524bbb31b4f6b0
da2d460b131a7b24403630d26f46aa6ed07156be
85294 F20101204_AADBVM wilson_c_Page_117.jp2
2a0b779c3511cda565fa375930205cbd
5b007f8351098cd4513526cf99a66432d7558bbe
23587 F20101204_AADBUY wilson_c_Page_163.QC.jpg
b7532f77e7a2866bb8ba7c70b9802e59
5cf303f9c57efa897bea9ad05faf89760bb696ad
13660 F20101204_AADCBH wilson_c_Page_186.jpg
3d94990f321d7ac668fce75f7e801f38
9eccd015ee8ea6a14fea4e2325893166e81e3af5
38587 F20101204_AADCAS wilson_c_Page_080.pro
2339d0944d6246ed8269665fd3db5c5b
ae3566498e214e17e3d9f2ecab0b2c15aafa94bc
22172 F20101204_AADDEK wilson_c_Page_070.QC.jpg
d2b43767b218c9962c151f4a452cfa43
9677db9e7f443756066335f752828a90b6d0d257
9644 F20101204_AADDDV wilson_c_Page_056.QC.jpg
4af3bdc4469b10940c6a9f3cd504a0d2
18e29fd46d93d54e590ea0968a108710a4ba8ba6
1596 F20101204_AADCZD wilson_c_Page_037.txt
d9093cd040b630c894d6e10ae35456dc
814314edf339d1caeb2d942f6ccb0eab06122dfa
1677 F20101204_AADCYP wilson_c_Page_018.txt
da4e5053d6c61b2d7bb1b2e3a3a906fb
69b8f5bd7bf5377f26a8617574e736374cf72cd8
91839 F20101204_AADBVN wilson_c_Page_010.jpg
852d1ff6123435bede9db609a65da0d3
539cb31fd2b5ff83a6a9985d7ce3dd4e2f966043
21161 F20101204_AADBUZ wilson_c_Page_148.pro
7571bf011ad02e68dbfb96b8acef0a37
40649abb0019fa36e1b1bf9030b017d0169f7f0b
F20101204_AADCBI wilson_c_Page_077.tif
12b9206b597ef89489cd52de2153a130
e361e6b691170b12ff3e7924ffbc10320f7c3374
53567 F20101204_AADBWB wilson_c_Page_188.pro
ad2e58eab2e7bfcca222b2f2360f2aad
18cba2d41578051a9fe93348537edb8281633fef
2037 F20101204_AADCAT wilson_c_Page_115.txt
a704ae1ca13e16a2b2e991e6a6ad5f71
7e5eae9e02ecb41d28779ccc9dfa14db669c610d
5872 F20101204_AADDEL wilson_c_Page_070thm.jpg
8f757d4bcfee91661908eaf0dfc03f55
fe597110d09da5ccc323a189bd477b6b73762719
3210 F20101204_AADDDW wilson_c_Page_056thm.jpg
7f7213e9874b99e07e7d0e286193ee6c
de30c68f5c8d324d74850640ca0a5f221581e2f1
1741 F20101204_AADCZE wilson_c_Page_039.txt
a736a5d67106f86fa82b628daf0d957a
e4d68aba2b8b621b946790160d1c9ff9359e7731
1893 F20101204_AADCYQ wilson_c_Page_019.txt
aca4350181695345e2ea5f04775911bd
2de8808a5a90a48854faf7e5b1038ca0549b5d17
95282 F20101204_AADBVO wilson_c_Page_063.jp2
1c83be667170f31717b863122bbfab31
183f744efc79889594a428c0f45532b1a77295cb
74707 F20101204_AADCBJ wilson_c_Page_130.jp2
f067b66d637e2314b50a0280dffb023b
41d97784af290e86cb8f7e7481f16e2da162ba53
17201 F20101204_AADBWC wilson_c_Page_004.QC.jpg
125ad64ce22ad235f6825856f7d0ef46
5fb0377a3db2ba843e135e9b8222535825b33615
1886 F20101204_AADCAU wilson_c_Page_120.txt
d0cc52d1a9374d59eb65510421758716
cade736a74e2f9b954144f2a1041026348ef5b29
4975 F20101204_AADDFA wilson_c_Page_087thm.jpg
0a9981002d34dfdb62490a084b5de8de
7093371ec6d292e8187c7ff19bb676064543033e
5347 F20101204_AADDEM wilson_c_Page_071thm.jpg
f651ca65883e98af1c6614b9916eb8fb
f90296149f49f9d60b94325738adf00a65093bdb
25061 F20101204_AADDDX wilson_c_Page_057.QC.jpg
938bc6da679d427526f510ccb31a5062
0b4fd31a1890f5b86ad68282d028ff2362072b67
991 F20101204_AADCZF wilson_c_Page_043.txt
6cfdddc0092acdfa011f0f2ac3f0418e
f2a603b5ca626c8baa63bee257fce559eb3859e0
1909 F20101204_AADCYR wilson_c_Page_022.txt
645acb0458f9e974cc4cacacffe63fb7
2e54a8ea3a6dc089b1194df06101aa88167cd9a9
5785 F20101204_AADBVP wilson_c_Page_083thm.jpg
5bc9c975b109cb9a28983002e2326984
7c62d44a8820ceb3ee95a6b18a8d36de257ded9a
24327 F20101204_AADCBK wilson_c_Page_120.QC.jpg
f17dca37c074133d5effd23ddb8f4fac
9ce63bd4749782850b2aa56d0b981f0e24c0598c
96882 F20101204_AADBWD wilson_c_Page_190.jpg
cbde57dab442ceaac6c825c9363e3f17
cac80a110fa53a2a2780793929a27eff4a006376
6591 F20101204_AADCAV wilson_c_Page_185thm.jpg
87a46fad885a8666848fcebb4de0d17d
6ca2b2565bb44318b3e14eaafcb9f9d4bc008543
5009 F20101204_AADDFB wilson_c_Page_088thm.jpg
04ad56ae55b7c0a53e56c1e9c1bace7c
67f76dab1a4217b46ac9635c71d14fc703980d51
16959 F20101204_AADDDY wilson_c_Page_059.QC.jpg
5648d69e9f20608041810dc3be87e6a5
974fb32a7b60f77e8ed9f5428f48828733fa9e2d
1645 F20101204_AADCZG wilson_c_Page_044.txt
d7b3807e76ad6f719becb7a292158f66
06a15a2ea412432f11487b2015131a79c40dd903
1646 F20101204_AADCYS wilson_c_Page_023.txt
63352e756ff38a98b67f09a100ef4606
7162aa022d651c28b70f542b35c7e319feba1919
81624 F20101204_AADBVQ wilson_c_Page_171.jpg
1b56e26956610c8ea1ffeb53b134f51d
769bc18109b920542325c973d3555c41b74c34cd
103492 F20101204_AADBWE wilson_c_Page_160.jp2
70eab48abd6dd9244db10ee1e8ccae88
4879c3bbd6126769e89880e0ef5468fefa3429c1
24630 F20101204_AADCAW wilson_c_Page_134.QC.jpg
c20007d0d0d5e40a40e2356bcc29db75
25373fbef0789e634ff85ee454f3e8deef2bb7a5
5505 F20101204_AADDFC wilson_c_Page_089thm.jpg
7ebf017d90e6f78d5cb6f0c2f7a10d35
6213f3820ca6960da77e0b639f844095d5b7d08b
16583 F20101204_AADDEN wilson_c_Page_072.QC.jpg
b2845d7f4bd4e3ce4aee31e04007a4ae
ba8ed2e58669636e72202fc02f469a92a8e0f364
4807 F20101204_AADDDZ wilson_c_Page_059thm.jpg
a16ea47a855a25f138bf6c01be761339
70d93e55bc8fa5fe7f89c6673175726fa6139c57
1122 F20101204_AADCZH wilson_c_Page_045.txt
062092bc6a2054b566638a0759e30a95
720bdcf8d15dab69c0551525b289b10322d96fc2
1843 F20101204_AADCYT wilson_c_Page_024.txt
36e1183da67d0bd9f4c17aff6a44c497
ef097f24e4d066252cdc91a71caf42b384b46828
6043 F20101204_AADBVR wilson_c_Page_019thm.jpg
0eec8cb126971d75704c44e8b3f6ae8b
3b657b6c0675d9f0b25f36547cd56527dbe036e1
22453 F20101204_AADCCA wilson_c_Page_132.QC.jpg
bd31095c0f99095a8cb2abc97e02258f
de30709cd55d057f40d68b4cdb480a35adf49068
60463 F20101204_AADCBL wilson_c_Page_173.jpg
1a55b7d417688fdde2232af7c39abfd2
7732e15b5db071de51023262611c3e0985dff6c5
94992 F20101204_AADBWF wilson_c_Page_189.jpg
e259613cebc86cb731ae837764b2e09f
9513384ca5dfbe4fe8871316f074380d10c5b882
23405 F20101204_AADCAX wilson_c_Page_166.QC.jpg
55a73a1a217c43ad42463b97d0d54432
17991393392ed07f0605d49e3b2bffd9376e9745
22815 F20101204_AADDFD wilson_c_Page_090.QC.jpg
a9b010c88a888fa2697a6f64c08725c9
d452af670d4f01a360461e7ea531491fdd703cad
4809 F20101204_AADDEO wilson_c_Page_072thm.jpg
9eee0842ecb4555c21ee1d4cb3338df4
aa0a441cb1351176f54123dca109c8792a850f95
2066 F20101204_AADCZI wilson_c_Page_049.txt
5fef6f45745980b7f9941bbf5f449e98
0364293bab9842453f250357be207821acef28d7
1454 F20101204_AADCYU wilson_c_Page_025.txt
1122aa7f9daa857553e59b1d92e14674
fb9fe5687c233a4899026546c0be9ed5f0608e65
91263 F20101204_AADBVS wilson_c_Page_110.jp2
58c445664cb7faa705edfa6fc0ef9655
4f40b43cf3c6e8d9d81da60aa18495e37d0a8965
21495 F20101204_AADCCB wilson_c_Page_070.pro
b111ad2aa18c3b04db7d9bb6e1626899
5ee26371e08aeb5cc15717a983a45f4734348b42
72310 F20101204_AADCBM wilson_c_Page_132.jpg
52a1e2c488cf2904f4ceaa1021b9950e
c58697324d271c1e339de8886505c350f9df2598
4808 F20101204_AADBWG wilson_c_Page_147thm.jpg
513d25844bb00ffe2b456799d92d589d
fd1d8c9deae7259aa8081b5b01b3623f1ec7aa04
1051977 F20101204_AADCAY wilson_c_Page_013.jp2
529f1f80cdd389cdad36e6ad682369d8
f17bca90b2ab78de72e0b679eb99a255983fa899
5643 F20101204_AADDFE wilson_c_Page_090thm.jpg
c523107624ada455e0ebb529dd242fa1
39c66534337a4aa42ed9b20bc76d54d66efc5c63
20232 F20101204_AADDEP wilson_c_Page_073.QC.jpg
835b20819d08fd3fa67b5e6c6a5f4505
4c69e5c089bcf41173c34f7a2636026f722ac5e5
487 F20101204_AADCZJ wilson_c_Page_050.txt
974b7b90e7c76835b3d0dc3a8f782713
13966afd2ad07cc3c20d9f8fb6560fa4a3abc71c
1802 F20101204_AADCYV wilson_c_Page_026.txt
fc1d9029b03eb8c0b99e9b00ef4b6aca
21f32bd0d75d5de4bb1d945c752062a1a29f7a0a
106722 F20101204_AADBVT wilson_c_Page_014.jpg
36ac950bc1566afd9fc6d6e27ca40954
94217eefdd73dbb68824d10ecea0bfa120c88092
F20101204_AADCCC wilson_c_Page_041.tif
165e78e5717088a32348a421c2e70841
b81219a18d60254888b8a1d39c11e621420d6478
49687 F20101204_AADCBN wilson_c_Page_127.jpg
93f92802708bd05ea71163ba18559e29
4b683f8f03c8fbddfb331430391062441486bed9
50973 F20101204_AADBWH wilson_c_Page_183.pro
99e473ca170841ed68455207d5d948a5
aef954dd592cda402e03d5e6467e18c40be43a2d
73417 F20101204_AADCAZ wilson_c_Page_039.jpg
7573db5cb0e13b859b5dc323efbd204b
196608f169d16e6b31ae1cf4a31f36f8e36b1960
24140 F20101204_AADDFF wilson_c_Page_091.QC.jpg
bea0e63b85f620237d093ce498f55594
668d8115579e184639aca1d3f0fa1fef5c009292
16916 F20101204_AADDEQ wilson_c_Page_075.QC.jpg
a681c5f265e64df5d1482bf5153e3503
14309151c43f3bf22205834ccb98dd61ceffa40d
1139 F20101204_AADCZK wilson_c_Page_051.txt
f37c2bc240fe834a2187fc5f20026960
53895097bf6736e2df4af1925a3966c93cc6a52b
1781 F20101204_AADCYW wilson_c_Page_028.txt
c948e1a7278560fa7f7e960336ec089b
7fd929d2b2cb09b33c84a3dc2d09806b3f89bfba
F20101204_AADBVU wilson_c_Page_154.tif
2a2281230fda6b0f18087a11e444f531
9d117f77da70c71a6cd7af76266983db69fb56e6
77567 F20101204_AADCCD wilson_c_Page_119.jpg
907e428d10511e9c6c7c42c8dcb4b5ce
ed61639a4d5e8b0eae61af580112d34aec8637c4
11321 F20101204_AADCBO wilson_c_Page_123.pro
9a97636d1edf542d772467b9be1f4960
88bdbec49a6fc46fb600f8474592f35714b19e20
F20101204_AADBWI wilson_c_Page_099.tif
a545a8e3b3221ac2e1d2419327f6a202
5fc6b41713f799bafdab5020ea99c6fb5d357bac
5983 F20101204_AADDFG wilson_c_Page_091thm.jpg
17890dccdfd02a1c4ef9cac0a50a9f4b
a9b3cd39a9b05f5e0e37e3cb323f24c7ac97e383
4575 F20101204_AADDER wilson_c_Page_075thm.jpg
111ceae3426d01293dfabb087a250f5a
4e6c0fcbf0e7649b34300ba4c9c97f6d12bf22fb
834 F20101204_AADCZL wilson_c_Page_053.txt
7da99afb5e8a7b02694df848b376cca4
4bfa4fd83e22312d011aecd0d8c7913c3dc35f70
1768 F20101204_AADCYX wilson_c_Page_031.txt
5e75e8b87fd75863b9e3412af0ec6fce
3564e0ed7cdca1a388a89623f409ff956504bdc4
27623 F20101204_AADBVV wilson_c_Page_109.pro
d15d924f28a794f3ce1c1dcb23b3297e
f09bde85f179eea34265aa4f0a5895277250ac08
4295 F20101204_AADCCE wilson_c_Page_170thm.jpg
626673a19c3104134cc60517868bbfc1
084330cb8016095bd22ec513c7cc6089a2046a37
6279 F20101204_AADCBP wilson_c_Page_188thm.jpg
a5f0be299bdf77624c3cf1bbb5915b79
446ccd8444d0d61cd3395d8b171d0cea2c58d435
36698 F20101204_AADBWJ wilson_c_Page_177.jpg
cbc26e268dedf4105a70b5c62ccd199a
514a0f8f99c8f2661ed6403aedcc26c21ea5d191
18949 F20101204_AADDFH wilson_c_Page_095.QC.jpg
4328e1d683f2a7f8712aebcfb88a8141
8f59c6d6aa8ae6d263dcf795ce89ee30bc3974ac
5962 F20101204_AADDES wilson_c_Page_076thm.jpg
8323c394d0dd774e0824174939a2356c
54e2bfe7f9e8d328f7c6ca294e3f2cca8e07b405
F20101204_AADCZM wilson_c_Page_054.txt
6aa78d7c841a053ac1d10d77dc10d727
49c42c2de69d74de511e06ae2117b2333fe32bd7
1839 F20101204_AADCYY wilson_c_Page_032.txt
d3665e6d00e44f303d77a27cf364d748
5f3b46238108b4acaf55eb1dc0879068435039cb
39173 F20101204_AADBVW wilson_c_Page_037.pro
16e3bd378b844ae16e6823e143767712
83c17b94179034b3d23e0e2f560fbc77c769d06a
24025 F20101204_AADCCF wilson_c_Page_165.QC.jpg
debc0a7dad22ecc93c05d8ac796a8695
cd8e919ff44719643482b718aedfc4c6b08f04e6
6423 F20101204_AADCBQ wilson_c_Page_114thm.jpg
3e932e5b0b36018200355b621c948454
b7dcc4f240cbf7d27b8ff5a610562bc6a4e8e89e
4276 F20101204_AADBWK wilson_c_Page_136thm.jpg
94b95381a564589cfc11b9f96ddf71f3
35a1e87a3d68ef1658c9e224a4b34fbe3d126f5f
5713 F20101204_AADDFI wilson_c_Page_095thm.jpg
ca1c7ab69ff6076d4006cb830ca56e21
b8a294b31291b223ba2eb140504a09693088a07e
7204 F20101204_AADDET wilson_c_Page_077.QC.jpg
a3b23665e2ff4a1f2cd39434cedb490b
947af719a81717d143a6bf42185f2272aab0f432
1914 F20101204_AADCZN wilson_c_Page_055.txt
d2e3158daf6dea033609f7a63ec173c6
d298206500662659cd29bcc391ce028003d5d605
1819 F20101204_AADCYZ wilson_c_Page_033.txt
41dc7732aca2b240b22a25513b548000
ced3eb84926265987f74cd84904a3c0b1e1e9db3
F20101204_AADBVX wilson_c_Page_058.tif
76ff5aad70078763c99a76dae92a8201
3dc56f6318feea1f38575e18a9ab356cd383c477
F20101204_AADCCG wilson_c_Page_028.tif
3ff6122073285dd1977e00f3db3c9784
6f4c3b086d65aedcc06c26ed0a6c192b7cba247f
F20101204_AADBXA wilson_c_Page_007.tif
c7f7f21324cc38c41c26d6a10c2210e8
2babfd32d765810452896b52bb26562bada35744
103942 F20101204_AADCBR wilson_c_Page_102.jp2
20bc298f214d2628c49737d15469716a
957eec83ebd47754334fdfc4f31d786bf31caafa
10053 F20101204_AADBWL wilson_c_Page_104.pro
1004f2febebc281b34e27e34f4196247
143f9b5447b281f21403b5aac5f0693b013273f6
20409 F20101204_AADDFJ wilson_c_Page_096.QC.jpg
fc6708b1df04c668e66ba5bee0ed69f7
79d88e692fb1195f39491dc8add7109129a269e6
5154 F20101204_AADDEU wilson_c_Page_078thm.jpg
f7f434b427a4573564483c1f805fe836
bd601beeb1b55d6571daed4b685c8e0a35c13cfe
1890 F20101204_AADCZO wilson_c_Page_057.txt
f8fc109a0ede3ff947810c7899140cdb
63772c019d5577534d9edb862677a453bda0a87d
49354 F20101204_AADBVY wilson_c_Page_146.pro
bf2f6aab3c35557e35cdeaff3800182d
29b6f00ec33546227af23a6e2fe32e8dd6f8ea1a
31493 F20101204_AADCCH wilson_c_Page_165.pro
2c07597732e6888e213b6142d317d013
7e3c3aacbd3c6800e611da58b3d5fd7666e5f7c6
1008 F20101204_AADBXB wilson_c_Page_144.txt
65f903d55f6b27a6f6a1d885c4c0239f
14de124d03853d35203a224aa4300a4b8949db22
46109 F20101204_AADCBS wilson_c_Page_027.pro
d4bf70d8aea9debcaa49db75e636fa9b
5eeb75ac71e5ecd3d6e455871cf38514b73d4546
79677 F20101204_AADBWM wilson_c_Page_125.jpg
ffc6f42ed408296fb59445feddf6c2b1
718c234275bfc8952e4343366ebfb4411b1138d7
22142 F20101204_AADDFK wilson_c_Page_097.QC.jpg
128ee70476754809fe9272f1446d42fa
7c7045b876bcaa20ff90d5677631816a4dd0d83b
5556 F20101204_AADDEV wilson_c_Page_080thm.jpg
7e62ea668a539578d48873241c65ba25
2e956ec78a1e1b2589e908a5accd7cd51fd7bd64
540 F20101204_AADCZP wilson_c_Page_058.txt
d2d89db352bd93e57a5164985613497b
cdb87e31da1947ce190f7295af6d2855e8c94bb6
127 F20101204_AADBVZ wilson_c_Page_002.txt
26c5fe8621ec3cbbc46c6fdd20e99eb4
5cb4ee6de4b428cba80ab48bfbba4e6027f4de44
90864 F20101204_AADCCI wilson_c_Page_188.jpg
0b6427e90a35e7907a8d2bc210f29b2e
c029e72415aa4db5f2854ada74583f60e4f3775b
50196 F20101204_AADCBT wilson_c_Page_144.jpg
e37069f4a1f9dcbe8fa23805a286f6f6
68e4841ad5b0183c0042f80006e25ffd9df9663b
3548 F20101204_AADBWN wilson_c_Page_105thm.jpg
11edb5e35c48f97b9cb875dc71642e03
6800bcaea5ba56bf31e4c1afd891dd3004256130
5412 F20101204_AADDFL wilson_c_Page_098thm.jpg
78604b54951f7ac7cafb94755d4ed6da
25a488007c041209d4c6e41bfbbb374a8712f3b4
24479 F20101204_AADDEW wilson_c_Page_083.QC.jpg
bd66c92d23b3d786d216e53971c46a24
003b8137b3df66c4bd9cf6e5e486d66a7aba1951
880 F20101204_AADCZQ wilson_c_Page_059.txt
f4ad3c34730a31470d6dc5e20f5f91a9
163968ec6441c72c836e04d204fb494ae1f0b90f
45275 F20101204_AADCCJ wilson_c_Page_054.pro
6359624baa036f263cc7bcc124a223d8
618a6a6d0592c5cd705c870150dee72a32bc5aac
5542 F20101204_AADBXC wilson_c_Page_073thm.jpg
5a1478b9ec8b39115583c76ce62b222a
6725c4c2f90dab0f8e7a0f6f5fca896ded03b710
24428 F20101204_AADCBU wilson_c_Page_188.QC.jpg
9902b7a306be974f26920f168cbd8d20
332e65dd5411064131e0e416b468dc705b0a66c2
50695 F20101204_AADBWO wilson_c_Page_088.jpg
4f6cf13cab74e0144d635e47b6453799
433082a06f5cdd497116d0d5a71dbf37997df7b6
3843 F20101204_AADDGA wilson_c_Page_112thm.jpg
6e31f644abfe2b3e045e784c29a7a8a8
4c12b9a570a5a7171efce54adac4f894ee54fadf
22401 F20101204_AADDFM wilson_c_Page_099.QC.jpg
32f8163c56457660598df524b159e50d
a9315311ed100cce30fcf82b3fc06c6e08cdec83
24272 F20101204_AADDEX wilson_c_Page_084.QC.jpg
a71cbfa821e60a8e6911cf799cec1835
122d433c9ddc52e68d7bfe552de36f59b88ac8ec
1894 F20101204_AADCZR wilson_c_Page_060.txt
8d10334a94a6e4478bc9013ef0dffd6a
40a0adf2ca486ae5c7a9a33c3ea266d9b5ea989d
23880 F20101204_AADCCK wilson_c_Page_118.QC.jpg
6c588bfaea87daef7076952c6bf0958f
63460bd8d6541951d52cfe9a64eed44986721775
51672 F20101204_AADBXD wilson_c_Page_174.pro
b5b236a421d8b4e2809ff78ca96a1100
db3eff530ace3bd1389fcd2bcb02212b942a5f71
F20101204_AADCBV wilson_c_Page_126.tif
fd3af34056bc9122af846c9e11811f1c
323a016c8a3af97b35e45b59f09a962a4dcf7056
12616 F20101204_AADBWP wilson_c_Page_092.pro
4119fc2486a92c4f30b46725ea7fe7b1
58cf1112810d30c13cf0a0db0fb010a348ef35c4
5613 F20101204_AADDGB wilson_c_Page_117thm.jpg
1746f77ac7a79a2757a74e04b0d32080
5a31750a2074389f149698223d056071f554539e
5729 F20101204_AADDFN wilson_c_Page_099thm.jpg
ed830fa95a05228dc63e66852ca47da1
c2516e81fa08859079cd3bda8ea0c2a3fe9fb9bf
6002 F20101204_AADDEY wilson_c_Page_084thm.jpg
de1f1dc530fb1a88ca7615c694368e52
2b1f8cdd6fc302b4654c70e48364dcc27189e4b2
1546 F20101204_AADCZS wilson_c_Page_061.txt
37fc5a4321961f5800f2e9d9d966e16d
7d9bf1c2488da537d71ae8428bef43dfacf8de11
23858 F20101204_AADCCL wilson_c_Page_122.QC.jpg
77b377ab46de5d77462b56c768665702
a9fb458440b45b1bfaa4b181c20dd7921491071d
17724 F20101204_AADBXE wilson_c_Page_046.QC.jpg
40bb9676d96cabf67b9aa2421c333713
f16fecf54422a93418240d3394101d342473e59e
5583 F20101204_AADCBW wilson_c_Page_086thm.jpg
6088a1b52b69399f6f59ed69e9e0b2f5
012d8a9fa114cef8beed9ad6ba1679d5573b5489
1051972 F20101204_AADBWQ wilson_c_Page_014.jp2
1ac2147f7e682d83acd505d12de04260
0de50e1480b3ae1646b253292b8e56ab2f245f61
5806 F20101204_AADDGC wilson_c_Page_119thm.jpg
dd73b705e698729333bf520b34a058dd
f99d6128981e220d5b0bdff92a3521aedfefb691
23088 F20101204_AADDEZ wilson_c_Page_086.QC.jpg
5e585b2d7c8ce3ac957f8f7ac40a846a
4900a6f62a791c3a01408123cfe8006e4de66527
1080 F20101204_AADCZT wilson_c_Page_065.txt
08356400142913c557acd07cad8c97c6
799158f38b7c404783858498496c7eed34c9e908
79399 F20101204_AADBXF wilson_c_Page_141.jp2
46530e231320eb2441d3fbe212f78df6
9d2f607ff95a7da2cff3aaaba24f2f1db57a2801
12748 F20101204_AADCBX wilson_c_Page_177.QC.jpg
06f7d6c6ca196c3dd603bcc273381705
bb3e64cf9f924b41c919d935923c7e93235f1f4f
5481 F20101204_AADBWR wilson_c_Page_181thm.jpg
81df81dd0bfa3414b93eeaf13edd165d
cb2ea251254409d926e3172f893e5190b854ed1a
25546 F20101204_AADCDA wilson_c_Page_172.QC.jpg
7c68b15d89072663140c749d7ba1aeca
75509d5db18077149afbd0135c15e685b4e32e42
25187 F20101204_AADDGD wilson_c_Page_124.QC.jpg
677207c2def8e523b94ddf37ebb6ba30
2d0c38c51fc72c75338a27a5a73e06578516b3d6
11914 F20101204_AADDFO wilson_c_Page_100.QC.jpg
9f5a24131cac1e4906d405a7617203c6
01f2fa2faf2bd6d86e5032a8d59fdf891a36ae42
1915 F20101204_AADCZU wilson_c_Page_068.txt
bbfe08b49e891efbab30e7693fd0b415
7c90deb4183ae39e68163f70c9e978e85e21550f
F20101204_AADCCM wilson_c_Page_033.tif
620ce293c69b05e23719bbe811baa7dc
a3ebca1e4696071a0fa3ece8f8abb66bf6f8b823
58467 F20101204_AADBXG wilson_c_Page_130.jpg
8912c9febd09712f6dae0e0493f2751c
db5caf0266d68447b0de43759e85ca767d611b47
6249 F20101204_AADCBY wilson_c_Page_139thm.jpg
24a75f83a1826ed974ed4e2acadae0ed
417cf11b97710000814059968243121b4778eca8
62375 F20101204_AADBWS wilson_c_Page_108.jpg
cc6f09c3141272cb64be87513d0da9be
24324c8cdd95a6fd2aaf2b8e81d017068f8cddc9
82960 F20101204_AADCDB wilson_c_Page_029.jpg
ca6f780029a3c8bfd8301ce0117b8529
b99f2537efd8a1a1e384741348278a46c5ea6a08
6215 F20101204_AADDGE wilson_c_Page_124thm.jpg
43b0ead87519ab8aefb77cda310372c6
3112b979e95849f6382aba2e987bee6ce4bffaf2
10831 F20101204_AADDFP wilson_c_Page_101.QC.jpg
b3f6b482b70d31b46929f747ad5e9d65
b190001d3472a6611d1a9826d53a28ce968bbbfc
937 F20101204_AADCZV wilson_c_Page_070.txt
039292f6e4ffe153ca1c08b19beb647f
1018a4711c2a4854f94c3fd7a2ae56ca78ab5ffb
6169 F20101204_AADCCN wilson_c_Page_082thm.jpg
e8d947286d985a8689549b2bc6046680
a30ec663910c5b2a40674e8edc65cc1bea683329
20969 F20101204_AADBXH wilson_c_Page_145.QC.jpg
f98ea7b86a4297ba415a119f3315b394
30e533f12e261b24292403d7ad71a925ceba6674
6661 F20101204_AADCBZ wilson_c_Page_014thm.jpg
997b5129d6a9a95ed84ba381d22b1cf8
80b9020434f8c87fe31ac7a7be59417d096ee071
F20101204_AADBWT wilson_c_Page_052.tif
6fd9c95e460d2762e9d24551ffa9592a
6685ce9442d59b971778b799e2ea90ab06931e67
F20101204_AADCDC wilson_c_Page_031.tif
06b13f0df64a031fd819d9e5071fa748
178170aab3b1958657d84c4854ed587ecb544bc7
5648 F20101204_AADDGF wilson_c_Page_125thm.jpg
3ce4b1cf7cf0aba4d8bbbeda6af30772
fcaf05c666f815b0e3e7381f7e0e7be6526f5e37
3644 F20101204_AADDFQ wilson_c_Page_101thm.jpg
fda3a0691eef5d53ae200383649a5ffe
1643e714d5314cbdcdc1587c96a0181a4f242e2e
629 F20101204_AADCZW wilson_c_Page_075.txt
c4a192308cf0f1bbee9128e67e5251cc
a6ff480155eec02aff8d8eeeb1b0a549b6ad7819
887 F20101204_AADCCO wilson_c_Page_046.txt
de6763a95561da32570749da85a4c502
39c48c85c6a563e7fe6aa6b705f508830ea5f67d
101257 F20101204_AADBXI wilson_c_Page_024.jp2
191e9d26b132fd4360bd7ded0e6d09a9
a83232df6bd9dc7982fe47f20785dde47e799956
15238 F20101204_AADBWU wilson_c_Page_116.pro
f593d2208c7ef060289362bd87370bb2
2c8d3bd2bba855d562cc1bd65a9aeef4ccc7f665
48313 F20101204_AADCDD wilson_c_Page_055.pro
4a163ea56bbd7c31b3965e24ee68e479
3b5be3fdc79e320eb56aada74fd08eb001fb7a2a
15151 F20101204_AADDGG wilson_c_Page_126.QC.jpg
f6c66961b67f2c631805c5488d0e2268
50775ddf5151275e092c42274a6d0eb437408ddc
6379 F20101204_AADDFR wilson_c_Page_102thm.jpg
b2acc2f4b68ca719ecb75b793fb4652c
d8a9434d27cd137f7917228c0f904e3b589910f1
355 F20101204_AADCZX wilson_c_Page_076.txt
fff81c30f745ff868ed1bc048aa012cf
08d42de137e5cce321d7c5f04031f0bfcae7b0d7
103441 F20101204_AADCCP wilson_c_Page_103.jp2
59b16073b74b70a76ea6d999cbfe765b
6079b4ed13a33200b013bbf2da17b5d23112d301
5636 F20101204_AADBXJ wilson_c_Page_039thm.jpg
600de44fb0c1a8e97876032affa89599
b323d407f1c94bccf0acbff898f31e4967c08f18
F20101204_AADBWV wilson_c_Page_166.tif
0755fdcf2b8212904a5d814c24a0211e
bdbb921967851df201545a840c5cc1173db626bc
6171 F20101204_AADCDE wilson_c_Page_002.jp2
2363e0238bd237eae05e5392ed60cd2f
1b7d941b61a1fb7e638ce1c9b9d96f249010b06b
15677 F20101204_AADDGH wilson_c_Page_127.QC.jpg
18e09dc7bcba1ee3f7489b129076f529
96c9569bad074ec489884d67bd8631d68f3f077f
6161 F20101204_AADDFS wilson_c_Page_103thm.jpg
f65e1cdee30499fdbaed069a96f8868b
af003962bcf2b14a636eff103d92aaac85944011
479 F20101204_AADCZY wilson_c_Page_077.txt
17796ae919b462c4a4ad54011d2407f9
3e85a7bf76f544db5d0ee365090325cb3867c96c
988213 F20101204_AADCCQ wilson_c_Page_070.jp2
43f4ed3d52ca42a01b3dbd5381deda6c
195a75635e6b468bbd0ff058a3d97ba35bc72b1e
814461 F20101204_AADBXK wilson_c_Page_095.jp2
160ba3855fbe731f6a1f89020476d249
41713a8fa5e3623919d3caa89e690499c14f6d9b
F20101204_AADBWW wilson_c_Page_169.tif
696069f91ea9083c400ce86fb7db3477
983c63ed66d19ec68b688e72216100b892faeac5
4846 F20101204_AADCDF wilson_c_Page_035thm.jpg
e5c060038c0eac84a548567354fcbc75
1c6ceff8876c60761df466253dc15f77092ba885
4724 F20101204_AADDGI wilson_c_Page_127thm.jpg
e7efe2733092d558cc11d1a42be269be
0b070e224e0c6a417d12be684a2f727a4b3e3339
14234 F20101204_AADDFT wilson_c_Page_104.QC.jpg
f2b493e111adf08e83be68ad6964ee60
7c0caca5e657e51e05d5a971196277837d200c31
1641 F20101204_AADCZZ wilson_c_Page_079.txt
4f99068028f587a5fa4b1f2cf50fbe35
cbd6bde268885af93a68a7a98b2a10738091f818
655548 F20101204_AADBYA wilson_c_Page_144.jp2
1c9df1917f66b28c77f50df9d165ea55
d9ec854c131aa521e8b7e3bfd34bd5c06c4c24fc
F20101204_AADCCR wilson_c_Page_037.tif
c95982b66df55b6842433ead8f661a84
c638f740280c99aa6d58c497a4a529d1e5c5fa75
6593 F20101204_AADBXL wilson_c_Page_156thm.jpg
46c3681009742b08a505a064f310cca5
4eb00ca3b5d06c393f7c4bc3356b27e68de1c5c1
46233 F20101204_AADBWX wilson_c_Page_074.pro
011cd3fabe5c4af3276ac0bde55381c1
d644584c228f8538d1ea62672022b55d80c127ee
80450 F20101204_AADCDG wilson_c_Page_184.jpg
17bca8402e0113097fc1932e39855f74
75c351081483d545def8ac27bba8ba9834d56626
10674 F20101204_AADDGJ wilson_c_Page_128.QC.jpg
388395de05dd4f55531111e200f8c14d
7b9c95684ae5b7f472244f93acff867be0f5d6da
10164 F20101204_AADDFU wilson_c_Page_105.QC.jpg
db9a1515f23bc059d8559e7522b9fae7
38b0457666376fed9d8f8cf5cfac8ae5d76dc3e4
25460 F20101204_AADCCS wilson_c_Page_093.QC.jpg
f3e0f040057cdbc2e0c2dc6baa6daa52
8d5943c5ef248b16e4c6813302f4d843552d7a33
1322 F20101204_AADBXM wilson_c_Page_147.txt
4bd16531566845805da99773136d25a3
c9c995dc05d220ab3d710641029d33d9d4eea6c4
47858 F20101204_AADBWY wilson_c_Page_120.pro
95a85ba5654af9851d295dd43733a772
eb6e4b13dcfbcf8ba2a0adcda079b4da8251925c
473 F20101204_AADCDH wilson_c_Page_169.txt
dc5547f65cf38bdc203dc3f4e9df2b91
a45b7555b45194410ba7d232014bc2c7d8346c43
5070 F20101204_AADBYB wilson_c_Page_079thm.jpg
013570f289dd63175ccda54239b55754
6e7f057a140b6ba3197b8402e214610b1341fdc2
3279 F20101204_AADDGK wilson_c_Page_128thm.jpg
5cf6f59260fc33957a856ed6c5331901
b7b1d2cdc8be37fa5e5bcc5fd1cb545666a673fc
15885 F20101204_AADDFV wilson_c_Page_106.QC.jpg
fced4391b935eca5cc02e30039dcc7e3
0daada99434fef2f3fa85a77064699f9e28bb2a5
13703 F20101204_AADCCT wilson_c_Page_030.QC.jpg
baa9073781d515fb673a6894498de811
430d111815c8af4e8c164d526018b7e4d2ac7eff
F20101204_AADBXN wilson_c_Page_109.tif
987358800b0695f92acb078de9ae8ef9
9a66a26a74fbe750a6847009f06137747e01367f
1286 F20101204_AADBWZ wilson_c_Page_096.txt
c7e265429242a9a897bb5e7a8c17f63d
f793fa160b5bf10dd1b60d9e9ebd34b9acb98ea9
25276 F20101204_AADCDI wilson_c_Page_146.QC.jpg
b7390c3295a867387c81be912034b243
887232137722ec7773020b656725fd87820ee23e
83159 F20101204_AADBYC wilson_c_Page_018.jp2
a98560c456277bf5c76c9baeabf87ad5
591273bb97a067f40c84b28c32d10562fe5be312
3171 F20101204_AADDGL wilson_c_Page_129thm.jpg
4108b30cd09543363837bd2a3adeab2b
80d1cc7ca5d958dcf1f40ef3452f5617660e1d1c
25791 F20101204_AADDFW wilson_c_Page_107.QC.jpg
e8d210eca3236e1319ffc5858b6f97d8
12d2a92a600dce98673c53aa985c1c7c5ac30cde
6862 F20101204_AADCCU wilson_c_Page_105.pro
0f37e9c18edd46c7da2f0c0d858f035a
4d161023194f45ff69c6edd84fde968e443dfdd0
1680 F20101204_AADBXO wilson_c_Page_133.txt
f5c147e1dbc8cf99c01f92847eeae0d6
90cc8fb0a65add882ec4c61adf1d56a90e9de894
F20101204_AADCDJ wilson_c_Page_006.tif
1d251878f36b31b31a63257b6808bdd9
dd0d96e32f31f3dda988de1b3b3f094aec5a96cd
3582 F20101204_AADDHA wilson_c_Page_149thm.jpg
1071ff5dc6ef5cfa8ad7eafe681854c5
4e5ea52d66bce66eeee3f2ea956e3710d4a92270
18280 F20101204_AADDGM wilson_c_Page_130.QC.jpg
a8e12a4c804e0d3ec5dc42d5e526da73
8565a010c2eefcc4d1e89949496d29e8b026c222
16480 F20101204_AADDFX wilson_c_Page_109.QC.jpg
c1e8e4bd05811395bdc5aa1d94ef691b
34172657e3e89a796b1841fef4d786b173b7ba93
1798 F20101204_AADCCV wilson_c_Page_082.txt
fb4b17605d739c994183515874e23c81
b099cdfd37c515ea2626e8d53f13c5afd9078cc3
F20101204_AADBXP wilson_c_Page_187.tif
48cbf22ac588bcb8585d6ca7f8bcba62
c482dded7d89458021e6cc32ae34339dc3bcb6c6
67109 F20101204_AADCDK wilson_c_Page_014.pro
8a2d6a0a740ec0d28d830fff42bbe47f
8b5a368264b9eba7150665f7c13995422d6d0b09
6074 F20101204_AADBYD wilson_c_Page_074thm.jpg
44035e4e7aad03f10799c312b2ca7e7c
4ab49eee8e72c6f103020e9248d3034b3bcaafdf
21621 F20101204_AADDHB wilson_c_Page_150.QC.jpg
7d99f61a1d1be8c8f98c5198f0cbc91e
59e64da59e0aba24600bebc9ba515c2ec4d7afeb
22097 F20101204_AADDGN wilson_c_Page_131.QC.jpg
58c19825c050c0c5ca5d59c433eab584
50de4f379c8f07bdfacc30ef87be406a0f8beb90
4637 F20101204_AADDFY wilson_c_Page_111thm.jpg
c315382e8bfc11d9278e7ab0f1bda71d
0857efc7b7f658ad88a076f0db1ee9c734b85626
43592 F20101204_AADCCW wilson_c_Page_191.jp2
6aa291acccf14c01da88caea04807f59
0ca058188ad56cc8a0f4ffea15ad7ca4b1a7df7d
44639 F20101204_AADBXQ wilson_c_Page_106.jpg
f93a938dd655e57ce97e736f8fe76b05
d94e693b676485882677513c14bb1b5cb188420e
11817 F20101204_AADCDL wilson_c_Page_162.QC.jpg
952e97bac290d58d95156cee88103771
4f6f358f7025002322b56dac70e85f3db1266289
F20101204_AADBYE wilson_c_Page_161.tif
541e8715e605e1b2d951241a972eed1f
c6d56f1a2f3074bdc22ab116d85d7fc6c42ac5c0
6376 F20101204_AADDHC wilson_c_Page_151thm.jpg
aa1f14f3d05bcc9b3c114e9fef4e3d2e
ab31eca9d6bd93c8e0ef32db14576b704def0a39
22157 F20101204_AADDGO wilson_c_Page_133.QC.jpg
b8b80251b86cd5573c381722c4399475
7700e9a0296494bced0604c9ba6599d475dc6aa3
13757 F20101204_AADDFZ wilson_c_Page_112.QC.jpg
97d026616e322b776832d38d54101d5c
54cfce1a773ab15604c7344abde072f0baa8d728
74610 F20101204_AADCCX wilson_c_Page_116.jpg
9677e0027db1ec02eb1b9ce2c0fd05bf
51e1a018431423915ba694421c5b8eb1ad5be30c
28338 F20101204_AADBXR wilson_c_Page_121.pro
f801a8a426254b17fa125843062f926d
8717abf3c1c3615cc59bbc0d89712df10e767219
4077 F20101204_AADCEA wilson_c_Page_109thm.jpg
bb016bc9d631f1cb984cffef915268b1
e3e49e681b8a72a4758410f6981c4baf7b104f4e
63793 F20101204_AADCDM wilson_c_Page_035.jp2
208bbc7f43ceeba25837b76715148f26
bcc87869666bb974b74dc317cda0267ef9686e77
63856 F20101204_AADBYF wilson_c_Page_141.jpg
d630c9efa90e804e91ab400a10727b2f
da179da6b7b46c680398f5373e99a842e73cac34
5993 F20101204_AADDHD wilson_c_Page_152thm.jpg
6156d9fddd73d4532ad3ed5907821a61
477fdb1ce6569d9237a76bc5a8a5830972d62fc9
66147 F20101204_AADCCY wilson_c_Page_158.jpg
532c3b168233243c5c8dadcaa81eaead
c6e87105019d9cf54382a1a0e3f21b2faabe5be7
62376 F20101204_AADBXS wilson_c_Page_038.jp2
2be9ce8105ec88fe37268c75bf1e7fa5
86796ddb1f3298d456e326d92434e799ac4bc533
F20101204_AADCEB wilson_c_Page_153.tif
0d0707aa78ac501ce8973f905d339df4
9ae46dde45c38d6e443749653980d7157b8b2d25
15196 F20101204_AADBYG wilson_c_Page_148.QC.jpg
e907d88088230844def27594a802c19d
5e5661823dfcf55e34ff4ab3883d79257409410b
11974 F20101204_AADDHE wilson_c_Page_153.QC.jpg
df5ae4db5d471381c626e0c6766b63a0
112bdb080cb00fde524d72f261853619c350c431
6185 F20101204_AADDGP wilson_c_Page_134thm.jpg
3dfe5b8d2c2858231e1c317a91fd4906
b8e2ab25f4b13cbc9a7af5a76bb7e691f82f9ad3
1051967 F20101204_AADBXT wilson_c_Page_015.jp2
df994007378699af9d73b00e90fc4b29
7a322ffd9accffae19fd1efe2de77044a604d113
1000605 F20101204_AADCEC wilson_c_Page_059.jp2
71a6fa1e35ad3f4829e949f207ca410b
2780b9b3bc4c786e2740352bc433a9e089b0b5d2
21549 F20101204_AADCDN wilson_c_Page_155.QC.jpg
28fe5c7216fe1d586ed43f152b22c121
0103b113ed22e613578c6122944b81b55521cd10
47303 F20101204_AADBYH wilson_c_Page_148.jpg
5cf025aef02e4c5f41494b2ec5f8a27e
2bcf5f269249b39ad4edfc127a17b8a43256b7f9
6311 F20101204_AADCCZ wilson_c_Page_172thm.jpg
884d91c5d58eb53bd26e21f3801ab599
6ccf87a4214f44332c9af73c35be62f380fb0f6a
15159 F20101204_AADDHF wilson_c_Page_154.QC.jpg
a4bb0611de5615b4fd47f5f4e02de919
06926f2da7df26422ce59014150438b675983b80
23869 F20101204_AADDGQ wilson_c_Page_137.QC.jpg
54191c72a85b7fea7297dc5897a0ded2
9d640ed50a16307a3ac8607266c192a46741ad6f
F20101204_AADBXU wilson_c_Page_149.tif
31b29a064190e2e9eea1613bf8d3ff15
be3da0a6047d1efd5523d6cea62d5f12e57c1cdc
5279 F20101204_AADCED wilson_c_Page_110thm.jpg
011375a2938493f88d8b83390b99a34f
6b39ee9694b77a64242f7649e891286aa4058a9f
882267 F20101204_AADCDO wilson_c_Page_009.jp2
82b9b6707729eaa7e540186e1dc68fa5
488ad59b2166d0eaee6e6676cdb54cb80e56d9ab
25205 F20101204_AADBYI wilson_c_Page_055.QC.jpg
0195aee25019a80be10a452c27ad45c1
fd027d7f005deb3dc79f798814c1b4054d7d4377
4634 F20101204_AADDHG wilson_c_Page_154thm.jpg
9cd2c831b700f58739f2dac1c682d6c0
efbf997b7db03f40edfeabf8fb8b3fea33e8ea77
5852 F20101204_AADDGR wilson_c_Page_137thm.jpg
31385ea0ed677382b35434f475309d0c
d533a0a0ee3eb34a30b121d37e1cac3e29fd710b
120482 F20101204_AADBXV wilson_c_Page_016.jpg
4dfdb722698ea47aa51030502cce49ef
bb65d779262a73b12ed96f8b124b4f991a615298
3528 F20101204_AADCEE wilson_c_Page_050thm.jpg
fce3c2a3c472afb24354f27616dd1ed7
60856b27dcaf05f94ddf31a501131636b01e4edb
F20101204_AADCDP wilson_c_Page_125.tif
5c1e35ecea494b1791be2210889565bf
d7b50d55a91936efdbc1391ff0f1a6ef10139ab3
27721 F20101204_AADBYJ wilson_c_Page_147.pro
ccb24d3e28e166fad59f119e7f636cb9
cc292aa88929f9e1534015508178f3cd0bb4bb1d
19476 F20101204_AADDHH wilson_c_Page_157.QC.jpg
316671d80e24983d268efee9b1a2ae1d
100cefa97b39cc744b1ec33ddeccbe172ecceec9
22943 F20101204_AADDGS wilson_c_Page_138.QC.jpg
043e82d0e88cea351073ba497ac4344a
0dbe027380de66e316dd8ab3cb487695847376c4
21380 F20101204_AADBXW wilson_c_Page_043.pro
d285573f0add07307fe517db3589de13
dc2b375321acb3073fd7bc8e16f044289797a6e5
F20101204_AADCEF wilson_c_Page_123.tif
b3d186c6e9c94bdd641af89d80b477ce
17e3029c9a7354933e0a251db05d1d127ef89769
F20101204_AADCDQ wilson_c_Page_092.tif
c01abe849a46ea7a0df267d27681ac64
deb4193f4b4dcef6c5536399824c664c6e4fed64
1594 F20101204_AADBYK wilson_c_Page_078.txt
9ea8c581cf807030f475a4c9f712606c
2dd253e85737041fa12a2b1eba93deb9e01ed53c
4901 F20101204_AADDHI wilson_c_Page_159thm.jpg
4b96080df73f0380865e12e631ab61a4
ba3357dae7447ad90ab7373c80ed4bbfaf6d14ee
25224 F20101204_AADDGT wilson_c_Page_139.QC.jpg
6e9a85fd675c182e0a3762339660b66c
121dabfd36e8e6ffdd48ad478bf87facf8801739
23526 F20101204_AADBXX wilson_c_Page_076.QC.jpg
1cd33f78d1002741330eff4162cb692c
d12783891f497ec4ffa6a0d637683621dffcf34d
61425 F20101204_AADCEG wilson_c_Page_115.jpg
d4787138e9ab5b93853f1f0cd3c33379
70f879859fbe06ea519dcd202ddf086561fdb617
6033 F20101204_AADBZA wilson_c_Page_022thm.jpg
fbc8b595fdc9fab9cc084edf9db18282
1ebfa1e9278df36e6108398ddb3ecd3a93bb081f
25309 F20101204_AADCDR wilson_c_Page_125.QC.jpg
cd1fe5dce7737f19a2980abf2253a68f
34d7701f6a72306d6125bd4ee6134943c49253e3
65831 F20101204_AADBYL wilson_c_Page_070.jpg
ae2c466f636a9a5b17c79a1a9063c940
00b96626692c3ae49221330ec56e78b1fa328f60
6098 F20101204_AADDHJ wilson_c_Page_160thm.jpg
dbc26d02ed05eb1d6fd312775fc59628
7536006612e1e0434eb4408d3a4072e26656525e
3596 F20101204_AADDGU wilson_c_Page_140thm.jpg
590da295d5482e9331dde6732b5136a7
25d1b4977e4d9e6d452503026003f7c5f7045900
10662 F20101204_AADBXY wilson_c_Page_123.QC.jpg
cb2ad3c72acfdb826ab24e4b6215b1e3
4531e7a9982073a92eda7f93e6d39edbadaab8f5
F20101204_AADCEH wilson_c_Page_030.tif
b890d2906e668dc63a5bbf42012f1b8b
5766d195dcff09b912ec7909fa5da0d06c3d5af4
F20101204_AADBZB wilson_c_Page_151.tif
9cdae01e6b24754797acf834a1f554fc
4ebbed1425391d06292f75fc8fb7b2f39320163c
F20101204_AADCDS wilson_c_Page_040.tif
9649e43facfcced319bf0e53227c6daf
9d1700e988c9109767fe6fd68ddd7c3fef2f64bc
969123 F20101204_AADBYM wilson_c_Page_176.jp2
2d65b55353f9c356750acbc3538ab94d
698e325442146bb19f153ce933ca60e594254474
12575 F20101204_AADDHK wilson_c_Page_161.QC.jpg
0e9b2e6974df9cf38d2aafbba6c1b192
e2aa450ed6df6cce7180a2c4b0d3200e3577b46a
20003 F20101204_AADDGV wilson_c_Page_141.QC.jpg
7622c31a40ed16dded9f58c22284e1b2
7bdd55dfaa7e17ab88b264461e3247409dc150ef
2025 F20101204_AADBXZ wilson_c_Page_187.txt
8d8a9e9a0a3d85a83a5506062bdfe5a5
b2678f10e0fd81ea7b4acf2157e1a5bf9fd2f209
21142 F20101204_AADCEI wilson_c_Page_023.QC.jpg
755520367775581358d3c1b3bbf3dd12
95a8535996b95ab3815f5d479d7c66678c031fe9
6369 F20101204_AADBZC wilson_c_Page_093thm.jpg
d312e675a04a211a5ea2a19693ce0986
9b60fd5ceec5d92780ff317ad64bbfba17c76fb4
95612 F20101204_AADCDT wilson_c_Page_039.jp2
7a3b0294f024a4be216250831f139535
6fac4806a6c3f4a17cde0158e9cac6d417528909
F20101204_AADBYN wilson_c_Page_054.tif
6088c3a5d568abc7ed7e82eefbe323da
7c9dea8c98b52c7343cc5526d3512a3d9557702f
5717 F20101204_AADDHL wilson_c_Page_163thm.jpg
992a26a342dffce67c97fcf9671a1e58
3cd10d1cd837b21468a8f13583f8220d08725b8e
15945 F20101204_AADDGW wilson_c_Page_142.QC.jpg
987053f6cc7a45c71139463553b9a86f
2ee7ccec742effdd2bf034da6c74a9200f0d4d80
24755 F20101204_AADCEJ wilson_c_Page_103.QC.jpg
126fc644b0508b8379a95c813a5e406c
a6be6a795b67e2ce447895fdc6b9debe9c849daf
F20101204_AADBZD wilson_c_Page_157.pro
1fc37aee89f460e5858fe935391d3216
b1d99d7ee2206a73b0840b5e6715541d25d982db
5563 F20101204_AADCDU wilson_c_Page_168thm.jpg
affaaa6b7ef145c11b86311fbaa4b23c
8cb14fd736fb0f1b6a0b5f157412fc76afb47570
21116 F20101204_AADBYO wilson_c_Page_061.QC.jpg
4f0581ca575cc293ef783c467efaf92e
b3d964510515f7285acfe3027e4f0b1bc885e318
25996 F20101204_AADDIA wilson_c_Page_179.QC.jpg
a65e66ed5abe8cc056c955d77d127e75
af80877b14725f939f506e367589130399bb2def
15387 F20101204_AADDHM wilson_c_Page_164.QC.jpg
3410aa06ac3bf3535ef98fd03e61d69d
d38f25de32e7b7032c2728bc6bcc7ca3cd98e84a
3451 F20101204_AADDGX wilson_c_Page_143thm.jpg
e5bb1d626c8ec539f52f9f96cc2e616b
26578d2de99e56dc53e1a1c3b331231101ddea3a
17377 F20101204_AADCEK wilson_c_Page_115.QC.jpg
b07f80bba038e1b126895a7a1b796c31
dd50f0860be21d071245c6c1b8236e420f553c05
28598 F20101204_AADCDV wilson_c_Page_038.pro
bad616d258cc59f320bdeb3a8f3b51a8
8bfd9ef8dda7f6d4ba4c58e42dac77f1d7020d21
385830 F20101204_AADBYP wilson_c_Page_050.jp2
c3a0be0c90f926a230c086c6efa525d3
34f1c60a30bdddf5636672f78d162ce13e5118fa
20036 F20101204_AADDIB wilson_c_Page_182.QC.jpg
da5de37273c79f501dbbd2b64207c962
d7555ecebcfbb3f9aaa6de46b9fdd5746eabf2d7
F20101204_AADDHN wilson_c_Page_164thm.jpg
d9c633b54acbee205bdabd5e7f30abbf
03db2a2989c4e4854cd95d8dc0301ed976c0dcfa
16576 F20101204_AADDGY wilson_c_Page_144.QC.jpg
bec662d7837ed0c2a50c2f3747da198d
8a29659aa063e4bc7e6b432e35cd74d938fe65e4
21178 F20101204_AADCEL wilson_c_Page_044.QC.jpg
d6b1efbd9081ba932304f5074355c9dc
288fb64e35121ec9c52bd8deb2600648c56baeca
27708 F20101204_AADBZE wilson_c_Page_180.QC.jpg
37276255399ed53ffe68a20c9f677435
49e50e406c1dfa8b0d04db3f2e3a7f1edd31bdaa
55605 F20101204_AADCDW wilson_c_Page_176.jpg
6be51b8003969ec03a93984427f992d0
bd47c59b76dea51b0866c4d0d97d7ce97a681645
14639 F20101204_AADBYQ wilson_c_Page_053.QC.jpg
16d237eb5a7fe6541f07293ea9a0e2bb
a847ee9b01184b1ac799cfe4d6010923676eac96
5741 F20101204_AADDIC wilson_c_Page_182thm.jpg
b1b87379bd7f3fc7f7ef6b759b2f1a01
5bfb7292a0fe2ec03446d5f138099b60af745b3c
5600 F20101204_AADDHO wilson_c_Page_165thm.jpg
98af3a4c29d7f6fb0788ceb0d90cd911
8fd2c0af4201c924770b7930a6bc7d5996682f46
5469 F20101204_AADDGZ wilson_c_Page_145thm.jpg
6a4f118514c9af71dec4975e7cb8ce46
f9b4de6853f1cf5e1f5d20ea3b4ec24e260a6408
1104 F20101204_AADCEM wilson_c_Page_192.txt
9c0b93d2bb9bb79c7d5be34f06ca5d5c
7cdde51993157c58032839ffb1360ce8b491541d
1051872 F20101204_AADBZF wilson_c_Page_168.jp2
fcc42d9973ad61076ee6cdb85508b32a
7d449c85a6d9d3bf7e521014945716cc43c9b5f0
29784 F20101204_AADCDX wilson_c_Page_140.jpg
db4e293650272c5b8e98ef0f23878c68
99bc439327a4b10f08b01b5be67f81257fdb631c
25319 F20101204_AADBYR wilson_c_Page_142.pro
031c8d72c3b5e13304d78a8278493cf1
e7c39f7f6916b36b8c5ec2ed27bf0578c6de3804
2383 F20101204_AADCFA wilson_c_Page_190.txt
54690059e53715a554d1cfeafddddecf
97421f686ced046b9de299dc779bd97c4d54f59a
26523 F20101204_AADDID wilson_c_Page_183.QC.jpg
96b9267c51f68d067ad6ab4811b0019a
b3ab25d89363957296330b3aef9c0dbb1030219c
14440 F20101204_AADDHP wilson_c_Page_167.QC.jpg
785b73b9a8a74e07186a4216c25d0e53
ffd4d95660579504358e86756d2eb2d800c86be9
20760 F20101204_AADCEN wilson_c_Page_037.QC.jpg
6ca8747d807429848c71ee3b310c6b98
cc9272944754c923bc95e8652ed20df9b0a9d7f7
919 F20101204_AADBZG wilson_c_Page_129.txt
e00c0c6ec00a585238382bfd60db932b
8e0deec0f59eb5f3162d029d53c21651d3c63d40
3831 F20101204_AADCDY wilson_c_Page_100thm.jpg
faa8e2d20fcffa4c652a20461db53e19
e246dcec076ee17d8d15b492232aaa41925e9673
105972 F20101204_AADBYS wilson_c_Page_187.jp2
0807b9eba89b55fa58cc1ca1cc4c79c4
a33691cc12b8d1fc6b9b2732e0b47b04a66c683c
48793 F20101204_AADCFB wilson_c_Page_124.pro
eda14f1424a8090423fa0f97c1d3ff2f
719f96c2b65d0d192716965e56cec3289eeb2251
24897 F20101204_AADDIE wilson_c_Page_184.QC.jpg
150646573d534ee956082c7c89280871
aec6aef6d7d585731d180f7626f59509429f0963
1491 F20101204_AADBZH wilson_c_Page_042.txt
af5f1913adbf12d6e72371244bc7898a
85b0de503b3749b89b431927d81f764dbe3968a0
37983 F20101204_AADCDZ wilson_c_Page_117.pro
aeebeedc1f86d102eaeeeb9b0f6f683c
114b72f3ac622f9200cfe30dc7a744c65eea5325
20835 F20101204_AADBYT wilson_c_Page_154.pro
901bfb12d35f4cc0f9ec788dc8349634
327aad34d40570001c11d99cb3eef2deef5db2e6
44664 F20101204_AADCFC wilson_c_Page_028.pro
3d91ca2262ca811fa52b9dfd5ecc54b4
765241f41e161e589e529a33a74b15fe597ba3e5
6147 F20101204_AADDIF wilson_c_Page_184thm.jpg
7d5c169c41d587f695e5e669032c68d0
8ac5552908387bf119f321c4aff99c035615d3e7
4297 F20101204_AADDHQ wilson_c_Page_167thm.jpg
33e2bcee29a3736d5949643f0067a539
e25f7961d4b72dbb6cb57361db1ec2e204311803
5821 F20101204_AADCEO wilson_c_Page_097thm.jpg
bf5f7dcf140535ef4dd2a5dee26ed776
b2f83e4d07eb655e2b3c208723057371d6fc136d
F20101204_AADBZI wilson_c_Page_074.tif
d264f76a67fbd44e594947530dd6f53a
3867b00a0c294de007ec99e1aab318584078c5df
22921 F20101204_AADBYU wilson_c_Page_181.QC.jpg
c9bdece3c762182de903f6fe0fc40e02
30ee026fe8833ce3fbae7ad5441d451045701701
75733 F20101204_AADCFD wilson_c_Page_166.jpg
188d468c17644d4eefd63f36416f2e12
08f3dacb3d48e8ec28130272ae3b3fd6bd9f789d
5749 F20101204_AADDIG wilson_c_Page_187thm.jpg
15b7f0bba6d954ebb80241010530007e
757f418cb6f658ecb90910f189d25c6da8dfc574
18028 F20101204_AADDHR wilson_c_Page_168.QC.jpg
31c3ec1833d87e7be9576bab25c719b8
00b7ffe73645ff8099a66c3ad870cd1991953896
6007 F20101204_AADCEP wilson_c_Page_055thm.jpg
0729993368500712493c8131765113ab
dde1f7ce54122d53760c0a0fa4fe299dfb2c6f87
83694 F20101204_AADBZJ wilson_c_Page_041.jp2
072b6f3765bf003411fdd2d3971d0f8a
9171b2f6aba3cc806e2337989839503cc63c5b5a
88715 F20101204_AADBYV wilson_c_Page_037.jp2
2753fa214b7999c47a99fcc4b2815276
0694e79e764d144b34fb68ab21353c7b35e7b637
4960 F20101204_AADCFE wilson_c_Page_092thm.jpg
b0d6f37b2480d7bf3e99c6438d0c7460
b928238aa642957b32377855aec0812e88fdc0ab
F20101204_AADDIH wilson_c_Page_190thm.jpg
ae8d4092fa580e3f7f50595150344ed8
bf0569c69077f984beba563c3698ee8100b03374
15680 F20101204_AADDHS wilson_c_Page_170.QC.jpg
6c38822c632ec832bb6bd972c8e9bebf
4c4edf05911a31db6bbc575dce8175565eed446e
F20101204_AADCEQ wilson_c_Page_137.tif
484531cdafb36a360c4d0fc6a9a4305d
f4ab3a08d82b1f1634ed461b1fc3cc2b66d3c191
42728 F20101204_AADBZK wilson_c_Page_132.pro
e8d53708f365c4966097f00179fa176d
4a999a4369ed6fcf5404625e5d177c47252d1280
506 F20101204_AADBYW wilson_c_Page_069.txt
1a79af513b3bafd67dd7e09cd999026e
472356bb8eab3c1a510aaf34b229209d7a9a93d5
2028 F20101204_AADCFF wilson_c_Page_179.txt
4274efb636dc873a023dfc2bfa6da5a5
e01e357189b4df66deb4dc777f06974c02fe4696
9708 F20101204_AADDII wilson_c_Page_191.QC.jpg
4c32254a6e20acb598943ac5e3d26032
22a7a6f52b854069959a1fa4b4e697d7b2a9072b
6126 F20101204_AADDHT wilson_c_Page_171thm.jpg
a114c8cdc6c03ae9559e9a98658f192a
83e7b749e9a7c9d3210e547ab916dd049bf386fc
1757 F20101204_AADCER wilson_c_Page_099.txt
950e464c758fc111abfe81069e487e28
858fff384d47b32dddcb9cb18585ff93f4a1cf89
102277 F20101204_AADBZL wilson_c_Page_060.jp2
ba17d70e33193aeb11fbda4871c12ecd
3b1c2d5a6658411a64c34f1e39087bf4fe9c87be
7238 F20101204_AADBYX wilson_c_Page_180thm.jpg
fcde5e99d27667a3b7639575c82c2b18
eccad04fb5b64b213432dbca6c0dc2580faf3324
99851 F20101204_AADCFG wilson_c_Page_084.jp2
4cbe18b6869d337be8bb7449e939305f
44e84bfa980cc7189d8cfaea5c6be2742e9c0054
5173 F20101204_AADDHU wilson_c_Page_173thm.jpg
57e79c87d79a451b0f3b82f1f513e29d
38910ae588ee4585c497a1d96f9da66ed4bf32f3
1907 F20101204_AADCES wilson_c_Page_102.txt
3eb4e65d6c1bbd064f5f089dd7069a71
795c6d815ae6c7d46bc276bf6d5a04739c412d59
F20101204_AADBZM wilson_c_Page_001.tif
ce9e0ce6761db4658c447923782c4d35
5ffe003a0e290691e130409aaa1193c11400c701
17645 F20101204_AADBYY wilson_c_Page_129.pro
c334022bf1d1574edb118a65f5c4fdb2
733e951f44bc8076857807b33dab2f4c0cf1592f
714164 F20101204_AADCFH wilson_c_Page_147.jp2
37a5ed7bfd6a49cef7f3c9b71421c2ad
93d162f6d659d1daefb7f613401da91895e7d4f7
27531 F20101204_AADDHV wilson_c_Page_174.QC.jpg
94ff19eab5d68f824f93067ba89de6fb
9f181ae9d34333b54a3795bd3964262229cc3f5c
26715 F20101204_AADCET wilson_c_Page_185.QC.jpg
dbc99da658d771bca4d95d4591df6557
9623af62b33a6cc0c7c50299f3d518af8a9e022a
F20101204_AADBZN wilson_c_Page_016.tif
b23fc089fbbeca56b4036ad7f5e8c5ee
d254de03d32fae5a7d427f00ffef47004ff697e6
44672 F20101204_AADBYZ wilson_c_Page_032.pro
c084d0784c6d97a39300bb6eb9580025
d1477d5f70c49002329e0a6eb1a8be8f501cbc62
5494 F20101204_AADCFI wilson_c_Page_106thm.jpg
8e12b71a14ed96d70acae3a7840aab15
5ce13e0481eaf75bf9efec3a36c7b749978c158a
19105 F20101204_AADDHW wilson_c_Page_175.QC.jpg
d8d33732d0d9a290a18dac29a815cec1
a5642e44a529f0b9be1b374721bd564cb0246a03
478597 F20101204_AADCEU wilson_c_Page_161.jp2
9e44e568785d252b51ba69b7894c08b7
85302674838fdb39b3d51110bd34edaab4c17568
1786 F20101204_AADBZO wilson_c_Page_122.txt
9fd2f8f3eab9c7d59a9b3568f8b45064
0cc06f094d2c28a9a006ca4b3873815f0df3906a
76926 F20101204_AADCFJ wilson_c_Page_054.jpg
3e7a5dfd8fb5f106b8bd25a5067f5ce9
cac5c98d925f927303e7d6d26a5bcb2fe8759b3d
16795 F20101204_AADDHX wilson_c_Page_176.QC.jpg
7f06f50460e75eb3bf15b18573ed3776
f4373215ffcfb21244840c060c9f392b1a14c0de
20240 F20101204_AADCEV wilson_c_Page_089.pro
47bec701e7a7a267ee5b6342a33b8839
cf961db3fa18b09cd3eadffac5fe58901a22a515
23723 F20101204_AADBZP wilson_c_Page_082.QC.jpg
b59afbda2e4176e33d0b3fb5abf36127
2b19ce2304507057951de19b79ea2f8e182961e5
20818 F20101204_AADCFK wilson_c_Page_108.QC.jpg
59a315db97d23da3edad8221db34978c
8e5d94b95fc6e5c73d1c5aa2d03678faeb278dd1
4509 F20101204_AADDHY wilson_c_Page_176thm.jpg
f850ab726d8936725865579d068bd441
dcfa1727221522eff156be51e44507f7ff2df12d
75727 F20101204_AADCEW wilson_c_Page_032.jpg
84e654718c92531e2ae3c79ebea1f682
9618bd4229a9d4e77048f1f46770aff8aba96cf0
376249 F20101204_AADBZQ wilson_c_Page_056.jp2
bff852ed2ea938737e0eddcafec4739f
f773d1449114760270eb6200ac65a5a9f48aefab
79560 F20101204_AADCFL wilson_c_Page_022.jpg
bad72022a76eddf8f8b4d00577e508f3
9538ac96ec93d73cfbfd92453351b6b0dfd3aa1b
18136 F20101204_AADDHZ wilson_c_Page_178.QC.jpg
b988c5ceb911b76fea8c9970de6f710a
92ffa0976618418c4e08de40ec874d4913f87ba1
F20101204_AADCEX wilson_c_Page_176.tif
02901f5a70f7bbf192b5f5263b792d59
01e396faed4faacfa5cbf34d7a565e59acc1016c
48513 F20101204_AADBZR wilson_c_Page_184.pro
03e882c386565ef7e7b1fa15269bd886
62c749cb0334e72be9d12d380142a87be1712672
49457 F20101204_AADCGA wilson_c_Page_093.pro
30b013e6402e483d56670b298c1c88ab
dac368989b273aa05ee5ce0d30db4c2b6439a1a7
F20101204_AADCFM wilson_c_Page_148.tif
bd9ca20eb99a8a0f36f9ae1f74553b2b
cf4bf58b215ce87d888224bb70e0854f3f97ad69
5418 F20101204_AADCEY wilson_c_Page_132thm.jpg
e9249f9772729a23239356c5ae00da2d
bf1ce63d7dabbbf3ab82fbc6e5ae5f23d6cdaecb
5801 F20101204_AADBZS wilson_c_Page_033thm.jpg
b26bb2c0e862b9f2ea715548e5487039
c6bd56cdd682dd1c1a560b597bcd14182262b79d
4368 F20101204_AADCGB wilson_c_Page_115thm.jpg
d5115c8268c728c5b306165f52ae2263
1003cf7a7ca5575fbd6fc9a3a03594d65a1aa2a3
10112 F20101204_AADCFN wilson_c_Page_162.pro
589c7d87c0705ad47a79c69292618100
62e05ab099ef2ecf8938eae1f550013a6584667b
1571 F20101204_AADCEZ wilson_c_Page_002.QC.jpg
fd9b2d6f24f8cdcb146bdd9f9ba5069b
4c54ce4320275cd48e70a3b4938380efd034eb17
80124 F20101204_AADBZT wilson_c_Page_019.jpg
28d5c053be91f7da75bf7a6928ccd2fb
5c3d38c7d8bbaedd64ed1113c8265722ef9643a4
F20101204_AADCGC wilson_c_Page_027thm.jpg
2f6af8f8a9fd2b330e807256dabe91b8
a34eef6cdc8b2cd6fe5a7d6f0cff803fd687d837
1949 F20101204_AADCFO wilson_c_Page_093.txt
ff1cc8d3eea0cd787b00e1006b2dece3
1547b5d3ac38b9d9c4340275d392e8449667fc0f
F20101204_AADBZU wilson_c_Page_122.tif
4d8eca2100c4c7fdb81fa1596d1fa0f1
11c72bc0bbcfc472382fcdb6641ec349bdffdaf6
662 F20101204_AADCGD wilson_c_Page_064.txt
92cbb6f4c54bd065466bcf3443759e30
081faec54bb622f0888d5276d3fde1cb70ca468b
F20101204_AADBZV wilson_c_Page_115.jp2
6ac5fc64bddb43762eca05fb5cec7144
b3f2f4052bc6b3ad3ff68949006b33b0a563aaca
6321 F20101204_AADCGE wilson_c_Page_179thm.jpg
082f09e1b10f887dc9a211851bea651f
d9ce5078f299db06fc773fd867630f8804f7617a
1962 F20101204_AADCFP wilson_c_Page_048.txt
3f72641ea5bf0ecf8ce57ec7d4dba585
aea298f4a1be2ac7e990cafeeb1dc85476fd6b56
25414 F20101204_AADBZW wilson_c_Page_074.QC.jpg
81943ce3529e28169e51f7925f79a51e
2bdd2d1def88b1286c94becc3c1ff66bb688ba0c
428129 F20101204_AADCGF wilson_c_Page_128.jp2
7ed0fa8ce1d67a47fe7bc7c6503cd9e0
0156b5f5388b6c685203d1370ea83faba1f170eb
5384 F20101204_AADCFQ wilson_c_Page_006thm.jpg
eb70f9cf9dc1baf45de0380267527f64
668b69334037a9fac790baf31b2464bc545ed654
71947 F20101204_AADBZX wilson_c_Page_090.jpg
1b05e02632265f31b866b7cd1d786fe8
aeda8c2f1bda86b39beb2e27c4af592e7271972d
94465 F20101204_AADCGG wilson_c_Page_090.jp2
6bda0ab9df2784eaf6875a8386baf28d
6d0d586189b9e605cd11d0db35b365b2c70d75ef
101659 F20101204_AADCFR wilson_c_Page_033.jp2
5d67bd422bd4b21dbb77342d106644ea
92d9428ba2d24d1e5782f41d840a173c246f4353
66914 F20101204_AADBZY wilson_c_Page_018.jpg
2cb14efb15a0de61ee6dafb4581b3489
198acacdb8ed21adaaca0a4a821cfc8dca73a688
4815 F20101204_AADCGH wilson_c_Page_126thm.jpg
1193598370229622ec9d86bf02ab69c4
32233ad22cd39f7f571f8dba8a4df06d79d5f6ed
1713 F20101204_AADCFS wilson_c_Page_131.txt
aaba62651f56bf79b6a2500ba8cddfce
7e15295f8f49495f02acd12602ce4aed67313db8
4760 F20101204_AADBZZ wilson_c_Page_002.jpg
c4c33e6b4be0a78c21014abcf01a5302
51cf5ac875a3d8ae86779493771d61131023c26d
52633 F20101204_AADCGI wilson_c_Page_049.pro
98f0f0be7771edb4fc8c51cda1524442
eeb63b16791b4e303e153ce184bcf4169ccf7762
111377 F20101204_AADCFT wilson_c_Page_174.jp2
e84adbce5a6e7aae65da65554e964289
5cdde67f9e13aa04c73ddc667076ab10bef3f053
1850 F20101204_AADCGJ wilson_c_Page_021.txt
26891ba101a004346e16f5a484c62347
6bcb66bf03ebd5926e054b291f534667e637152e
52064 F20101204_AADCFU wilson_c_Page_167.jpg
2403446f59c3a3ce76885aa769b572a8
5ec891ea67abf41d109beed1be2898893b8fe027
1327 F20101204_AADCGK wilson_c_Page_038.txt
9e617c378aea2f7424a8b3dd2a72dd3d
955c38e4765ab1ecfe05577d22966da846e0cd10
701491 F20101204_AADCFV wilson_c_Page_135.jp2
4afa18027ae8ddd5b60ac2ef1010197a
8c7a2b409750ed387f15a880df0ae9372b4c2994
1838 F20101204_AADCGL wilson_c_Page_137.txt
c933e36ef5a82c5e41df039d337e72df
8111180cfa8990b374c64dbe734c911418a49215
18712 F20101204_AADCFW wilson_c_Page_191.pro
9f2df1daf4c074b150fcc6aa8ed192f1
33b398fac7825f3b27adf49f2fde2c66cbc58fad
80559 F20101204_AADCHA wilson_c_Page_094.jpg
6fdd42e71ad0e4077636c6386a562ec0
225e9b87c3cf0ff5d27133422d5f90968a8d3916
781541 F20101204_AADCGM wilson_c_Page_121.jp2
9419e0890006efc8fa144458c15f23c9
cb3ebf096ea093fecb0186e1fa78cb09d67b10b0
F20101204_AADCFX wilson_c_Page_141.txt
ba6cf19bd0ae0b6ef9a5305b1fc57d2f
f228802076714e69cee484991aedcc518065c23a
21726 F20101204_AADCHB wilson_c_Page_144.pro
573db6fee13b0eac0b391b882b9fff17
7a7a5cd363efea11fddc0815859474c89dbf2030
50914 F20101204_AADCGN wilson_c_Page_072.jpg
28163eb829ad49181ce9703247cd5adb
56b763145751604f8acf2c1c84fe65a8a583ed05
32441 F20101204_AADCFY wilson_c_Page_073.pro
850f40d885edd4850a4f6ef4c2f566e4
15455fad7b4262ab7f4cc2861bfdbcb9e3292e59
60397 F20101204_AADCHC wilson_c_Page_045.jp2
96e3c7b517089d2ed2aa35eca1888c4a
2a3e190e3a6637b66b0eb7bbe08c72ab11fa8d8c
F20101204_AADCGO wilson_c_Page_075.tif
c938da4376ea44be37a048915a0cd947
4f61110cc784bbe4f04ded7b18858cd12dd666b2
19438 F20101204_AADCFZ wilson_c_Page_009.pro
ffa99c0b36e7fa4b84320d7d58749465
a675b7077b642bdcb2263b45c746faea19eea407
20817 F20101204_AADCHD wilson_c_Page_158.QC.jpg
c1494394a9e53138ffeb716363687fa5
0a6ede2c246eaa646565b9eb82bf08c6f5055dbe
1752 F20101204_AADCGP wilson_c_Page_110.txt
e4425fb4d010cb37e8a8c9cdb66e18ba
a062087b7861c1dc0c1295c5212d611a949b72f3
57961 F20101204_AADCHE wilson_c_Page_075.jpg
a8d6b45b418f0719ee6d795a751b9d40
01b5a816b4965f5831874f1d732ffd7758b2346e
83164 F20101204_AADCHF wilson_c_Page_076.jpg
0953b7eaba51596a67290ecbd1a50e88
5b34b7ac45c0c28371548d881bb3c7b78ada4bc5
69638 F20101204_AADCGQ wilson_c_Page_133.jpg
89b29ed23f0febe0b6f53b4b93d448f4
6ccf8884fb2c14a7c70e2e52a8861f293cb1ef0e
19060 F20101204_AADCHG wilson_c_Page_121.QC.jpg
293b72883805592b35ceb99aa91a6aa6
3631f4e4e61f9bdf248bf9e37e2bea08b85e73c3
106018 F20101204_AADCGR wilson_c_Page_139.jp2
8bb1b3d2e7cd9784201378c913507f2c
d9f9890c86dc3eae80cd78a78285f6fc798f20bc
1264 F20101204_AADCHH wilson_c_Page_121.txt
1f5bc25ec3fdcbca034ce2cb9c85f217
08267d7dbeeacd534c34fecdc7d94558fc0c5b4d
1392 F20101204_AADCGS wilson_c_Page_127.txt
4696bd09b37b0c5f454a0b385ef32761
65db424a7c6ee317e29345d5d9eab443c5db8620
F20101204_AADCHI wilson_c_Page_135.tif
7474bb7bc2fd9fcbd824ad84fab24cca
819cc286ffbe363d5857b4623d17f418b70a5efc
1634 F20101204_AADCGT wilson_c_Page_125.txt
a4a647e46ac35b55b9d3b7ca9f8bfb18
f74e82183b0c1db54833f5a44a82b22717f1e494
99783 F20101204_AADCHJ wilson_c_Page_166.jp2
096719e50c26d21599270e1d623f24aa
6f3241bee39333080a4ee10ce712f4ec1a1d38f2
52802 F20101204_AADCGU wilson_c_Page_185.pro
08268c6af78d2fa93521c22bc6d5e804
a1a1aa5dc2d8509526624ea3dae13ac297b9b05b
13241 F20101204_AADCHK wilson_c_Page_136.QC.jpg
8f731840c10412e5fdedd4c93e3c7614
658e1e5c63e1c842d8be6d8846c41b24da1840ce
17677 F20101204_AADCGV wilson_c_Page_005.QC.jpg
336795bbc9a6d49467f13c3201a67bcc
bf1ed3cb5b48110d91cda63352117539fdf26ff1
1379 F20101204_AADCHL wilson_c_Page_145.txt
73b4d26a977d8d03157b5b706d969408
9b5cf8e343223cc05d2f7eba4d3fb9709668b509
F20101204_AADCGW wilson_c_Page_002.tif
3b44aeae97546453d63541ba315da997
ebf42962a2a99e6e2e966b3c0c77cd5085b6d1ed
1930 F20101204_AADCHM wilson_c_Page_160.txt
e3f1568e092b1351b7965f4e665e5931
04d86bb64e3561e2b0c3b965f1e8a84cccdbc974
57325 F20101204_AADCGX wilson_c_Page_168.jpg
03793a358fb8c31ef938475b4b5de935
a629f91769c4c34a435e19e2c12987dae938b6aa
1051973 F20101204_AADCIA wilson_c_Page_157.jp2
42a75c2d157d41e34179fc7650f44964
96b5c8c8268d812df8b239fcc50bacb36d598772
F20101204_AADCHN wilson_c_Page_146.tif
72fedd74f1768329aa7f5aae16667625
4a4a13df11fe1efafc8244c3c9aed4bb7b23dfe5
30095 F20101204_AADCGY wilson_c_Page_050.jpg
d781622c44e5bb505451401f0a7d28c9
8c7d1af2e8ef4dfe4080916cfbef74752d23768d
94493 F20101204_AADCIB wilson_c_Page_031.jp2
3777374f5eb0f5851d75dc982e0b32cd
c1b59eb1c4836245bb31b8378966e213fe3d5dd6
71404 F20101204_AADCHO wilson_c_Page_071.jpg
f3a75731be2e84ce0a1aca52521342f2
0fbdf21f337c0afd9b374c6b3d9a32b8a6874457
F20101204_AADCGZ wilson_c_Page_093.tif
7f166858cd7eb9087d270c0fbf296150
ecc5fb6b205de987f7f29800e0e25bc430b317aa
24395 F20101204_AADCIC wilson_c_Page_021.QC.jpg
2975c534cbce654b075e3a716b824c31
e1f743982142b31b95faa9f5e68ad12b22e88695
F20101204_AADCHP wilson_c_Page_060.tif
a0ac9a1328f7d184a3636aa647ae3a6c
60c1252a69a44276d648647962c2b89e192e389b
799579 F20101204_AADCID wilson_c_Page_182.jp2
07f4629fa0b6f41a89a610e96d98e677
fdb2773ed486192c617070a61e35e02fcd0ccadc
22569 F20101204_AADCHQ wilson_c_Page_098.QC.jpg
b33c897d96707d7c2612d393a02a554d
6c406fc9e2debb3bc9984b84d9dbfb4f032b675f
F20101204_AADCIE wilson_c_Page_118.tif
331b92b8fc14a27002141eb9a30e7d86
6c97ab7dc7a277f67fa3125839377a2b89e75cd8
F20101204_AADCIF wilson_c_Page_041.txt
c9fe6997ee2b0565cc3ee7414e4f981d
c7ec9ab0fb9673401b1f41bd2f3c15a78168b273
22763 F20101204_AADCHR wilson_c_Page_116.QC.jpg
10c38646910bc4c2c88fa42f4041a77e
d8c61b6d5220d222e819cfddf29d9270af16c761
12391 F20101204_AADCIG wilson_c_Page_053.pro
cc98f30109ff451636e9c5d0ab3918e0
d05ceb85bc8a73ef329d11d0cf5d642dba3426bf
F20101204_AADCHS wilson_c_Page_057.tif
4218ee04cbf045f9ae1e160ff89bc01a
1ee2e6875c9819ae9a6c42574f4c2637e4f1708c
1051964 F20101204_AADCIH wilson_c_Page_175.jp2
ec1b3069a42e99c7ab24a33efda8c4cd
c9ec9ca7788143027068ad66a4890c483dd6869f
92133 F20101204_AADCHT wilson_c_Page_133.jp2
309de046ade9b62b190554bc4bd30cbf
06b8a8b8ddcb76b7ca2b5166c175fe61ebb7a4d1
791184 F20101204_AADCII wilson_c_Page_025.jp2
e697a535785b4f2a04fab6359d15ba48
7928832c5baba6a2aee0d42199759a3c3eaf31ac
2437 F20101204_AADCHU wilson_c_Page_191thm.jpg
c153a5ebe239de8afa6af274809d6c66
3165a7ee9a6ef6ceea60a00cb0b8715e6660e4b4
31297 F20101204_AADCIJ wilson_c_Page_108.pro
c32ac6c5d98b87d7b4b3622fc67e446d
bdbeb936b0af7e387e1d1cb9f6f66904681ea78a
76357 F20101204_AADCHV wilson_c_Page_091.jpg
002478f2dac20b5e24dd4d3075308a60
c29da1e291f58c7a2d218290034d27b4ee3c1905
1263 F20101204_AADCIK wilson_c_Page_180.txt
bb6528f5fe6487008a2eba5b094fa859
ff91ab544fa8af12e92b98e054d12d495e098352
5498 F20101204_AADCIL wilson_c_Page_023thm.jpg
25ece36345a13f7716b5764681309458
e0a263fa98cd7d9fcfcb700f4fe1343c4c220a80
73294 F20101204_AADCHW wilson_c_Page_031.jpg
9d9e48da1fa325933b84e97b11c23427
d6b08d3cabfa26e2a8a6cec0ac696eb781041485
16342 F20101204_AADCJA wilson_c_Page_088.QC.jpg
50ddc48cb0afefefb5b32b1940a6118c
befc35c6019b10c802f0ae8e8d9b36f771992ba2
2711 F20101204_AADCIM wilson_c_Page_014.txt
2483b393f77a209808870bc10bd23ee3
43d6006e7ca2f5c513e10b93ceba7d71d81a96d3
50224 F20101204_AADCHX wilson_c_Page_052.pro
1c5593260915831e6f738e86e027207a
c2794c9a0651e49c5bacd08b8de9cbdf74bf5942
14700 F20101204_AADCJB wilson_c_Page_192.QC.jpg
1190997d4d41012141137c26397c6de6
772e4a68db58ee1f3b64691009805dfe7ea01ca4
18471 F20101204_AADCIN wilson_c_Page_089.QC.jpg
f498b5e8dfb441fdb0220a8b3d43c7f0
8f7f99772ef8d8d4ed3efb5d2f77e27c0a40d880
6068 F20101204_AADCHY wilson_c_Page_048thm.jpg
016bc50bce94a033afdd04e6e3f4a868
9ca7f5d2ef78d5a774247a77442fa169fdb1f89f
77775 F20101204_AADCJC wilson_c_Page_021.jpg
81408ce5a82869b1ac3c2c1a4fdf86d5
292187ab0081ca1b5b9699f1093659515b9a72bd
5770 F20101204_AADCIO wilson_c_Page_061thm.jpg
63a5db7583e8d1456d62b68d800be368
a5f38963612c2cf3458619af416fbc22f3783fd5
52481 F20101204_AADCHZ wilson_c_Page_142.jpg
48112f79b5f0cc2dd1e1d6d03a82fada
e007c18817e795332a3ca50bb75b279010b602b7
5175 F20101204_AADCJD wilson_c_Page_037thm.jpg
26335656161498460001c96f2d26c28e
f329e1dbc67e425b347f7549f72fae3ebdda66e8
651374 F20101204_AADCIP wilson_c_Page_040.jp2
be4b3187c49d8778dc18d5b4b11ab307
1e7e45f6bad8b057c0741eb02aea1d1fd28df3bb
1978 F20101204_AADCJE wilson_c_Page_114.txt
a9e67021af1a99fa717c01aa29259f66
6e590f21c07b295c4898b18d129e4473e105257e
23874 F20101204_AADCIQ wilson_c_Page_006.QC.jpg
27ca69a954f29c358fe74035b9212d61
b37e89476e704728ea21f60c8b652346bacd0560
98717 F20101204_AADCJF wilson_c_Page_137.jp2
b9944c02405c0e617b08a474d7b4cd72
f1235b3f89693d61ba3357ce221bf562cbbbdcbd
73409 F20101204_AADCIR wilson_c_Page_096.jpg
1de111c6f25384a62300eb0e1dd5cb88
e4be830284f98c65df77f8d5a9bf2fe1658d8783
49858 F20101204_AADCJG wilson_c_Page_034.jpg
021ffa716217a0f227cb6b9795313db8
9e1ca89ca03317f1325fd8d33d61177a451f5331
325050 F20101204_AADCJH wilson_c_Page_069.jp2
35be72dec016ae694a4858ffad7b779a
4e5e2cac3b3e54042a3c1579728e68f8969632c1
47703 F20101204_AADCIS wilson_c_Page_060.pro
0e0a9fc28a08bed54f5e6b7e010907a4
365176d639bbb58f19d076d6afe292895600eb27
64918 F20101204_AADCJI wilson_c_Page_079.jpg
f7f1d7dec3a32e93f5f5d1c60c9eee36
93fa28def2abbc58f9e1cd494ec5e277c0f892d3
78223 F20101204_AADCIT wilson_c_Page_083.jpg
c32badc93ea530ec8ba2257ec3451721
a8e0e28c63310a5bef6782dc0b618141e42b7246
F20101204_AADCJJ wilson_c_Page_190.tif
0d0726dc226af5cd8fee8b2bbf80a121
21b105947e83343556ad8960b1a1ad41fe13fb47
37667 F20101204_AADCIU wilson_c_Page_078.pro
43a7d06cf8fffc134ad72d8bba327de5
88faece5cedf9ff51d8e99f68976c86ca3f38226
102964 F20101204_AADCJK wilson_c_Page_022.jp2
32a351d2de0861ba5aab4eb0e2664ad7
fb4df7c1dbddf08c64ec9206482cda622544727b
63781 F20101204_AADCIV wilson_c_Page_073.jpg
fc4b69440b3458c9ec844d81d123e800
c2fff7cad15ddb3d463ce538ab9d67a7e39e25b1
26194 F20101204_AADCJL wilson_c_Page_192.pro
b3d50700bacfdcc406eaa069359970cc
3c534feef717eb950e7f4fc151d195e2c0702aeb
8790 F20101204_AADCIW wilson_c_Page_164.pro
212eff10dfdad35a37d555844bd37085
9f40948d106eca2e0b6f1dff41d0ceb8d439299d
26744 F20101204_AADCJM wilson_c_Page_049.QC.jpg
957894b8941de69d222141b4e239891a
2eae2ef46166e45b50bd0f59dd58bbfc57a8ed00
4526 F20101204_AADCIX wilson_c_Page_130thm.jpg
0a0a3e882a25bd37d0aaab7c51662850
c24485c54f55511675a1d690b9367f36a8221940
F20101204_AADCKA wilson_c_Page_047.tif
0449a61608383b05310a8efd874cfb8e
0871acaae0e16e37297a3224bd450066e4b9039c
5310 F20101204_AADCJN wilson_c_Page_096thm.jpg
4832428e550e3fbb803805ae9e08a73b
b0cb840cb12ab1cf8fa959e7035f9e93b93bd15f
25241 F20101204_AADCIY wilson_c_Page_019.QC.jpg
e41a00ceb7b5603b96ff024f9d94f2c7
d72bf8c9561ffb681df10ec2ab3a37d1c00072ad
1052 F20101204_AADCKB wilson_c_Page_148.txt
bcab24f4ff3d5eb1af60eeb329bf58f9
21ec8804d9679d78c6c9378aa088b855e9aa3d6a
760 F20101204_AADCJO wilson_c_Page_088.txt
46ee6d4596b7cd58cabe56f899cec198
066357b496a63ce7b7b93ff7d8574359f7f87bf0
43534 F20101204_AADCIZ wilson_c_Page_063.pro
bc414319d5ef4845c3f7ccf7b8315bf8
e176a1d119d12ce990a0fd5ad2b8d1ae674a04c5
19790 F20101204_AADCKC wilson_c_Page_081.QC.jpg
a0282a1975e9d8a93c74900a5e63e6be
f4ad447bd20d505a7e9c15930560c641ac67142d
19037 F20101204_AADCJP wilson_c_Page_111.QC.jpg
33fdfdb0ef2daf9706ce519cd6e9f876
05197b411ccd4bdb82b67f237801aa79ed2fd46c
104612 F20101204_AADCKD wilson_c_Page_124.jp2
fa6b8b72afc27a10965b2547c665b481
e87de3a2f3de5e203b6491d8bb02bcf426cee374
8806 F20101204_AADCJQ wilson_c_Page_149.pro
9d8619322cf153fbffd82e21cd9ee159
0972e05ea9267b02bb11055d06402529a4378273
3583 F20101204_AADCKE wilson_c_Page_161thm.jpg
843ae8c1b601edc95c1787396924c1ca
0c757eabc18e8a3ab51d59626472f9e94c7037c4
F20101204_AADCJR wilson_c_Page_067.tif
aa9b89eeb5e772d078b88fb584c771cf
260b4e7c90fc91cbf54f3bfb8b3f8c5d4bb01eb3
1688 F20101204_AADCKF wilson_c_Page_086.txt
fa7591acbed336386d67f66696bf3248
f045117b284c768d333c4fb93dd661578cb601c8
6953 F20101204_AADCJS wilson_c_Page_176.pro
b110f5558235a324b9505e1bf3aed326
da953a07f0873749a1ffdd3d86c28854918146d6
F20101204_AADCKG wilson_c_Page_144.tif
134c6a83cf70b5a6a0f5085f96fbb2b5
78d67f71eebc1beec59da2e7ee85ad54ec032824
4783 F20101204_AADCKH wilson_c_Page_169thm.jpg
aaac7e83b11225f7fdf780fc8fcb8b8a
ac44e14b5a24b3b75faf05593f49575a4aed3cfa
F20101204_AADCJT wilson_c_Page_112.pro
412710cb5c20d7b5d2a596f98b92f695
c607c41c602cb623f07b777deafe0420d7f90f8d
F20101204_AADCKI wilson_c_Page_163.tif
6b2be480501cd868ca83d4ec4a89ef7a
4d0ec42b31d5d1b391e876648df8cbfa6f65a4e5
2001 F20101204_AADCJU wilson_c_Page_183.txt
14746c8a63550d7ebd72461e6e3f8f72
67d9a781ceb50b6d7952a9b94c7ead5846576f87
38635 F20101204_AADCKJ wilson_c_Page_098.pro
e6936795e8a9b3c2e5c1cab3fddb30f3
24172a5a6c5396c23e8fddb6d058144380489637
1137 F20101204_AADCJV wilson_c_Page_072.txt
2fbc197fc5dcad8706bd325bc3d3e1a1
473589a8ec936bd9a0707c3337b9425385708579
49065 F20101204_AADCKK wilson_c_Page_187.pro
5e8395825e835abdb6e623d1af92083e
5219417e38e1574476bddf42047c44efc5563913
F20101204_AADCJW wilson_c_Page_152.tif
3a0874192818230394dbcd68850ccdde
3ba252ce42e4269acf4aadc89933935c80251af0
9948 F20101204_AADCKL wilson_c_Page_064.pro
6242bdbf882d2fc98a34e46395cca6df
7463b1412aefc02d69277fe5f4fb1d661c981866
86349 F20101204_AADCJX wilson_c_Page_174.jpg
0fe6b9f64a314b5dc46fc59e8ac4f5d6
a390676188a26e5f75517c6a7526144b33579349
443 F20101204_AADCLA wilson_c_Page_056.txt
5c94c805c95d8d97f0d8ca1ace59319a
e04f90e535c2103d049364bb78f45b92a6fefe98
5914 F20101204_AADCKM wilson_c_Page_118thm.jpg
2c9071dabc0e8d0b550497ea2b88f782
4c25d860bde3322f4f5b3e8af68df43f54a30fc1
16354 F20101204_AADCJY wilson_c_Page_087.pro
396bb74de0c5d84b25ffb203a5bf2a52
7c951172062b49557ee74fe1b9d688889ebd7e97
F20101204_AADCLB wilson_c_Page_081.txt
e5af32aa1fbb9cd99126595f1c91815d
3eb9ad2f38112c42fac0f8cb788cfef028bccde0
21700 F20101204_AADCKN wilson_c_Page_110.QC.jpg
fc2faf236a37c13f85d94ed111e2a4d1
4884e21dcc5afec5414b15779e36f7b4b545f5ff
3299 F20101204_AADCJZ wilson_c_Page_030thm.jpg
48d06c94866006853576aae80793ccee
81e7be350d3c617427891477254d88f94962175d
731 F20101204_AADCLC wilson_c_Page_101.txt
d049a8fd2d4223d7bce8c64999f97ad6
04980b47634d6ade28b7f01452ac1679979c71d7
30329 F20101204_AADCKO wilson_c_Page_016.QC.jpg
a14bcebc7269c23869ea7010b6380bad
e1c1ce7df63ed64083f7ab76d0bd4d0f374f4854
499668 F20101204_AADCLD wilson_c_Page_129.jp2
5c3ff0a19f5b014c58a1d016fd127bf7
2df71b60535f75e0acd26bf4a865ebf8dc10b481
53302 F20101204_AADCKP wilson_c_Page_147.jpg
7c0ce6809359687c03461b2faac7279b
45b63b433932a71f8cdaec3ed784b89bc3378b12
1764 F20101204_AADCLE wilson_c_Page_119.txt
c17911982a2022166ad951c2da69dcce
cd503618150143246536aca2828b80ea8e6f1229
87908 F20101204_AADCKQ wilson_c_Page_098.jp2
9fd598a5f2c59cf6e0d7e004e6edeffc
f6a01aac4662167806fcf7e365d72f6692402093
81057 F20101204_AADCLF wilson_c_Page_055.jpg
85b63b79ab81be610c68a4b70782e93c
e96ad06bc657da97879a888ea75f7cb0ebb4e5ed
8664 F20101204_AADCKR wilson_c_Page_175.pro
a5164bccb7eb8274b0a170463504a9ec
6d32ed1fd4a3f7aed9af507e268271b5f406dfba
4899 F20101204_AADCLG wilson_c_Page_018thm.jpg
347cc26ad3ce2a76febc022870ba2cb8
fcfd1ed4520985e1de33418f683d1975abea6743
4907 F20101204_AADCKS wilson_c_Page_141thm.jpg
ab1fc17152e644ccf61a6066a7d8b2e1
8b26a230f52977ec1ab8c72ebd37261d4c325ee8
79756 F20101204_AADCLH wilson_c_Page_057.jpg
f7c5b18f4b5c3ac264e2057eb871b45b
f5e532e925ddfeb4f9af262772c25fca33d69610
F20101204_AADCKT wilson_c_Page_008.tif
cbc3625a8ac749243644d8276e50f6c5
cd36df92b51a9d5ab2ced7697d0f33ced994d0cb
37476 F20101204_AADCLI wilson_c_Page_129.jpg
0d6b8d39ad81512d16ef507cba208849
e4eb87831d184e10d144d496d2d27a3328b30c4a
71078 F20101204_AADCLJ wilson_c_Page_080.jpg
05def363f7d6f6bb7766adff72e06267
5614cd3c6fc7af4dda7971efd5869244a6557f91
9668 F20101204_AADCKU wilson_c_Page_064.QC.jpg
8703ebafd3d2c356dc42e0319c8c5c8d
c3693286bbbab652ca16dd2d751cf6adfcd6a1ce
46803 F20101204_AADCLK wilson_c_Page_021.pro
4b26faee094b806a8f30114aed185674
2603fbfc3aa598bbe1624a0626e3fd562de221fa
85663 F20101204_AADCKV wilson_c_Page_150.jp2
b5dccdf85935c70e0bf5fb1ee8148d55
684552a8275aa5eb825ea45699ccdb23dea176f5
3668 F20101204_AADCLL wilson_c_Page_123thm.jpg
728d2463a05b1bfbb6aa98410182359d
47f49e22157161c9a479c2b7e0967b3ce19279f2
F20101204_AADCKW wilson_c_Page_072.tif
4bf725c11d5f3fb6a3d40f0b8ea75b4e
45ee16916447b91d724ce3f656645cf35ebe9f5f
76798 F20101204_AADCMA wilson_c_Page_026.jpg
9ec6b0cd41ef612f3b14a3684e9665f5
0ddb3f0d876b176c15d8717923914eccad4c1794
22495 F20101204_AADCLM wilson_c_Page_187.QC.jpg
2b6de94e24ba362ce56342c33105ad1b
dac27d7b87a50ad8ee95646625890e5d33a0e3f3
36070 F20101204_AADCKX wilson_c_Page_079.pro
8d1d7e40459833f32936b8473edac8ee
a64b7625835001affe5a8a45de256a39c85ad126
62779 F20101204_AADCMB wilson_c_Page_041.jpg
63ea5b925cfe70beb3598bc8ddd6b6f8
2bf3ae2c262f8d553e6b053c6851d9412a9a5f21
285951 F20101204_AADCLN UFE0011396_00001.xml
f22d66fecdec6d618f7505d710ecfbb6
88cd25f2971405b1af71c6f1f375a261cdfa4f0c
F20101204_AADCKY wilson_c_Page_145.tif
9ab09029c18961372289d6800b76c9e3
2470f5239beaf7c61ba9ccf96c248cf2654f4bbc
F20101204_AADBJA wilson_c_Page_025.tif
8219b2410e3af29f21f7282bd4c1fac1
469e3856e84081cffd8dedc7179a5cafca7a4cc5
60821 F20101204_AADCMC wilson_c_Page_042.jpg
28e2e3fd9b669c31c7ff4957d7f1a775
e66bd994fc44b0d64342f1525d1f79300d1cece8
6919 F20101204_AADBIM wilson_c_Page_011thm.jpg
63cdac05ec5b4091369ed81c16d1d450
c71a63f1e0b9dc0f2072bb5d8d6367e4bf8f862f
16105 F20101204_AADCKZ wilson_c_Page_169.QC.jpg
aa30016c06cb93a163bd272fe711c42a
eecb329280766a4a14af8d66d5b759126a4c0275
89865 F20101204_AADBJB wilson_c_Page_125.jp2
2b1c03c8c6380fb3a1f4654fb62e9388
bd98d3e343a0ee1b2a794acfd72ed69c167693dc
60969 F20101204_AADCMD wilson_c_Page_043.jpg
305d0c43329428065434711c95d1c2bc
2f2458f86682b611ec36b9582a2ae66ba2eaf7ea
24045 F20101204_AADBIN wilson_c_Page_027.QC.jpg
7d1b5ed413fc00e4cb8d8c6ca22f9297
50ff38eebad8cf42be4cc231fa7ad0227b576e82
24769 F20101204_AADBJC wilson_c_Page_160.QC.jpg
a04ad36828c2a0331181dc5261e61bf9
010dd3ba77fac356a054f995de5c7b5b29b433a6
67171 F20101204_AADCME wilson_c_Page_044.jpg
dce3a5737f942036c9c56762394a60ca
7a12ff2d7c204a164f643db8ec25814275082d02
80783 F20101204_AADBIO wilson_c_Page_160.jpg
d31bf0cfdb5218b8f77dee885e439e55
715ef06d8281515b90e3353cf5b9e71baefb94d2
8418 F20101204_AADCLQ wilson_c_Page_003.jpg
43865cf0f11e38e5fba6657968d145de
71718cff45e0873890d2d9e2413ec34bd06622ba
4191 F20101204_AADBJD wilson_c_Page_148thm.jpg
a972811999f706bef64e448f3746232f
eb6ceaa92478ac95fc74b55d2df504609ed5dd0d
85118 F20101204_AADCMF wilson_c_Page_047.jpg
f8e87295e35a4bd498f112a68cb7d8d9
0083252622ea75dcae5bd9b9779514391697a51c
93056 F20101204_AADBIP wilson_c_Page_118.jp2
95f36053aef0d3f22473e85c099bbddc
6c8712b83571628e897be59143d39ba37f259307
54923 F20101204_AADCLR wilson_c_Page_004.jpg
dfcb70b824a0607f02373857f7d59739
f40c6a63e80611d17cc8097c8dc1aa5df404fe85
F20101204_AADBJE wilson_c_Page_186.tif
e6fb159697958439bbb018ae2eb98bf2
e489387278ad8351bc4744cf73a8351742cc737a
85907 F20101204_AADCMG wilson_c_Page_049.jpg
42eca13f449cfce147d7559a8bca723a
ac954badc364172449ec66fc3ef45305f51bab8c
8137 F20101204_AADBIQ wilson_c_Page_058.pro
0a23fda026377945e9acdfed8c91ec46
542b6dccd3fdd9b7a5a5d170bf3b8ac9cf1a453e
112841 F20101204_AADCLS wilson_c_Page_006.jpg
f20c8a78bc18de3fa9da3b68083ac133
cd9c5a3cb377d070372bf4a51ca98ec854f0a200
F20101204_AADBJF wilson_c_Page_008.jp2
191bddce8edd38e36bb832cab13eb9ee
2015590e3596f86e089cdf4339dc163dba5c6c47
65143 F20101204_AADCMH wilson_c_Page_051.jpg
0442b336b81949bafe7dd17ca04354a3
1d6526483365a5af6a14de6e73b50f25c8285778
104952 F20101204_AADBIR wilson_c_Page_094.jp2
f5fc4de7d6c12bf05783d62f5c470dd2
183e5375e3915d4952cf3690838353b083186031
46098 F20101204_AADCLT wilson_c_Page_007.jpg
7d137c480890c1e2c665fef8b552f481
853909c75f813a0bf8d599d2cb5ce9cb99729f57
18550 F20101204_AADBJG wilson_c_Page_043.QC.jpg
7bbeb5aa49a081b01d9537ea07494535
d630dc59bccd04a6abd2c2cd18c24ea08a1b9802
40928 F20101204_AADCMI wilson_c_Page_053.jpg
984fb5cb814c7c8f91a40fb635437527
706bb2d687194cc2fd82ccbc601aa0803329ac70
5683 F20101204_AADBIS wilson_c_Page_155thm.jpg
f8242b66597b65cc2313e8920a11433b
627eba512ca37fc58e5532e72eb841b1a338f51b
37524 F20101204_AADCLU wilson_c_Page_009.jpg
6ecf27b0848fbb512db51cea372e6638
36a94f81c66754ac5cb918664ad6eb51771f6f20
27171 F20101204_AADBJH wilson_c_Page_173.pro
b4ae8f224fde34f2fdd88622d33ac657
9116727325ccb069dd7eabeaf1fec48958c1318e
29275 F20101204_AADCMJ wilson_c_Page_056.jpg
900ad14cd99bbcd42f8ae79a3220d4a3
d67220c3f80dde35bdad83ef410ff839e18468ab
11104 F20101204_AADBJI wilson_c_Page_050.QC.jpg
9e49f4ebd36b595f4c33b61e8c206f60
75ccfb83a20e4de194dd086b1e3c2135c7e39361
45346 F20101204_AADCMK wilson_c_Page_058.jpg
3b4702cebb6d9cbf4fbd0de041b7594c
6215b55af151bcdde754d70d9a74bbfe05c31805
84493 F20101204_AADBIT wilson_c_Page_114.jpg
81e23728b45505436ece457cb42ef1e5
3c3957b3fee2a582d3493e055a71af50e3b7e27a
111449 F20101204_AADCLV wilson_c_Page_011.jpg
88c2c7382317cc0a17214d8e502a38e5
cdcb2ce01d505cfeea5183f0911878391a60fa15
54262 F20101204_AADCML wilson_c_Page_059.jpg
61b4594de7a6d4b4b1aff60c644c8f9d
75002ce9da1876d0025ac585b00357163a934752
4842 F20101204_AADBIU wilson_c_Page_135thm.jpg
d3183604fa1c7c0c8f2fdc20f730bba7
3a728abe90e9c63f9f5e3f53c1fc790cd21b48bb
123905 F20101204_AADCLW wilson_c_Page_015.jpg
7286b2dc9572e9c0e1fb4d2500048ac8
2a910c352e5c724c4ccaf329d1a7e616c86191b2
16211 F20101204_AADBJJ wilson_c_Page_040.pro
d2e31e466729882678da96522190bb09
c0ac21f0acb3c4a3afad3e958e361c43671aaa01
81431 F20101204_AADCNA wilson_c_Page_093.jpg
8461d449e75614ed29baf14ff920b3bc
2d3d570e113e6e8d1edf7c4842a799c728c86900
77636 F20101204_AADCMM wilson_c_Page_060.jpg
cbd5ead26a15b1566de61956894c8d8b
e4b0b8fc0247091417a6674a9ba61514963cc3d6
47684 F20101204_AADBIV wilson_c_Page_126.jpg
8a0e305608e303874b1fb89e18d49c85
f1ae6a43cd36ab6004a355f86cf2993901f14226
24991 F20101204_AADCLX wilson_c_Page_017.jpg
d1f9ab1a53276a1ed44e9c9e65b608cf
08d1701dfbb5aaf0f2dc9b7c1dce9e3ea0e6e035
4538 F20101204_AADBJK wilson_c_Page_144thm.jpg
64b70837eeff94b4d0868399b9bebed0
4e216846aa2b3ab56753149d4263978448cddefd
34213 F20101204_AADCNB wilson_c_Page_100.jpg
bd60690a014ee72866f446caefe5aaa8
46ee090be3ee8ffe17d16ac05abcadd7a89e63b3
67759 F20101204_AADCMN wilson_c_Page_061.jpg
7e5bcf0c23945a5318b360452b88026e
9671e23449e3ef84dc12bc3096067f11bdd339e9
24868 F20101204_AADBIW wilson_c_Page_024.QC.jpg
4e2527d83a4c55dcb3f29a7cb2edd163
42cecedb9fa5cc997b0f7aed139f5b8875d84899
68145 F20101204_AADCLY wilson_c_Page_023.jpg
35290a985f30ad03db22873a060c464d
bf3f01bc7ea62b2c46d5f560c010dced55be1d57
F20101204_AADBJL wilson_c_Page_067.txt
29d7c1ea2062b2c00bacd38c586a4c73
c1b60b05be784d306e6315e56bc88f2d2559230b
33028 F20101204_AADCNC wilson_c_Page_101.jpg
4206d9e26626b087bc30edad63862df8
0fa2c54b1bdbd0b6fded32f43d45e48db467c9f8
52670 F20101204_AADCMO wilson_c_Page_062.jpg
e7042b06d3516821d2faa14d2563baef
61ffadbf33e375b493b47668c44de933ce3049c6
20181 F20101204_AADBIX wilson_c_Page_051.QC.jpg
56e36197f19ab7e9e8272768408fed53
7a0897e3b0f1ed33378c5f582c8b2c4ec7bc3ab1
78556 F20101204_AADCLZ wilson_c_Page_024.jpg
ff73950528de9eeac5c4047e75e013c6
b571abf42003221151e5bc324df71a962085b04f
101527 F20101204_AADBKA wilson_c_Page_074.jp2
bef1710d7d179d629c3838c1c61696a8
33c06ab6d51601b75b79fe552febe6fbd89e8c2f
47692 F20101204_AADBJM wilson_c_Page_134.pro
81f4820ac1d648961012f8f00220dafe
d7335636433d887ea534c9512554ea3b7e839c9e
79285 F20101204_AADCND wilson_c_Page_103.jpg
8204a0cf5dd1951ab4a107cec75b16b6
d1e4f9bbd2cd369cf7acac17c896c3ab3ac46a23
28603 F20101204_AADCMP wilson_c_Page_064.jpg
fa3cee65a68b7c2a049e1d5393b51e4b
d2b765b07e10089169de2e92723c1917a9a755a3
11211 F20101204_AADBIY wilson_c_Page_167.pro
b68b83675ae3d8eb1deb869c94919902
50334f0ef33b21386c710a4dd2d1d326ff01ceeb
4570 F20101204_AADBKB wilson_c_Page_065thm.jpg
3ee42964567c43e97f6f92d1297e73c9
43821f33890c49e4887968450f3c5dacb478e939
6020 F20101204_AADBJN wilson_c_Page_166thm.jpg
87fe241c4b1172a49dcaca4e8c995185
3f7621268d5b796a9977942507169b2aeab91d85
42063 F20101204_AADCNE wilson_c_Page_104.jpg
14209b3387fba5ebe8533ba76a45cf98
9df8e5d5151d4c9eb4497ec2442ac82f59d73255
65726 F20101204_AADCMQ wilson_c_Page_066.jpg
1cce0c06c374560798dabe5332615966
ac4f85c2b44b5be3bada14be6291aaec52cce91f
14362 F20101204_AADBIZ wilson_c_Page_136.pro
bda74c09d2279337b62bf6662ef18dd6
90e9b937d3205d3ec066eccdc5bc69c17ddbf4bf
71104 F20101204_AADBKC wilson_c_Page_110.jpg
20ea63c0dfbcd5bb14c5ca6b3b492c42
6deb7ac9ae4165485e24485bb5fdf2bcb29d7c09
10887 F20101204_AADBJO wilson_c_Page_007.QC.jpg
dfe97d3ce386fdb44f2b2c6f49bdbd82
3324846b7a016b42420bc6d55f1c4ea319032ca4
51338 F20101204_AADCNF wilson_c_Page_109.jpg
d133c3ee93e130c6c634301b71aa9778
d440e8338430e567600ca53943d863033f75a4b5
79005 F20101204_AADCMR wilson_c_Page_067.jpg
5a6cb2b8b93156e1bff62984e6296fb0
b4ced031a321a1ef3537b124e221dc8f172d8661
2207 F20101204_AADBKD wilson_c_Page_188.txt
a8626b941187059e81bb41643d133c9b
05fcff904751870921de99291b2b9331de96d2af
23775 F20101204_AADBJP wilson_c_Page_085.QC.jpg
ed94e0d60d684640bb2adc00b1b63e3a
7afffffe62199f3ef2edec4b3fdf8ea0164db535
40741 F20101204_AADCNG wilson_c_Page_112.jpg
5515333a44257d7c6008632e8622e16e
2d900d6fdecf07c0ae805c8d6d3f67b920ea4322
27354 F20101204_AADCMS wilson_c_Page_069.jpg
dc1eeb06d7e7acfb780aefe6d69ad58d
5983d14e9bfd9016cd26a36e88c300d268c8817b
F20101204_AADBKE wilson_c_Page_017.tif
fb3c26f0b61964cc9753784b9d6d621b
c6221eb1b2d65cc26de46198e9bb42a84f549466
49563 F20101204_AADBJQ wilson_c_Page_171.pro
f723c47d8e6ea38bc50199f425302cb9
b147053c433c8a08dcf5ef8ed4cb009c7952940a
47626 F20101204_AADCNH wilson_c_Page_113.jpg
2fcd7d8143914b4af5f40c13e37f89a5
15619c89cb5b404d07cac07afa4407b3897d84d8
81487 F20101204_AADCMT wilson_c_Page_074.jpg
191661f47cbd6387f916d3152a0e8eac
3113f1afff68f9330be24c147040bc425ee1e98c
968 F20101204_AADBKF wilson_c_Page_030.txt
0ef277c723cd9c676a8ec7eb24a502ed
0f927fdb5641dd25156b0eef8128b62a1f594ee4
24874 F20101204_AADBJR wilson_c_Page_156.QC.jpg
08645ba1b69bebad829b8566ec5352bc
b822e8c678b3fb95ff066ff265dfae06b4b12575
78176 F20101204_AADCNI wilson_c_Page_118.jpg
7c78f28c756b6ea32db9620ba56ff1b4
770e963147a866706fa6a7bda6b749aea93f4cf9
64802 F20101204_AADCMU wilson_c_Page_078.jpg
3aeb24989a3e6af1d9fee39e970060c5
2c3e62aeca036b486392cde7be745fb6118d95ad
49010 F20101204_AADBKG wilson_c_Page_160.pro
c897281241c9d9bcbdf990fb277c670a
0d70ac0409f64e4862ec30c27c55d348a41c0fb0
37731 F20101204_AADBJS wilson_c_Page_041.pro
9ff73200b52a0c36ed0b142b7a7c6975
554013302728fffb59babf86e5c3f6a02bca51a7
30169 F20101204_AADCNJ wilson_c_Page_123.jpg
44c16b30502bf13b00d9d2ec18c0cf5e
5de93329a13dbd25736941804fc8dc521a4aed61
62245 F20101204_AADCMV wilson_c_Page_081.jpg
82a18c55c453df9a12ffdd4d13c88f00
f9da030007c4f629c39c407a14037c657312a455
46002 F20101204_AADBKH wilson_c_Page_040.jpg
8a19253ff624512fe1e1e522e6d00243
0b2d0d19dcd19668edc467acb34c9ca8ecde8305
49077 F20101204_AADBJT wilson_c_Page_087.jpg
23eed0730f5baaace74ca4bf77c6403a
328e256be752409b1f7f786afa3f405cc2b029d6
81024 F20101204_AADCNK wilson_c_Page_124.jpg
d690cafd6b07db8c6227477f70020cdd
5601186bb091812c84d6a4f7b34ae91f239374aa
2800 F20101204_AADBKI wilson_c_Page_013.txt
5ba79c8e0bcded354833a3eaa5700f21
3d14c1352218e7edeea197c8453583fe1a8e4507
71820 F20101204_AADCNL wilson_c_Page_131.jpg
370a0e28507985b903818b2a9323ce19
8b37902c679b0f04fb92c80c986bd9bea2b0d9cf
75810 F20101204_AADCMW wilson_c_Page_082.jpg
13b690cdfe1aae0b21685d3939422dbf
eb156e04de8640c7612c674f9de2b39a2abea2f0
29178 F20101204_AADBKJ wilson_c_Page_011.QC.jpg
e45948ea2636d00e03356804e33d4caf
2edf938ceae8e9cb7f8e979b8fbed0e6ec3feaeb
22344 F20101204_AADBJU wilson_c_Page_077.jpg
3e864cce7799b6cb8077d136e21d673f
5ee749f8a99283e5d747e4ed831afe6b0011511c
80718 F20101204_AADCOA wilson_c_Page_172.jpg
84775f526a43cef3af42541453ae27a0
a4af0c6afd5627b2597d4ba6f891e7465ca1ecf8
52548 F20101204_AADCNM wilson_c_Page_135.jpg
bdb0ceb5ea27bc1549800390babe67b3
3c641f2798fb95be6aa814dc2d1471b04e74ac11
76996 F20101204_AADCMX wilson_c_Page_084.jpg
786c3bab4157480dca845349918d8d66
27a9524815dd34be261a1eb3cbc6ad95f6c1abe0
566 F20101204_AADBKK wilson_c_Page_153.txt
db889a920cec6e7b2a5a096fdbb8a950
0bf015fdf7caf5524f8c9f0131e1beeb808e8ae2
F20101204_AADBJV wilson_c_Page_065.tif
086daee84575529b00d838d57b49e302
322a9fadf8375004285d6a01eddbf208396de41d
64635 F20101204_AADCOB wilson_c_Page_175.jpg
3f53e52d394e62dbd81a615b0868bbb5
84c4e8bdcceee6b61169e2378cd27a4946755564
72316 F20101204_AADCNN wilson_c_Page_138.jpg
4390370558a5327bd1c94833e57c8a5e
2e98e0ad350881e0525b46765bbbab6459e88e20
76719 F20101204_AADCMY wilson_c_Page_085.jpg
c464f2b02782c53e6ad166857f94c8cb
00cd973535ea5897e37ae488af53b3ed8f10461d
48169 F20101204_AADBKL wilson_c_Page_094.pro
8d0f51b4dc08bacba56447992c438e55
c68aa0a83fa95d6d5a4da8709d07896361571096
8023 F20101204_AADBJW wilson_c_Page_050.pro
f9134fe88f6c3f284ae33065ed3d27b2
58dbd7980520cc8f1c330e36d054ca01a9da64fa
64098 F20101204_AADCOC wilson_c_Page_178.jpg
a47bf85ca2485feaa7c706bbfec069a2
7a075851459a3398ae17a86dc88e159593429705
81381 F20101204_AADCNO wilson_c_Page_139.jpg
d310fbca974248db2782503618ccc70c
400f69abaef3285771e3787ee1698539f079978c
49413 F20101204_AADCMZ wilson_c_Page_092.jpg
0e773a39dce566bf58651151ce30d60f
b87fe7d3ecfd243e20b46f6e9cad767b659eed70
6412 F20101204_AADBLA wilson_c_Page_183thm.jpg
4fc71e109b33ef46a5a6764e66229d85
2328c5823cda847f5b8e3f734fddfe18f5b40a06
106782 F20101204_AADBKM wilson_c_Page_068.jp2
fc334173fd3b97edff79892e9936359a
6215623c35e302e0202c7cf9457f2c2e5b0984aa
100006 F20101204_AADBJX wilson_c_Page_054.jp2
da5dddfdb766d2f6baacc90354a8c1c5
1169e42b0cd1f4bf05d3f46d5c8a1d1c697ca92c
83400 F20101204_AADCOD wilson_c_Page_179.jpg
ada3d3007733844e195780395133f1ca
413edbc0a39efd38ac16e8becb10225d2292144d
29568 F20101204_AADCNP wilson_c_Page_149.jpg
5d0cd8b9086bfc4b3a229afb05c1f22c
f449d0a5ee79a10af23d7834664d41bb8f3a2ec4
61668 F20101204_AADBLB wilson_c_Page_113.jp2
dcd3e4a404ec9e6f9b13447432a8e3b5
94eaec1b448d50f813c8993a7b9df18ca39a293a
11262 F20101204_AADBKN wilson_c_Page_075.pro
6e7053ceccdf7a8284c8c1f511109df5
eacab0044ee57f86b3ca337eb9696541f41a80a0
1763 F20101204_AADBJY wilson_c_Page_077thm.jpg
5965fc6457aa64f2c5e89f32393a4923
166f0f84f816148662f85611ba7bfba61ddc9500
88736 F20101204_AADCOE wilson_c_Page_180.jpg
9475e2106389b26f3b9214c0e1b319f7
2821c3e585b20f4b827bac587be726108bc37895
70283 F20101204_AADCNQ wilson_c_Page_150.jpg
dc1000adf3d77af35a903fc7fd128580
50168b76fe187f49c41b15ae6719e6ac717bc131
47906 F20101204_AADBLC wilson_c_Page_019.pro
9c058f10a760810e35b7c975c9ecfe74
162bb162d419708485136d6094abbb22a79e1255
2901 F20101204_AADBKO wilson_c_Page_012.txt
8fd913065f1d8519cb541ac866f20382
d6f01b27f3a97c5d83545266307387b9a715fdf3
747745 F20101204_AADBJZ wilson_c_Page_127.jp2
3032480e3aa632c87861d0aea01a22a4
60732de6cca54595af2f26ac0ee27020434ee78c
74185 F20101204_AADCOF wilson_c_Page_181.jpg
dec267131f669084563f13f4cbe054ea
c0cadb273fb314a97f2554d099aea0d5c1d4ff94
79136 F20101204_AADCNR wilson_c_Page_152.jpg
9aa2791e4203e0b324064dd483f456fc
b2d5b123da182c4163b0a45565c501fabe0bb04e
20342 F20101204_AADBLD wilson_c_Page_079.QC.jpg
de55bf2a6ce6533f2c484dc05465f4e2
5658ea6d66ab84f04f9d0cacd1167dbd097b5c63
9080 F20101204_AADBKP wilson_c_Page_100.pro
c646405c3a0b4f454e07aa0f37f098ff
d435ea2b335f54d08c7e2f5ea952f45452147838
64164 F20101204_AADCOG wilson_c_Page_182.jpg
14a7a03fdc25e68c708da6c765e156c6
3cb5003f69b108776d06544fa1b4599b479eb613
36230 F20101204_AADCNS wilson_c_Page_153.jpg
73653c4c42b5f7b8cb47639c09197256
b59b9ce925786d7d71ed790fa20406700212449a
85889 F20101204_AADBLE wilson_c_Page_165.jpg
b868a187062739f04668e125a9a7422b
86661ea7d8c1388bb2313339d4527f75db4ac901
25588 F20101204_AADBKQ wilson_c_Page_189.QC.jpg
032f6504a128b1c4a454949478567c56
0b42abcefec0f5e547a748dca2eb742477acda42
85091 F20101204_AADCOH wilson_c_Page_183.jpg
c0dfaf90ea3bed39c78d0f1f56595142
756325a947ae4dfea5ac0ebaa587630b26f2a825
48929 F20101204_AADCNT wilson_c_Page_154.jpg
b1436f34a80ff6313f3f290783b51ee4
8d8f0df4b99d96f85a8cf73763f9bf7717bc39bf
F20101204_AADBLF wilson_c_Page_138.tif
ca47b5987059465c00366220976ee940
45748d58497f75678f62605ab8f18a90346f1c00
F20101204_AADBKR wilson_c_Page_014.tif
5a28d1499eb20cf59a1748b5955dad6d
46d7bc6b093050f811621e87ad59ac361b05a0c2
86114 F20101204_AADCOI wilson_c_Page_185.jpg
5892b14e8804f779068f579e259e2bdd
72c89d3bef4f116bc0edf531e3b240c6abfd70fd
67369 F20101204_AADCNU wilson_c_Page_157.jpg
82606b85fc6aad04a937de41f6f25504
0d57534ffe5d348408946e3c68cced9e4b5632a9
1899 F20101204_AADBLG wilson_c_Page_094.txt
2821120e21ee1c1f3297d9cf5939aa29
181acc405361dda82b9e4939e3e200f45cc8ffc3
F20101204_AADBKS wilson_c_Page_038.tif
048713df2a555684e5a677988ce23331
182899567c9a5b96817ef54d563942e004866def
81003 F20101204_AADCOJ wilson_c_Page_187.jpg
a985e9376f49b2671b087c1354100bf6
10b3ee0fb31a305d824392b7b2eaef9e891746f3
61379 F20101204_AADCNV wilson_c_Page_159.jpg
ebd3bc554e12008db061cb0c5e04dc1e
abdea02dbb780b7f33fbd1c1fe2d123c59a43de3
19225 F20101204_AADBLH wilson_c_Page_041.QC.jpg
5b73300d6181dc610c32f332ea7de19a
f055d93ae035b1a5ad10aa4d296e37d49a5e8af0
F20101204_AADBKT wilson_c_Page_009.tif
6c06a8428f13106c144942b08854aac6
fc33eb46842030e1712e1fe001daa4fbc9b955b9
34523 F20101204_AADCOK wilson_c_Page_191.jpg
5c31ad64612f65acfc164b8d24f448fb
f035676d435f538fef257ce97675253341d4ac94
75224 F20101204_AADCNW wilson_c_Page_163.jpg
d9af2d40c94b0cc8b4fca03661216760
5efa701a7ec9fdfefe4811c7bb8c7ac9b3ced798
96512 F20101204_AADBLI wilson_c_Page_181.jp2
3b00e3ec3f13c0b2d523b0342ebc6e60
a069f0408a8be28a180ad29a26a39a326463164f
42399 F20101204_AADBKU wilson_c_Page_030.jpg
d420b92e9e6896ca87f5eef19fda52b9
e2a54ebe04c0c9ead1b2f7a6426dc356ced01cca
47647 F20101204_AADCOL wilson_c_Page_192.jpg
6fd355e27b284bc6190bd6378f9a4ddb
d79736503ba505662abe427e64f8719be2213335
F20101204_AADBLJ wilson_c_Page_044.tif
58dbe5f9cf4d779ef7a150da40cc7187
22a729d65ccc63e080f3a4fcb93bc81d1b4dfff3
722537 F20101204_AADCPA wilson_c_Page_046.jp2
1e95cc96b346ca37bcda1128486d3e09
af71e5b439fcbd5c7cd3b438909ea2245ed81fe2
25688 F20101204_AADCOM wilson_c_Page_001.jp2
6ff68b5deb94c0f69cf7c94ed94f8259
bbeadcbed05079bb155e7e28f7321b64797036d9
48239 F20101204_AADCNX wilson_c_Page_164.jpg
61a32224f49dcd3a273c50706f192a0c
e8127ed8131eecc9f2083199bf0cf56041cd81f4
1365 F20101204_AADBLK wilson_c_Page_007.txt
d3654f9894d4178514011dd2a19e5739
effe3af06afb82bf3f908e461d2baca2716b1a8e
4638 F20101204_AADBKV wilson_c_Page_045thm.jpg
6e2188713377de9320d83ebe43c282e7
ec02685f409583f53b40d10aca431f75e61e57dc
110340 F20101204_AADCPB wilson_c_Page_047.jp2
cfbc679ce148e34798165d5f509ed18d
3ca43e30c45bdaffa9856458f973847314c8d130
1051948 F20101204_AADCON wilson_c_Page_007.jp2
3e66e0bdb9c94e3cdc79e1a64a9ff206
e1bfdada8757f3354bcf3c974af516c195f6990a
55628 F20101204_AADCNY wilson_c_Page_169.jpg
15165c81b96e68796c15750d3df1cb79
9a7ebeb815c199e47bbe939a67aca5e565d6556d
111880 F20101204_AADBLL wilson_c_Page_185.jp2
f4912ba16d1eacc38a6433127b18052e
e105912be04e112fd830f329ac7f9375e82ef2b8
367 F20101204_AADBKW wilson_c_Page_159.txt
e56c73f84ddb192b09880d93c78b09cd
7b70e0a26a2ae7c67740683c6b241fc6987ccb98
106614 F20101204_AADCPC wilson_c_Page_048.jp2
a6aa558438c1884fa53ef5a13e843894
4097783f9b8dbcb9aa813c803f2496be343ab9b3
1051981 F20101204_AADCOO wilson_c_Page_010.jp2
0db256663ffa2bb2df625ef752ee76a1
13f0445a5d79e0da6c67e288ea751ebf2dbe3883
53817 F20101204_AADCNZ wilson_c_Page_170.jpg
21f5714e089986e627904b35d7ea95f4
8d060a48fcea5462865352e25679c32744aeaa24
96387 F20101204_AADBLM wilson_c_Page_099.jp2
ad56c9788366c867ae8e2a89b5a458fe
9038b26c356cab781ff8902fa894922f33fb0661
F20101204_AADBKX wilson_c_Page_139.tif
dbd9c213fa13a427f2741b98b7b0ab8d
13fa0472f968662325310f3210bc85f203b93a7d
74131 F20101204_AADBMA wilson_c_Page_099.jpg
104fd204d9a96a8d9dd50beb657c9438
de8f13a56f7557a6e489c091ad2a93f729ae426a
112751 F20101204_AADCPD wilson_c_Page_049.jp2
425c58c57aa4bbd3505f673f6978a365
2e9347309eb0cd032e29986e9013ac6f1ea8ea6d
F20101204_AADCOP wilson_c_Page_011.jp2
93a1f7c1182e8b32c32d665a3385579e
4b45f829207574bd40cebaa32dcf57e1f64909c8
F20101204_AADBLN wilson_c_Page_181.tif
5b528bed72518a071eb0565668ef06ae
66d9c71eaa73a5e95530862bd3ab07b27fabdb3d
63458 F20101204_AADBKY wilson_c_Page_145.jpg
77f2469c5155d4afd963d293999e67cf
c14e4073840b2cc0f372d52d90a7957ac67cdcf4
F20101204_AADBMB wilson_c_Page_005.tif
7912fb3fd45ebf0d9f59cfeb86be072a
5fa7727be097581d0dbf60453199f144208a870d
1051965 F20101204_AADCPE wilson_c_Page_051.jp2
b183eb06532e811ae3968ec5f14b551c
ca93354df7b7f249b69d5ecc6608af1b6cdf2b89
1051986 F20101204_AADCOQ wilson_c_Page_012.jp2
294e6264326b65ee472a5863b8f413bd
6332bb8ad757496d4522eda78a184e33a36f0f10
24007 F20101204_AADBLO wilson_c_Page_119.QC.jpg
22dd2696e887be0ce46372b2f5f83d4f
01cc84e69d2455dcfbc2fdc9b73c2136abfd3a7f
F20101204_AADBKZ wilson_c_Page_184.txt
815120e1dacbf96e4010a4387248abf9
e4caf8ed33161f1c90a7a51bbccc939e8c65924d
41147 F20101204_AADBMC wilson_c_Page_133.pro
e52ef7622a8dfe21c069a9e7bcb4aca6
61edf3597e546a032dcf139adf4b2cb18a730d2c
107055 F20101204_AADCPF wilson_c_Page_052.jp2
21dd5ad5ca2491f4f06e31334d03f996
1a0e588a6090d353c0a15724f7f3bf58a7350e9e
1051980 F20101204_AADCOR wilson_c_Page_016.jp2
4109537f77419031453f08b2528456b4
44b44ddf2032bfca776ec8223c08ed4bf7eb23de
99421 F20101204_AADBLP wilson_c_Page_091.jp2
f5363271aae0fdf42f1211cb6189dbf2
676b15ae077f3e8e2f01a57ba87a142acb28f5aa
5402 F20101204_AADBMD wilson_c_Page_131thm.jpg
ccb5ef6084a47e67a6517958aef7d9cd
6ce679e7ad4757e4cc4dd1f7232fc0d9829d29e9
587554 F20101204_AADCPG wilson_c_Page_053.jp2
a06ad53b6da0334728c1bd22e71f08c4
30366854395de7c39325551ef9985b563e7bd021
104678 F20101204_AADCOS wilson_c_Page_019.jp2
73c17ab80466cd063e717188725a9905
b44bfa7455c89d3597e0bb384332de527ecef5e8
1449 F20101204_AADBLQ wilson_c_Page_111.txt
67bbe6413992a98cc90e5598169bebde
2e56111fc81362152cb419d87d38e85ae3440676
26148 F20101204_AADBME wilson_c_Page_190.QC.jpg
ccc58ece4f015ccd82a1af35b102e9a8
a9954604075be997aa57643d70887bc80ae8493d
106579 F20101204_AADCPH wilson_c_Page_055.jp2
5ac0076f348ff2296595e84e463220a5
158219369274c933850193c9dcc39d7325019f4f
101311 F20101204_AADCOT wilson_c_Page_021.jp2
13e54737d17700ad083a0e42aefa5c53
8052c2ff82a2aa9808d4425e0100957cd2b0b758
5291 F20101204_AADBLR wilson_c_Page_121thm.jpg
2bb63a134c095946c3625cc3af04c7b2
52b1812b683a0f8b059afb7bd612d418375ef625
617323 F20101204_AADBMF wilson_c_Page_058.jp2
bf5231b33be019f93249f39081dc9dd0
17bea6f5c0ce71834f78c8cced92ef0ea222ecad
104249 F20101204_AADCPI wilson_c_Page_057.jp2
5783ceb7d8f16dcfb3dcd0808baea599
ed68a68168cc57fdc13b7a0004c1cb69ab9e18c1
100236 F20101204_AADCOU wilson_c_Page_027.jp2
99c2d1f5f998735a01215617d73f58df
c4a5068e21c2cf497d8f3e59c0e084bf3faa0d42
24992 F20101204_AADBLS wilson_c_Page_171.QC.jpg
732b781e1aa9b7542760d8f0fba6dc1b
ac7c13c890bcd49bf07a3ed0acb76cd91f54f440
F20101204_AADBMG wilson_c_Page_177.tif
910a334c58dd31143fe9d9b972699ae1
01fdb73615f2364d03c83a7ddc2e031136196a91
852543 F20101204_AADCPJ wilson_c_Page_061.jp2
ee4e7a96b0f23a976d09a60395d348d6
4b685f469f58362bcab3f1efb1f5fb32645aa637
108552 F20101204_AADCOV wilson_c_Page_029.jp2
a0e3309ca46f37980b8a79589e2a9048
3119296bf17a9ef1abc9c46beacf54a91db7e84f
28683 F20101204_AADBLT wilson_c_Page_182.pro
7f3e47cef2323b1f9d7c7a77227fc672
215c7072ae86832694e4705d0f527fca835a439e
74567 F20101204_AADBMH wilson_c_Page_028.jpg
632a538284442231325948c5126e8bc5
654c441292726f80ad1d07f51ba8177fb8f93a14
30087 F20101204_AADCPK wilson_c_Page_064.jp2
3dd3744b065ad361d7546d44723488d7
bfb64216b010d52ab4d4c0c2585b17d1611d80c6
53974 F20101204_AADCOW wilson_c_Page_030.jp2
9b25e6e9341ecb2928226397a077d6c6
4861dd390f7f988c18313328b46c8e83ad22ad05
F20101204_AADBLU wilson_c_Page_116.tif
9f8c37ec6e74b9816d7ddb5448913f70
db84b9097dc01ffe68a100de848a7a0509546a2b
14054 F20101204_AADBMI wilson_c_Page_058.QC.jpg
b082e95c7a17e2a8ea8ad252198da962
1dd8a248260f6db25e85eb74463b3e8d6ff82b64
685225 F20101204_AADCPL wilson_c_Page_065.jp2
14e2168cc73a4d7ec30a0331d8063af7
366752df07f1addd1b473a8341844a4680244f63
98910 F20101204_AADCOX wilson_c_Page_032.jp2
0ac7c1839d06861378529549a473219a
2953c453492035c809c186424fda699515008d0d
63605 F20101204_AADBLV wilson_c_Page_025.jpg
70d1b6ca60cdaee3c0a740d6f6a8bde9
be4bb73f5d40b021881a2f91ea8b9f9e798c948d
24803 F20101204_AADBMJ wilson_c_Page_060.QC.jpg
23d59687db1f08249e6da15453230bf6
25374ca055e7f1d6142fbfaff84aed38db4f8228
866398 F20101204_AADCQA wilson_c_Page_097.jp2
4afe4d65fc2abd7618ecfc1388644272
d6bf4f3a974ff8d7e108d7657ec274af37042914
102525 F20101204_AADCPM wilson_c_Page_067.jp2
c34e5fd5dcd7a4f8624c3e14f88d89cb
75981f5a2f274d3e74ca9911661fe94ddbb6fbb3
30228 F20101204_AADBMK wilson_c_Page_015.QC.jpg
805f1f7531adba9cc5d02248f22bd11d
ba1f8eb67fd39f265e453b2efa3bab37dbef3dcf
426130 F20101204_AADCQB wilson_c_Page_101.jp2
64727cfec24fbeaa9767831f62c03fc2
dcddd9709427d201eb123ea453aedbcacb1d940c
90699 F20101204_AADCPN wilson_c_Page_071.jp2
e4b623cdb8086a2f161fb1a355ee11d5
c67b063de85f2eee1d54ef9dc3c1e17780f43b7a
751753 F20101204_AADCOY wilson_c_Page_034.jp2
4902f2896fbdc89e2e18d301621f2f78
6417201f07acd6490d8c38e7cf1cfd6a93cae152
F20101204_AADBLW wilson_c_Page_111.tif
589a50910b9315fec007b4113864291d
80c715a6ebcca1daaf903073f81bd3ba8bfa8145
26728 F20101204_AADBML wilson_c_Page_114.QC.jpg
f8852555b09c1b02e1b3aabbadb70068
f8d3a19678bb1cfd0432f09832c7b2c1bd5f0442
607981 F20101204_AADCQC wilson_c_Page_106.jp2
75d1377c2d16d0eb1c3cf746d01f3e21
1442c9d5411f683c6065cdcbc230a6028e0021ea
795615 F20101204_AADCPO wilson_c_Page_073.jp2
28fb12a3286b86dcdb2211d65ea495fc
41cba1d2d174ead9e6a06f07652a6033ea956191
88725 F20101204_AADCOZ wilson_c_Page_044.jp2
737abdd99a0a3742d703cfb94008171a
b6b06d40c5121727e98a50137e354c51a1a7980c
25373 F20101204_AADBLX wilson_c_Page_035.pro
9536cb2ac612ee0f6c22c06c7aa573dd
e1e7848b503bd8b70903fc741f965f8cce6f8e0e
5879 F20101204_AADBNA wilson_c_Page_116thm.jpg
421f3b7247687c37d5f9f3a3f7eb8386
4325266b53f5ac2e76d6604944ebcb653a519062
63791 F20101204_AADBMM wilson_c_Page_109.jp2
d6f5a4240e711f683272ed2c3bdb5487
222b7bd54c7e2d5915e35cac6ba3c88d1857c486
107314 F20101204_AADCQD wilson_c_Page_107.jp2
6b3909120aae6e419a0bba21fb9e7381
9f1a48f1df75e0374d15431634d135fe5dce94e5
1051889 F20101204_AADCPP wilson_c_Page_076.jp2
ef13a7b1a5961939f96e0e8848454409
d830a1306d7e91fe5389c02ac906d7669ce63a2d
F20101204_AADBLY wilson_c_Page_073.tif
72625845cc261c38e1699abd2f236e06
9a97c45745a56ae72c2cf5be31262795fffe5c40
1740 F20101204_AADBNB wilson_c_Page_063.txt
7f2dd85800f159dd0fe08e873dce7486
99d1cbbf4c05f3faa4ded3746fd8bb669f3dabeb
4485 F20101204_AADBMN wilson_c_Page_058thm.jpg
3d2a712e7dc169c9114abaeefa0ac3c2
4a8b533ebed5d1c4f950fa063543deb1d71a3a6f
801994 F20101204_AADCQE wilson_c_Page_108.jp2
ba51554d6a217f309b40c9d2355ba99e
277fd4f495d7cba8662f3bfad6bd11aee536c640
29099 F20101204_AADCPQ wilson_c_Page_077.jp2
aeb27085d11dd50e0e19f235e7b2387a
b1e85b736da558db0222d94d38a87d708412726d
3896 F20101204_AADBLZ wilson_c_Page_192thm.jpg
32375ac0e56f5416f7e73e54a5f46257
3cccb44114b3302021e54372856e018ad89d89e8
1742 F20101204_AADBNC wilson_c_Page_020.txt
3c5ee83a4248c4476f501bc0f95d79c6
6eb3f32c86aa13646e025b4a29b12e94be4eec21
99241 F20101204_AADBMO wilson_c_Page_082.jp2
5fbe25dbbb7e3e16beed823279ab1556
d2a36bb144b93fbdc03a86ed9d2bfb212ba213a6
76771 F20101204_AADCQF wilson_c_Page_111.jp2
eca5049acbaf9fc431e4764f122aab1d
ac21d27878cbef1b49d5683a811cba2d9fe65cb5
82706 F20101204_AADCPR wilson_c_Page_078.jp2
7805812b505d08e13f6f4ac10bfd6576
bc737bd58a5380458d77ba6a69f60188e56ff2b9
949 F20101204_AADBND wilson_c_Page_154.txt
f88f68a7388ff486680c3b0820f3435b
c7b7dda6463f25890ba62fab6e4de7d4df165ce5
6154 F20101204_AADBMP wilson_c_Page_189thm.jpg
0e54b55726336241c21e43111353bcb9
8c06e1d8f484c75ca7dc89a7d5cc9a16a75832aa
48577 F20101204_AADCQG wilson_c_Page_112.jp2
cd95ec941b949a361def9d17319d5868
6586e8131f6577ab545bc8a86a0736139c12f574
826221 F20101204_AADCPS wilson_c_Page_079.jp2
7fbe766550fcc1234d1dfcd7d7fc3c53
52fb3bc4c7cf48368b4acb32215e5c201375b3d1
45534 F20101204_AADBNE wilson_c_Page_084.pro
1784e1d84daa21c9270680a60547b8c9
ceef8be2ee74b405eda2c777b3a26b9a3ef16d50
6846 F20101204_AADBMQ wilson_c_Page_186.pro
55bc10e9ea0e04cfbc9c397fb9feec35
9cbef04fdbf156de9328b6324da435adeeca5b0b
109543 F20101204_AADCQH wilson_c_Page_114.jp2
7b905c6ccca8e5f6e4e76d400fe42da2
d977e9eceb1e2a1378ac45925461eac6d1272fbf
928315 F20101204_AADCPT wilson_c_Page_080.jp2
11dc91d6b2c17103532340b9b267fb7e
0eb0d582e0b0b63624bbc61463642f4efacda784
46703 F20101204_AADBNF wilson_c_Page_038.jpg
9e26911b07253d784629a34351540dd1
809e3d871802cd26f16392072f98fe9806beb6a9
26640 F20101204_AADBMR wilson_c_Page_151.QC.jpg
25a034b0137b7b8722795ab037011219
52fb4d7d42d6c1703654aa925f5cc6199c4885ed
F20101204_AADCQI wilson_c_Page_116.jp2
404fff950738e3b52961631e93912ebd
fecd93a36ac35b25a06e19c551a36c5f99d7ec1f
780601 F20101204_AADCPU wilson_c_Page_081.jp2
d027809bf9ffd7a5aa21feedf12cc7a3
a730582e8fa214eaeebefc1697d2f565e2b07dcb
42872 F20101204_AADBNG wilson_c_Page_131.pro
46cd683c1bca4ec3e0678660898346ab
8d954805d4ca28dfcf42b61309eaa6a3711e3fdf
24065 F20101204_AADBMS wilson_c_Page_039.QC.jpg
36bb8a6caedda04e4893ab58f75bcd4d
8a8fc67d99612b7e73f002be372275875d277311
93061 F20101204_AADCQJ wilson_c_Page_119.jp2
77c309969694b9fd88e36c1895c2d827
7707c0fef8fae27bee7dcf3aecb9895479ce91eb
94146 F20101204_AADCPV wilson_c_Page_086.jp2
42f0d88b8d203d928a954cbf230c4c16
1ab8738fbd2efb8057a0745a9ff68febdb455128
7042 F20101204_AADBNH wilson_c_Page_056.pro
cda3a7b7b7acbce645e3c96d34294886
b28a2070e3e0e12b7df16623decb4766dd263d89
57644 F20101204_AADBMT wilson_c_Page_095.jpg
ea0289efe0bc2f89f53ae6549c3ca1ff
42f66d2d5256a93ff6468f81233d019db84fcdc2
100953 F20101204_AADCQK wilson_c_Page_120.jp2
2ba9ed262b243bc188372b851b80dc4c
e8735fe04ff0ffe4935f8b45407d2b6f27e1df8b
555296 F20101204_AADCPW wilson_c_Page_087.jp2
d221dce65873bf6a41fbc5237f1523a2
4666f32c3837ee673c94d3d404fd8dfa35ecd5f3
5811 F20101204_AADBNI wilson_c_Page_054thm.jpg
3bdaed1f5f4e2b3013a57d778753b9cb
940236c70d28917662a1d972263b9867ff6d4298
82132 F20101204_AADBMU wilson_c_Page_048.jpg
f568f49f030497f3fe3aac1f82799291
94fbc68047041f93ca0a794047a94c7bbde05a26
97949 F20101204_AADCQL wilson_c_Page_122.jp2
cc65bf3e2435275b922e3d998924045d
5997e4be7dda74c11d6e24fd56135cfad045f6ed
711686 F20101204_AADCPX wilson_c_Page_092.jp2
da3a314a9454f2220503fa3240ff4908
a914b70a075e243900b002d203282bcc472f5033
47845 F20101204_AADBNJ wilson_c_Page_057.pro
81acba6f7ad91871af87b722e34b67a4
2b5a2ef26689586d473ddf9f8938761f84e9d2a3
425 F20101204_AADBMV wilson_c_Page_164.txt
77bbe34f57c07a2e7259b1b0e398d14d
7fa0fc82de29c1b7506287aad8b317c13b13e6c6
1051825 F20101204_AADCRA wilson_c_Page_159.jp2
90e36b3b9ebb0b0b02b2a0390a1ee654
407e61a8a5d9757613e74d9354e6cd42b69c76f4
425207 F20101204_AADCQM wilson_c_Page_123.jp2
40e0cb5ca5c4ec0c706ffcdb82166324
4e92bf912bdc8545563f9495b5622939574e4899
106098 F20101204_AADCPY wilson_c_Page_093.jp2
00b351acee72b9400b4c7a78ebfd7b63
9e1fbef10ffc04ee993ff47f65e8d13586edba53
17754 F20101204_AADBNK wilson_c_Page_035.QC.jpg
aab117086d85073a94710aa6a1215b00
46863d8631bbc0f8673d6c5dfed9ea6ff4bada18
1277 F20101204_AADBMW wilson_c_Page_186thm.jpg
2accee3de645a2cd235cf4c624ae478c
f0d4fbc2f74a91e8bd5ad420cdd3305727f80caf
426947 F20101204_AADCRB wilson_c_Page_162.jp2
ce93966c5712d44bb428d1ada4789a55
0f7d5e0486b9a76a09e8bc36d89470a62113be20
676228 F20101204_AADCQN wilson_c_Page_126.jp2
0c1004e74fda7c73a23c85499d159a03
4818c74545368f3f430888e9820b94f86f2ddc2e
4630 F20101204_AADBNL wilson_c_Page_186.QC.jpg
a74255bdb78ee1a1a55ea06462199b0e
800b60389663b8e909f19ed394cf5c19484a6f8f
95010 F20101204_AADCRC wilson_c_Page_163.jp2
43c9c626d3414e173abfc4e0f28ebf90
c07ab1ba2c0905d89b014c7ff1aaa770d01e90e1
94053 F20101204_AADCQO wilson_c_Page_131.jp2
cac75cff6ea850519f4e0867f3ce4d23
b3b01f433b652823bd2b149762fe2749bc07c278
1051927 F20101204_AADCPZ wilson_c_Page_096.jp2
483f90dd29692a8ecb1f115d885485a3
9a23c4825fb18d30252bef00bcbca50d3ce0e4a9
11687 F20101204_AADBOA wilson_c_Page_143.QC.jpg
94ab1523b1fb946defbde94cbe4aa47d
2c99d4ded9a0d9eb45ca9b4297c15965b2fdb0fc
25226 F20101204_AADBNM wilson_c_Page_102.QC.jpg
968407960e1baddc2134a174ddbfc4f2
14d75bdbb1f7f6f34b7a6ea1b1e0e4893567dcb0
24337 F20101204_AADBMX wilson_c_Page_022.QC.jpg
77d7a4a810bac91fcf83a321a85e1c20
5e4d43d9f5a95d91f91837f5973962e79cb67805
987136 F20101204_AADCRD wilson_c_Page_164.jp2
2bdb76421e287bcf55ff7a59ee39d719
16aa336d78603bb2ecdcf30657eebd46c6137ed4
94842 F20101204_AADCQP wilson_c_Page_132.jp2
1d08de8f83b7501e3944beb6240e3940
0b27bf6b96769c0cd8fffbce79414a9ee10b81b6
6023 F20101204_AADBOB wilson_c_Page_094thm.jpg
6b4233e6a4083706a6dc0a3bfbb82ee2
0c987e50d4c22990f36e30aed38727d695e42d85
5547 F20101204_AADBNN wilson_c_Page_063thm.jpg
5c1ca0c63ee4918102f44c1debbb3a0a
c199a91d0cd3a32da6407e3525a388de16a0c092
534986 F20101204_AADBMY wilson_c_Page_104.jp2
8a7da0a50cd1405d406d0edf1cdf63a6
cca0761b2d69afb7c0ad700b3bf4e7539218c07d
1051946 F20101204_AADCRE wilson_c_Page_165.jp2
4a3104b416b5f9746fc7da646b112506
e21304318e453e8ea55c783eeef8bef345b05a68
102699 F20101204_AADCQQ wilson_c_Page_134.jp2
3b629e1213741ec8511004e29931eb13
b936a3f3d60552d210311bbb2dc2b394b77e98b0
4011 F20101204_AADBOC wilson_c_Page_153thm.jpg
276d705b067ce8ce7c0fef71b399100a
fd076bb06a66c55743236d3841a6c0b14dfca55b
16280 F20101204_AADBNO wilson_c_Page_034.QC.jpg
540a95e62ceb2b3b14bcb7713d8a7b39
072225bc350091fbdd243cccb28eccf975b018f1
F20101204_AADBMZ wilson_c_Page_010.tif
7590d0484839317a6949ffdd14a55a53
c5571b4e85792ffef8ace7ff0d92752a33e6c92c
1051917 F20101204_AADCRF wilson_c_Page_167.jp2
bf936ff3e04085cfdff8705b1eac1ecd
75a88e93052aec41b4704619560243741e656cce
949202 F20101204_AADCQR wilson_c_Page_138.jp2
95066c13b969c784fb232e161909a1f0
8d0e4fcf59e58095d62ff23bab6dbbad917523f1
6029 F20101204_AADBOD wilson_c_Page_066thm.jpg
f43194abb3f067ae740b5edcd9326125
144c03c11e105902aef44c8fdfee608b2d30841d
1660 F20101204_AADBNP wilson_c_Page_080.txt
db2ff8689b90bac814bf699cf59d26c7
f26ba2071a627e3aa0a3180e5c09c57c3c5affe1
1051879 F20101204_AADCRG wilson_c_Page_169.jp2
48086eff6fac0ac1a283b977f696810e
26a2f02e08bd905c598d3f6ef3a13c37bdd3400b
402069 F20101204_AADCQS wilson_c_Page_140.jp2
8fd778825cff154386ba8262c4468e07
e230a14132dfc7f2b414ba5efcc6d97a463e25b4
24179 F20101204_AADBOE wilson_c_Page_152.QC.jpg
3ede521221925acc083a133b09451c98
ef75c7d4c33e947fc5e5f53f90d1094c25be3c77
550 F20101204_AADBNQ wilson_c_Page_167.txt
bb828d34b81fbed837b07bd3ce054ceb
a46489223af44fd258ef0c0407bdbfe8eeb51089
910131 F20101204_AADCRH wilson_c_Page_170.jp2
74537ed7d56d3f721e0c8dd63d7d324c
5b2d965f3028ae7527930f8314fb111a9335454e
512329 F20101204_AADCQT wilson_c_Page_143.jp2
b55658e4772bd20c3f8e63bea6c88240
7c6b84c8e9f624d0318154101cae3436a9c55282
1821 F20101204_AADBOF wilson_c_Page_166.txt
5df0db8c8e6fa1de08e7df9599409acd
e495aedb9f853ed27e67fcf0e84c7ada3eb4a701
57932 F20101204_AADBNR wilson_c_Page_089.jpg
c0a9110c0244736a9656ecc5e46a143d
259a1227a80bfbffb2ed703bcad70b96b2f72fdc
105887 F20101204_AADCRI wilson_c_Page_171.jp2
be17f8a15cbbfc76f39c365b1007b215
85f1fc779d9fc9f1facaec6b87c561a60bf1dce8
105108 F20101204_AADCQU wilson_c_Page_146.jp2
c7e8d74b2225f02ef53cbb55f68b6f7d
f3e59dc3457629f13a989e5a2ba1bdbfeadf103e
70622 F20101204_AADBNS wilson_c_Page_098.jpg
f269e3886577c563b5fd56ec927b5011
282ba8d3eecd48b90bb08cef09511dd395a7cef8
32101 F20101204_AADBOG wilson_c_Page_128.jpg
f82c3698c3333cc303e88178e91338ee
f8f303b1b29c8db3cbab11e02a4cfe7d1f65cb72
105970 F20101204_AADCRJ wilson_c_Page_172.jp2
5de934b0cec9e7ede8c02f4e8dd20818
a6b2abd2b4ecd793995a93df56a3695326b1d10c
111226 F20101204_AADCQV wilson_c_Page_151.jp2
de21b02aa82f5ada291182bdf7a76b79
defb6da3c9af77756a495e215d57eebdb1dda4ab
89639 F20101204_AADBNT wilson_c_Page_020.jp2
eb086ba82f9c5d8bb5b50f13bf4fb58a
d6b32191ac33196fcdaa18364a665b765801ea67
23316 F20101204_AADBOH wilson_c_Page_010.QC.jpg
e187acf255b1a9502b1e7ec5a573548f
ada75fdf89c5e645eb0444faa78822a9ac5e8bbf
803573 F20101204_AADCRK wilson_c_Page_173.jp2
031a3c43c3ee461cf5a404e7f7b222de
551460846031ed230bf8cf54bb69813199840374
102189 F20101204_AADCQW wilson_c_Page_152.jp2
01e926d969d9ac0728a6550735f3911e
00ab0386ae886e10720da23d0c85b6c56084f0a5
65321 F20101204_AADBNU wilson_c_Page_155.jpg
966bb75ddfaf3fae0aca08f38a9c1229
2d9add485247f7c002f4615adfe0349b76c56ad5
82283 F20101204_AADBOI wilson_c_Page_052.jpg
12c486b7b187cff7ec89b3ce9558f796
10ccbbf12982ac0b019336683d2a671629b41810
540136 F20101204_AADCRL wilson_c_Page_177.jp2
c94ab5c41715a7953316f82b864af638
e694faf7ba956ae660ccb9d7a93257cc687b4b9a
623656 F20101204_AADCQX wilson_c_Page_154.jp2
651fcbc46b84a498837ace2ccf2320e1
4611f414182a59d7c9e1026f6e8548b7eed53dc7
25216 F20101204_AADBNV wilson_c_Page_052.QC.jpg
6b758ad70e48d2f9233995a3d1b014db
814923cbc6b63936e2cef1788c8a34f2d529c56d
8423998 F20101204_AADBOJ wilson_c_Page_157.tif
27cd798e5c8da3716ed48b9a727df71e
c0e0e85775c911ddedec35a27b88d1fe6d397cd6
F20101204_AADCSA wilson_c_Page_023.tif
e09d6cbca44e74a7a4876b8e1ab55dc9
3b961b0a498b87ad60974e72e1a06bf540bcc0f5
F20101204_AADCRM wilson_c_Page_178.jp2
d29af782d7f678161698c9e30ad982ed
b8426a6d4293d5dd24bc3915a4e6d2e10dfec1c8
953363 F20101204_AADCQY wilson_c_Page_155.jp2
fcdadb79347d848bcde3f635a2458663
7268b3c8320ff86fc98d8a2a71d0a334fb212e8a
81031 F20101204_AADBNW wilson_c_Page_008.jpg
ee37a03775c4b980e39e42213c67cff1
72bc533082a1844c37b4772a8d7aba38cb9f235d
5545 F20101204_AADBOK wilson_c_Page_150thm.jpg
29b08ff4add916650636a3c0e2d2d8c5
af6e93053bf38bfaf6a7465a1b2e89f0f7c91801
F20101204_AADCSB wilson_c_Page_024.tif
a212239bf761656e95f5a90a19e26a17
b710be43f7bbb2f1b4c1efd3f517fa1051ddb33b
109094 F20101204_AADCRN wilson_c_Page_179.jp2
30b05bff80d05cd9a78ae81cfbb1610d
7f611f9cef0ee628c3b8b59cbed5b79f332e383e
F20101204_AADCQZ wilson_c_Page_156.jp2
0105f180ad57fce0aae2d1cea265e59d
e4dd4b606861707738e1f0a5f5c0a77b69d06bd2
F20101204_AADBNX wilson_c_Page_051.tif
88ba1782b03e6c515824d8aedd5c94ee
5c2af0a7dd4e116edfaa601f84d09c90caaa93f1
4078 F20101204_AADBOL wilson_c_Page_040thm.jpg
9c4a495d88c57978e6184445aa8f45b1
ec12e974d2215f72f4a5a43b02abfb61a8233817
F20101204_AADCSC wilson_c_Page_026.tif
03676671a9c98cdb09c8f2f3953cb07c
21f56a9a0f487ee39be2bff6d3afb40ce3d0ee61
17727 F20101204_AADCRO wilson_c_Page_186.jp2
8f6aebc5247dc6b1c86d13fc937da3c6
970ff4ecac77ea2172c90b0aa0b50e6abff15aac
F20101204_AADBPA wilson_c_Page_192.tif
e6ec04de2c90d32927a2aefa01e586ec
3a64b52263aa521e2d48001701232d0c902d5dff
21157 F20101204_AADBOM wilson_c_Page_025.QC.jpg
c0af91c0d08d080e0fd71a3abe799338
c7feaf968e6661a4ad27d655063360bff4ac4540
F20101204_AADCSD wilson_c_Page_027.tif
dc0d2da9903ffbdb19bbeded66d9e54a
c40f61937c69713763063db0cfba1c5f98b15682
115463 F20101204_AADCRP wilson_c_Page_188.jp2
917a99b4774d550b784f4b6bd2090d59
e709462a7399751aa8f66402bd2bb1161ba50b61
16221 F20101204_AADBNY wilson_c_Page_087.QC.jpg
4fee600032f115aef5f02fb0ffc2e6ca
6cf2be31a317a0aacfe1eebcab6476b3cb41e9b3
28826 F20101204_AADBPB wilson_c_Page_105.jpg
2ac2755a6197d314b68802d056c0f7e1
436af5695e40f6df310aab72179d5d3f0c00dafe
5013 F20101204_AADBON wilson_c_Page_157thm.jpg
6977247294a93c06567459163b117120
bd739de8ecbfd4a1d8391a2f683f1a1d9dd49a16
F20101204_AADCSE wilson_c_Page_029.tif
66910ce9add02fad66e7171eb7b29f51
2cbdda3f6b8c851b382e30f570422159d1ab0103
122028 F20101204_AADCRQ wilson_c_Page_189.jp2
c04095ace4db0973208ea13e94bdf71a
0a6db750771576498a2ff31cd71ece8afcf0a835
47856 F20101204_AADBNZ wilson_c_Page_103.pro
17c323547dfecc0fb194ca28dda2355e
c33b51ac9b67bd0d890af34d362e8f78a32a97bd
16183 F20101204_AADBPC wilson_c_Page_045.QC.jpg
419c24435111db210ef14d57751c5102
ecb7b9dff116d1b313d8f6cc25bdb77e4ff2e61c
17601 F20101204_AADBOO wilson_c_Page_159.QC.jpg
a1f874824d796b05ed14dff7e496e209
8928730d51f9706c474d5db8c3a1223d5cba2f5c
F20101204_AADCSF wilson_c_Page_034.tif
bd06852017d87584950ecc6f71ab6e3d
b0bfe7b0c36b4e3342f5dc66b167881bfbd4f5fa
123734 F20101204_AADCRR wilson_c_Page_190.jp2
896c1e4e9cb565529a117063576a7222
58cd603bc650c30cfda8443863df6102da78993f
1031385 F20101204_AADBPD wilson_c_Page_075.jp2
59a6ee0697d56f67cf49353c1a7bab87
839496aaaf207c85bdd9699ab73a0b02802897eb
79973 F20101204_AADBOP wilson_c_Page_102.jpg
7998b5925d8371e90438746c25017cf3
fa9d8c299be8742a0d317e547155f4bac57f3419
F20101204_AADCSG wilson_c_Page_035.tif
87a10399db8b2f61579947f59c7ec50e
911cd6d30a47527d95c0aa6ca77ae56de8990d65
60492 F20101204_AADCRS wilson_c_Page_192.jp2
a303a548a1c89f298d7a3f54c2ec2622
1eaee9bd61a9a108a818b480c2a126771551903a
F20101204_AADBPE wilson_c_Page_087.tif
380bbe6a6f4d9d4d653e6646c0ef96c0
ecc3c257d714d1596ed1e778955d74a2faa6ffc2
5898 F20101204_AADBOQ wilson_c_Page_085thm.jpg
aa76efa05099c394dfcd272da3f6290c
b23d77d31386091f935fcef96ce7a0053fd3522f
F20101204_AADCSH wilson_c_Page_039.tif
69aec7faa1d26c2aa12c681b519786e8
1382652c383ab20f7e0c05ba4f005dd8d7bcb388
F20101204_AADCRT wilson_c_Page_003.tif
72a070dcf1323f82c2c67df665a19915
6c13bc1ab18ab20016485174d1d37dcb95ed069a
22660 F20101204_AADBPF wilson_c_Page_117.QC.jpg
fc5f646b53848d251f982b0c964fb35b
61436603e9415ad6da6bb4c1fdd003a7fe3a7e43
22998 F20101204_AADBOR wilson_c_Page_001.jpg
cf468ee60d2b56f12470fe89c5d1e962
5bc1616b9d14a41c6a66c35f4555f5cbffea6171
F20101204_AADCSI wilson_c_Page_043.tif
7248cc6f23560ce6c96e92bb9c1fe25e
78c6a48310fd1812db0e4ca3e3d4b5d8a6cc5e27
F20101204_AADCRU wilson_c_Page_004.tif
176d6e3bd3a2a7c15a4c9dcb5e793f4f
f3b9bafdc5166b3c8ad5934b1f1d21b37167768a
24557 F20101204_AADBPG wilson_c_Page_067.QC.jpg
7ae08e277381b89b2561903327c6f53e
63064d06f3d391959151d336a7fdd6817df94448
212 F20101204_AADBOS wilson_c_Page_003.txt
52d1b5c687f959ee79cc5e2663c746b0
2e6652e7d94ce4fab39d2b556cdfec845c381e92
F20101204_AADCSJ wilson_c_Page_046.tif
2ea1b8d15b8d412604139d8cd974add6
e265df8e5cdf5107201502e5ff1802e4512a3e40
F20101204_AADCRV wilson_c_Page_011.tif
f129d3a73186c6252615f46c69cbdc62
0ed8a44a407d73006373bc4819203685feba7721
41622 F20101204_AADBPH wilson_c_Page_110.pro
87e84c562e44ae4554346fa1bac51625
50f1b33eee2a1cedb4d7c7eb7b41f391a390256d
6406 F20101204_AADBOT wilson_c_Page_052thm.jpg
f3cdea06ea34208aabb76ecb81e0ce23
f4f5af0f3754348f6fda61c297d2a9ceb5e51e89
F20101204_AADCSK wilson_c_Page_049.tif
e6a5c9e3575bec1e5e18bb46bab5df5d
0da432d846d3e187d4dc2251093098c2f51ad2f6
F20101204_AADCRW wilson_c_Page_012.tif
5f7245224de588fb1b517040d1fc9a2a
ed95aeca632b7ece3382a2917cc1ec75d841cbd0
510 F20101204_AADBPI wilson_c_Page_170.txt
04d1846b53e0dd1451a206b1b9c46c04
e59aa2d8d2168434ca6cc844c72cb903264d20bf
89098 F20101204_AADBOU wilson_c_Page_023.jp2
cb36094bb5fadcb4ef94a65254e59767
0717b89afa7232e8e28d382a33252760962ce901
F20101204_AADCSL wilson_c_Page_050.tif
cb65645442453c3cba044cb90183972d
76852a5bbc4abaf80da04f7a462f758383419558
F20101204_AADCRX wilson_c_Page_013.tif
8d082d345c01c54bffc1228011d07c9d
9bec7d62ac736c487b0e8cad3576100630432753
82126 F20101204_AADBPJ wilson_c_Page_042.jp2
b74a9f0dc6e981ab332624bb3002dade
ed1f7845a3e47b4c68094201e54939e2b104a5fe
48409 F20101204_AADBOV wilson_c_Page_022.pro
a0a4d45aea5cdfbe33b3026ce2d8a10c
c14129233d3fcfcc9d03db218b9c413802e8ac74
F20101204_AADCTA wilson_c_Page_086.tif
ad5c82c984dc85dfcad6dae68ce0a5e5
9ad6d653cddf71993335460db78ba5567000abc4
F20101204_AADCSM wilson_c_Page_061.tif
47b44d8b9be1f7634876f5df76d99401
77221c732c308b4d4d86bb94b7f3d0efd57f5d19
F20101204_AADCRY wilson_c_Page_019.tif
622d8103f585b8d469639fc4a3bee3f6
69428ff25be5ad7402fb88d12f8ebbd09e4d11f7
11704 F20101204_AADBPK wilson_c_Page_129.QC.jpg
c6996d3238c246553f7455637b57924d
b7ec380a06f40b2187cc247e55340ba23b9e5fd6
410 F20101204_AADBOW wilson_c_Page_105.txt
8f578768a9f65574e26b4d2491a98c67
33030fcf1f5804c343e3c0bfae38bc6c5527aaf9
F20101204_AADCTB wilson_c_Page_088.tif
0f92cfc8f11be4707343d1c04827a383
362a80f2ce85000559b5c7b3d13282e763503399
F20101204_AADCSN wilson_c_Page_062.tif
c111d7e93f3293eebe8362fbc168d0f0
6e0a1f2ff926c9828eadda13872a956c7a08aff7
F20101204_AADCRZ wilson_c_Page_021.tif
6f89f5eb82094305f2d39b161cf592bf
023e81a80d09719d4be31b6651f3dd9aefc536ea
76905 F20101204_AADBPL wilson_c_Page_027.jpg
ac1dfdb1919c56a7d7a5153f16766f64
46b10b001ba4cadd63475336b14f58e4e15fefc8
82109 F20101204_AADBOX wilson_c_Page_158.jp2
6cd30cadba58f6d7a220f22d7a50259d
c6302c244390fad4612aa5d2eec1b4b1e0c04bb9
F20101204_AADCTC wilson_c_Page_089.tif
b2dfcc5806cc2a4774b07d8ed949e10a
d25c40c88ed760767bf6a40eb7942752de8b2417
F20101204_AADCSO wilson_c_Page_064.tif
d92997f98600c172635e623d4bdcc611
d5c2287f170844da56be5013af0c3d1c5118aeb5
49432 F20101204_AADBQA wilson_c_Page_139.pro
1b2e0a9e59ced1bd11d26bf694c98499
6cda80b7566927e6b9f08f78341ec315df7a7eac
F20101204_AADBPM wilson_c_Page_053.tif
34d83262e8eec466d6cb7a38e6bb0db0
a4e2fd28241fe2ef9d6fcdd94eb13e1f0aa7c073
41187 F20101204_AADBOY wilson_c_Page_136.jpg
63e9b2a594dab29330f975fd4776bd27
a711523cca72b5eb265030b766d4cb90bd3c4436
F20101204_AADCTD wilson_c_Page_090.tif
46037c31e371818c6ed27ece1f7a0e7a
a4682773122c5dfca5c0a9a50f03f7a6267c4f34
F20101204_AADCSP wilson_c_Page_066.tif
5fb555b853e6039d33f4b99ea8463557
2f57582dfc22f9c9fadc5c80002e2dd7bcbfd81f
73320 F20101204_AADBQB wilson_c_Page_063.jpg
6584a03455baf49932496e1d992f3b76
70ad875ed13b1ecc002d28141e3dde1b7a1f0422
50851 F20101204_AADBPN wilson_c_Page_045.jpg
feaf7d6f4f12e5f89bfeb395415bf3a5
4574fae06cf2f229fccaca52d1259fbc202e9731
F20101204_AADCTE wilson_c_Page_091.tif
7bbdfebd8e8f7b564c746c4f87b7af9a
d10137abed34b11032f8d7695b64d4e3fa16803b
F20101204_AADCSQ wilson_c_Page_069.tif
cfdd515e093c4ee47ec51ec0c606dde7
6dbdda886480a5fd2e05365a26b2bb350767f380
2000 F20101204_AADBQC wilson_c_Page_029.txt
c96bd64edde403fae5a516d26ff96cd5
4939b52269d80bb211226c9eb521a5279a5439b3
F20101204_AADBPO wilson_c_Page_036.tif
27236b6437981de0ca862c33aa355bdd
87e9a19003ba64a7a6c3d7b71e215b0d8f832105
F20101204_AADBOZ wilson_c_Page_101.tif
fae54ee68b54c7fde1d1571658f45d4e
666f5aad1ab068b61e19a0313884ac7275e82a91
F20101204_AADCTF wilson_c_Page_094.tif
054c954567d1daa9552a910b649314d0
ec696e0d0233d57ed7fa45db12b1f4182c418547
F20101204_AADCSR wilson_c_Page_070.tif
0a549fc886781db672df7d8cab8f2c4d
88d063741fa8c974c702e921b64954c3c6570148
1051823 F20101204_AADBQD wilson_c_Page_180.jp2
4211c65b5766af6906d53975bacdd16f
c670fa4cb23897a8ae03cf8a55faf6a595961c3c
79335 F20101204_AADBPP wilson_c_Page_134.jpg
d8764f64d8907d9987425a87ede042a5
5bd44a058b28da59897fdd9e4307d83095cabc4b
F20101204_AADCTG wilson_c_Page_095.tif
cee73e1ad622e28d99128d473c1d9bf8
56e0040d9a33d9ea74cb977a3591a2c19741668c
F20101204_AADCSS wilson_c_Page_071.tif
f06db8c041a6f21adbd6c514e8bfd4a3
e4f4a4aae87e22eef3ae7229e958065728f22e7a
F20101204_AADBQE wilson_c_Page_178thm.jpg
ddd3fefabeca462e30a0c05732debb6c
283a018a4f2753f1e420c8532545265d574d9ef7
6166 F20101204_AADBPQ wilson_c_Page_107thm.jpg
761e7e1569d01386869077ca4ad8f2f9
bbaa1c4b1998d55ce7044c3c40b255646894569c
F20101204_AADCTH wilson_c_Page_096.tif
8f1546876f1ccfc590fd57fcc434082d
4f8652d169c76e629bae2cfded0ed617cd5ccd20
F20101204_AADCST wilson_c_Page_076.tif
946129469c94bd1408900182e8cbf57b
b057bd2e9ed47c9bbcc05ac4ac2e11916afb0b41
5152 F20101204_AADBQF wilson_c_Page_081thm.jpg
f4b2e8bb481a932d612b4f70c1b030f1
71cf7f0fe5cb9e41d4c05a40285555363b4ed888
403778 F20101204_AADBPR wilson_c_Page_100.jp2
a75e1eb54ac8fc710418dd5d15411b23
b0891ead1fd930eef89500e841a235f8d01df7c4
F20101204_AADCTI wilson_c_Page_097.tif
f92183a5d31090e0e14386ccc9f9a196
c84191ed8c71141842ab701e0f0366449a69943d
F20101204_AADCSU wilson_c_Page_080.tif
0d5aedcdec4f61d8bb0a0f0d24f51088
5ea7e80b43d9c652e1084427944e86bc8bb3ab57
24830 F20101204_AADBQG wilson_c_Page_036.QC.jpg
013d4c48ee26fa755d01a5fd6865953b
8ddbefd79ced041ddd7d4aebc7689f6f8cf224f6
75392 F20101204_AADBPS wilson_c_Page_122.jpg
6fa4407bf4a81848733e1f015b824c7a
2056e556ce4efc25c35799416cacd84a3906518a
F20101204_AADCTJ wilson_c_Page_098.tif
2395e374c50cdf2b36f775890411da30
bee662ce668ff9f0b589c29289deafcf26511904
F20101204_AADCSV wilson_c_Page_081.tif
e6a173ab55df0a451d97227eceae2800
db86c4a1aec87a8d0d7e4bf1cf4735cb77911d78
24952 F20101204_AADBQH wilson_c_Page_048.QC.jpg
3c15b6ad8b0dece50251b52096946e64
108c7fa5bafca8129e19bbdc461fd568a61efc07
F20101204_AADBPT wilson_c_Page_189.tif
8e7bf568c403e24f0e346366eafecfc9
d57acd4cc900372f7f32cf430a4f8da42336b04a
F20101204_AADCTK wilson_c_Page_100.tif
c7df4f3825f997af771ef9c470047b89
f6fbb98f77baa0f47111ac56995fb683f9ab232c
F20101204_AADCSW wilson_c_Page_082.tif
2f12da2323fc862776abc0b375dd71f6
1c324ed7cd6e334d8fa25e30590e7498851590b3
F20101204_AADBQI wilson_c_Page_106.txt
b3b3a8d40b10470297d48ff0bb81ac10
4c2638a62f02372953508a800235872984afbd9b
77642 F20101204_AADBPU wilson_c_Page_033.jpg
2e1c86dd0d9ca6499127bd895bbcdb03
681080e94f1bf9c50d235a681c2535c215818ae2
F20101204_AADCTL wilson_c_Page_102.tif
324f28a812e18fc54b76825a6fcf3997
bb6861d114b40344cd4175ca3a2d3b86ffa17302
F20101204_AADCSX wilson_c_Page_083.tif
1cea4eeb75b82400f08d3665c575c401
fea39fc96383337ad0d054e5b7f65035969cad07
968080 F20101204_AADBQJ wilson_c_Page_066.jp2
15ec6f1d6f8e9317271b2d683cbb669d
a021576012a9970fae7c9082c722d22d44836c5b
549155 F20101204_AADBPV wilson_c_Page_136.jp2
4d1ae1cb98f71d3aa0ebcf71e9a72b50
9a380b129da4cceec402d7e4e9df52c2b95729d8
F20101204_AADCUA wilson_c_Page_131.tif
697921b5b7656c25fe06223e439a272d
6b95be63dcac02904b65a968370137594de8bc4c
F20101204_AADCTM wilson_c_Page_104.tif
2f3e953d463214b87c11db12951770e8
e60d6ce3e8ed507f8d9c4b470dcdab15e2ca791c
F20101204_AADCSY wilson_c_Page_084.tif
cffda3d9bc719b636beafc524ce44e63
81b1a787b283819d283bc7c4fc058c08d2e3fa88
10281 F20101204_AADBQK wilson_c_Page_003.jp2
66e237f9bcac14e8705e0b4a7e98084b
554f0ebb694798636ad93168d2e0f2bd3f1cd993
3751 F20101204_AADBPW wilson_c_Page_162thm.jpg
70ef00b755234bfe83f778cd77c36827
0048dc081a83ac8eb1f16fcf70214ea04222609a
F20101204_AADCUB wilson_c_Page_132.tif
7644044b07b5351874ffa2c889b2c985
e20d95847ae9880bda4d36545efa10bf5645a8a4
F20101204_AADCTN wilson_c_Page_107.tif
9d8ce3a9e0118264b640cdb696ec0e71
3115875512fd07c1b68658e4d7a9e9db0807edaa
F20101204_AADCSZ wilson_c_Page_085.tif
3d80b019f5378d05145156f187babe55
9731745f5bde4e7797640b29ac5ed8e72d6870c3
5954 F20101204_AADBQL wilson_c_Page_108thm.jpg
38db85d5df8be54fbab71890a0fc21d1
4ae934be4234921cbd731304255ba50162f8da28
48485 F20101204_AADBPX wilson_c_Page_068.pro
fb986c5d4fd9540746f0a6ad9821c8b6
268c5a2c27e37bcf854c04654fbf38fb2eb9f7c1
F20101204_AADCUC wilson_c_Page_133.tif
8ab564b25db1405228996c05a408ef21
09e7872a50d88487b4ae298cc95fa590c7ec0ae6
F20101204_AADCTO wilson_c_Page_108.tif
33c20d3e540ef501d2faa4beedbf7f8a
23a1226e247ba685102b86ab7a7f84a10f729cf2
6153 F20101204_AADBQM wilson_c_Page_146thm.jpg
a8580ec1cb7b21cea836a08e5fd16eb8
a2553b49020a7060f7e2f72c5526e220252e8cee
50007 F20101204_AADBPY wilson_c_Page_114.pro
9686884cbfa6f315a47f5fc76dd353df
19a6e73ec860e7e4235b6aae5c49f9cf1f9bb24a
36714 F20101204_AADBRA wilson_c_Page_042.pro
13eac4e62d3938b4b4b962a6fb461eb5
5a850499c0833225ff3d76e175cb466007458b2a
F20101204_AADCUD wilson_c_Page_134.tif
bbbafe90a0c9421042aafc00411eeb19
9ae90acdc421b8399f58433d6d7d387c3a69145e
F20101204_AADCTP wilson_c_Page_110.tif
44b6c405fa273257b1d0d91cb68b6b1c
806fc95a4cd73f4c760c1d9a5f0326416de676ac
1972 F20101204_AADBQN wilson_c_Page_052.txt
271f784c482b48e290fc2e9c633ad743
67e47bfb0221c3016e77b68d529f65625a0c6f66
19410 F20101204_AADBPZ wilson_c_Page_173.QC.jpg
903678711d72c11e19422905a783bae3
e926470a45c3a07267192d0318bf2cc314c2bfb2
29417 F20101204_AADBRB wilson_c_Page_061.pro
e2871a9a9eb4fc93728fe9bcb9f2d27e
bcf9055723094374a742a4832a69b476c5a4b74b
F20101204_AADCUE wilson_c_Page_136.tif
cbfc44683b27a14511f656f42170a199
bc5072d1f18b1d4bcab605562187d62b4ee0df62
F20101204_AADCTQ wilson_c_Page_112.tif
ba607244254480107d7afdef0d664ab2
443765930cb3e3005db5ec2cbc22a13cb99a1349
F20101204_AADBQO wilson_c_Page_056.tif
71189cb8ed0d51f66a8c3f168d67275b
9afa4a5f3744e2a2ddfbf5ffddc68bc6ba60384e
43920 F20101204_AADBRC wilson_c_Page_181.pro
7d654dd82d9c61c2f48fc6512ed3fa1f
76c259a36b3cca8b38430128bc06a6ccb434f23a
1828 F20101204_AADDAA wilson_c_Page_083.txt
33bbe0ae5f71244567264e94e4db3028
871735da4847ff2bc25e2d8edc7551005d1af1d0
F20101204_AADCUF wilson_c_Page_140.tif
db6e360ff58a973d33395e8fc354a483
7bee9b5676740b164aac3714c47466bbc7832f91
F20101204_AADCTR wilson_c_Page_113.tif
b21eef31ff4e20c012e04276fd3a7d8d
b851a4d59307b5d40d2d530ab6fa95613dde5003
53271 F20101204_AADBQP wilson_c_Page_065.jpg
5edf12409cfd2aae3eea381e7b4a3077
14bcfa10324e298036495383d0cb336f0138f3a2
4069 F20101204_AADBRD wilson_c_Page_177thm.jpg
1dda7c8573d9e2b064b4405cbd7ff262
b57337b0ac704cce81517eaea484edaba9471275
1788 F20101204_AADDAB wilson_c_Page_085.txt
39ae23ff3d36b39aa839332ac11e87b3
47f9562606ec8f0f84c6ff7a714e07724caa3b78
F20101204_AADCUG wilson_c_Page_141.tif
96f28751e6a910259ee24732e2685282
737b02e25f0e539a372055d3d4154b0967fcd6d2
F20101204_AADCTS wilson_c_Page_114.tif
a50d9e80e81355027c02806d173f26af
2603fe4a27a29e0862b32b88dbb37623f8043ef0
1589 F20101204_AADBQQ wilson_c_Page_156.txt
96e5c4336be2b76e30cb68093451a276
ad6624d211691f751deb8ce03e943ea2633a887d
1456 F20101204_AADBRE wilson_c_Page_073.txt
00bc1b1f1014ae2686df31287b333a2d
8d1042ad2290509e1c03ed987fd1d1cfa5eb12b6
818 F20101204_AADDAC wilson_c_Page_087.txt
b71cb2fa1fc5dd66d65f38f192dddad5
16f6ca3317f244443bbdeba645f7510e3bf805e1
F20101204_AADCUH wilson_c_Page_142.tif
65d370a8eaf4b51c0ff43f630055d0b8
1d17c15c3677aeb444a85e7f76236265119087b4
F20101204_AADCTT wilson_c_Page_117.tif
084c9b731947611bc3c0ad838fe04231
f1ae4a9dba36a54358ac99857f4511041464ef46
108631 F20101204_AADBQR wilson_c_Page_013.jpg
443e7031e1394cbe8d40833f8e50b93d
47cc0753ae07d3b4c954f7872ff2e76607387d08
F20101204_AADBRF wilson_c_Page_008thm.jpg
faed81a063072210e2c27b7c6cb5944e
36d060f76dd78d479e0e068fb6a9114f7be8e753
897 F20101204_AADDAD wilson_c_Page_089.txt
29589b13170c3f3e5a2ea5b02f40c016
5220ada65e21b2327912db240b0242c779db174b
F20101204_AADCUI wilson_c_Page_143.tif
a9ddacb4f3f19d1a7fcb5c3504c4c5ea
7451b56059071943c0213958848f81b424d475d0
F20101204_AADCTU wilson_c_Page_119.tif
4c559ff459a6daee7810f05de5a9abe3
32b901105075417fa0bcd56859df5d22abbb33fd
55215 F20101204_AADBQS wilson_c_Page_046.jpg
726bd9ed14ce19e50298677af66e82c8
bb55efddf72f464ea720ee5b427de3a33489462f
1801 F20101204_AADBRG wilson_c_Page_084.txt
a781e94f2405aa6c20fcf541fae2f55e
df1a63f8b5046e35d72ff20d20457690047145db
1707 F20101204_AADDAE wilson_c_Page_090.txt
5d69f7b3a4b5e0f00f8deb9e2dfff634
ea85e93370f9959871f742e51e9ad45ec2bc6182
F20101204_AADCUJ wilson_c_Page_147.tif
9cfd1f8a5e86392df8bf010751d7cf86
3b0a616f26cd0d8ccff41d0c7cb7341fc7239eb4
F20101204_AADCTV wilson_c_Page_120.tif
b5c6d71dc00e7c98c3764da00409a42f
1f20337f4da44d08f58e9f802c038a8f85716ee1
F20101204_AADBQT wilson_c_Page_022.tif
f609d6303f5a0f8fa8dbe8249bb90fff
dbab16f97f415ac5dbde2acc14e55e673b1bf33d
83349 F20101204_AADBRH wilson_c_Page_146.jpg
5e4218565a8335c2856ff267d56db5c4
2e367b3e036a4f46cf644d05dc2885495d4c554d
607 F20101204_AADDAF wilson_c_Page_092.txt
57b1e0df77c70a7bfaa941677b64fccb
0ce69d48c7d3e338282f0a9dbecc2f12695c4787
F20101204_AADCUK wilson_c_Page_150.tif
9263d43c6ceb6fb3702620df6ed61ed8
e5c52ab9699b668c1dab63f6316c9e07f2f918de
F20101204_AADCTW wilson_c_Page_121.tif
372dfa2eabd4379540b347e1360d5e84
3358cc7e6f7aca79654ad8e04fcd01a4a6d06d00
68436 F20101204_AADBQU wilson_c_Page_012.pro
fbcce20fb011331dfed7a004f9ffdfdb
beb237c0b2b3150c66bfee3c04b3fa739bb35ed5
83296 F20101204_AADBRI wilson_c_Page_107.jpg
caeb8694c7b4b41ccc99e26aee192f62
db73314a26ce888455816de0831fc077128a1f49
918 F20101204_AADDAG wilson_c_Page_095.txt
8ecf9a4deb6a3a8ffe20094d82167ec5
60f5185f9151a32ca769c6a1b24378f0ff4975c4
F20101204_AADCUL wilson_c_Page_156.tif
d9ae44dba03b4baedd2af11154256ee5
b5ad91347f5efb3f8cdfb8d0aac192c5b4477cd6
F20101204_AADCTX wilson_c_Page_127.tif
a9720ff2d090d83bab0f75722f4632a7
4cd5a754cf13e489339e5f6b7fbb046d6afc7073
1249 F20101204_AADBQV wilson_c_Page_066.txt
fc34554d30dc701cd8ed91037ed03f0e
9bf99f56aa5e151700cfcd4f336f757aa9c74843
F20101204_AADBRJ wilson_c_Page_128.tif
6149426718638516e10f531df3989aca
8897a7de2caff9dddb31a260854fae51e85a3af5
1479 F20101204_AADDAH wilson_c_Page_097.txt
5dac42cea9140e5f5d2a76771af51b22
5f3a6118d1340e54f78636b0756ca46a795a2300
F20101204_AADCVA wilson_c_Page_180.tif
d83830b1d44cb8d8bb52b586e45956dd
921359e42cceeddddd251df03cfa4d1eac1325f2
F20101204_AADCUM wilson_c_Page_158.tif
9bb72ef01c6457fba5751860b6a7d8b4
d829357d007c3b0ca48b76e6d5234f1579d165e9
F20101204_AADCTY wilson_c_Page_129.tif
dacc63548e639fc095b1d775b601ef8c
5480ea6f86be934037edd53ecf19b5bb57fc9955
35197 F20101204_AADBQW wilson_c_Page_143.jpg
0a6ee58145cc43bd7bb1527ad31ab4f3
136e74622b5c0f1abb23db557b6e49486a4d38bc
1883 F20101204_AADBRK wilson_c_Page_074.txt
0affd661efd0747843e33ab3c2e7d38c
3bae8f1758d0349a0f96b64eea69297c5e671155
565 F20101204_AADDAI wilson_c_Page_100.txt
aef14acdc1cbc63f59711af14e692dfa
92d0f481b5d3ff84b7103ca7bdee05a46be60e0b
F20101204_AADCVB wilson_c_Page_183.tif
e35db60c3fc9a555f3bc28f48095b52c
53f01258a1707902b753913d656ec91d2b1c7f71
F20101204_AADCUN wilson_c_Page_159.tif
ac50ba206f590053463579cf0d417517
698e1a0b1e2ee6477d66694ae01e705363334187
F20101204_AADCTZ wilson_c_Page_130.tif
bd04c0026917e1bd5c33cb4964a1064f
3d78c3df49fc5ebef4fce505f54d03d86e347a19
25117 F20101204_AADBQX wilson_c_Page_094.QC.jpg
08f1cd2d55ec834193751fd6b0f34d16
a505795f0f65cabdc3a3742cf8f007da1f454a83
1051978 F20101204_AADBRL wilson_c_Page_006.jp2
089bf290cb779952f30a392deb7ac88d
d5414555e4908976858b219ec5998039e3d06ff2
F20101204_AADCVC wilson_c_Page_184.tif
c0db917bc413ffdc8c53459a5d466eb8
91bee853a61bc1654ed0365c412cbd73aae0af5c
F20101204_AADCUO wilson_c_Page_160.tif
1253bcf4f3f2b9b0d0cc7f665e0fa74f
c52a5302dd031ea679764443b99fae179b8c9881
F20101204_AADBQY wilson_c_Page_059.tif
9529061ca7251ec720c31e78e3c47e9e
2f841814c705638d39b06dfb71e3ecdb420d1d56
44290 F20101204_AADBSA wilson_c_Page_099.pro
a8f4ae7c431606caaf94b0ab63175f58
9c0458c2a5fec682a789ac100557ebd0a65a3a28
102183 F20101204_AADBRM wilson_c_Page_083.jp2
ab046489ee98480aa40e2cf7bb89857d
108745ce2702f23b92f3b9b3b7496e54b3dbdf9f
777 F20101204_AADDAJ wilson_c_Page_104.txt
4840c71c6e30d67ed88545e9eef13d9f
01cfb9e441e13798dde91a13beb8319c13b3d9cf
F20101204_AADCVD wilson_c_Page_185.tif
43ba745b5f349162fac0a258486df687
c66a05e5bd2a6cb5207677091f34dde764e3bc03
F20101204_AADCUP wilson_c_Page_162.tif
87f1007d6359ab496390d5b5c68688e0
992f4b2ce796a63e253ec1d3f59d968a38c8022a
28409 F20101204_AADBQZ wilson_c_Page_113.pro
8f72fd105ac6ae704a00a9c062f9982e
d47395b2a8afa5e63d3181ac6d184ea57e5227c8
F20101204_AADBSB wilson_c_Page_068.tif
4b9639e0cebfa0dbcdb87658d7d0a570
e36c74231b3e0c19d36a7d3f53abf254b9a5416d
106803 F20101204_AADBRN wilson_c_Page_012.jpg
0541fa7d96c1b4a6bc64db4b5b4d8558
4d2eaa82ac7944f37ec3b4734896de9596bf35a9
1973 F20101204_AADDAK wilson_c_Page_107.txt
088b4575529b75f212c70753bb9968fe
9791231e12e65e1491c29cd9d7c72db7d4e91364
F20101204_AADCVE wilson_c_Page_188.tif
74b0b5a1d94825c6a037a7b5a988d66f
d615825e96ec29580a33b40560fb417ca8e6a9bb
F20101204_AADCUQ wilson_c_Page_164.tif
763dcb917b3412b912ea94e642e18d47
12f3928bfa689e02f251c66e548885fca9facebe
70917 F20101204_AADBSC wilson_c_Page_020.jpg
feb2031101004868cf88333f60277fee
02f935ddbb27508f7831bb1dad9d69e2e27f4b0c
F20101204_AADBRO wilson_c_Page_106.tif
4d77db4724d476f627bc28e1ee4c9af3
e4d79310c1b56dc8e7c8ed6b1a47fa55364ce4e0
F20101204_AADDBA wilson_c_Page_149.txt
647bce0dc981d57d6055654846bba78c
75513d91027730459ad0f01283151263418c2946
1431 F20101204_AADDAL wilson_c_Page_108.txt
2fec0f2c753e4e3e28d3b9bc451ecb71
bfcbb1162c3673c58d36084a9b9e6756d5044684
F20101204_AADCVF wilson_c_Page_191.tif
fed297a15fe385c2f99dca13faa0d192
f4e0ede427b66998dc7b3ff5b69d807f67d49bf7
F20101204_AADCUR wilson_c_Page_165.tif
a393688878206f75e34bcdf85f009b31
26eed039166eaf9ec7a459bddb1152a7baeb5b88
36345 F20101204_AADBSD wilson_c_Page_158.pro
e103fd3599f680a3d704138eb9459698
81cc2e4daa7177c143f52973f71feba390275e0c
45935 F20101204_AADBRP wilson_c_Page_166.pro
6009106c18f52a64be085efb8de0409a
e7336b0e127d080e77cb0f12be0e0fa9d1276160
1585 F20101204_AADDBB wilson_c_Page_150.txt
d59ae9d4a9994154772b5d365e3ec2fc
71cb9a02b84762040a2e9e96964db0ce1988bcfc
1151 F20101204_AADDAM wilson_c_Page_109.txt
b7f4ed164adbf10ec7c7e3596f15d5fe
5cae1c4ba941d78fb2e3151f00d49eab8f61f050
8713 F20101204_AADCVG wilson_c_Page_001.pro
0e680a7b08577b33cf922489b00f4c22
d43029a564488a0b72a96eef2b5f3c2e7e8e2dad
F20101204_AADCUS wilson_c_Page_167.tif
bb4247e4aa74904387c5ae2d8f8478e0
4142b16f8ccc6aeadd3f596e5b5b75d4a713f57b
6450 F20101204_AADBSE wilson_c_Page_049thm.jpg
259b9f0d8873c2da269c8bdfbc3b361d
224b518bcb9b13ddf695cbacea2ca05f3a9d290e
1130 F20101204_AADBRQ wilson_c_Page_142.txt
9d6b9d77c9315d8b87353497d9f66842
51fae133a14a6c15a0d1d49d49d9950a52ec81ae
2058 F20101204_AADDBC wilson_c_Page_151.txt
de43a59e81c72b81cb874d72a030fe82
b85e0936580fb8c1cd5066995c3baa185e702e5a
1061 F20101204_AADDAN wilson_c_Page_112.txt
d7b76523b9aa368e19252475091273a5
e16d32b738fbfbf96a43067200b75bbd34e84795
1406 F20101204_AADCVH wilson_c_Page_002.pro
fc2713e81f1129301b36c27541a15123
2e946e640102e59f4be15a77bc84e07359eb3831
F20101204_AADCUT wilson_c_Page_168.tif
289ad1debff8afc7dc47f3dc1e7dbfde
943eab95d89c898949ad1de285cc2ab5a932d1fe
59060 F20101204_AADBSF wilson_c_Page_121.jpg
18e86d80e82f556051f83863aa790be4
d7831fd635dd12a09ff2a3d6a54d90a0b551817d
59289 F20101204_AADBRR wilson_c_Page_111.jpg
8bc8bc2789db713ea0aa1613b3d9803a
ac055434a5795e916acc827517880c496beaf7fd
1233 F20101204_AADDBD wilson_c_Page_155.txt
f7f8d9ec5b04143aef58ab0b9943ffac
7f61bafd0d611a6d430031390e6ceaf2a852b059
771 F20101204_AADDAO wilson_c_Page_116.txt
37111dbaa575d630bc862aa45156df2c
4c718a546e2bc08204756be62f2aedbc90bedf27
3341 F20101204_AADCVI wilson_c_Page_003.pro
7737b3f089b8f9d4a360e646b980abf0
057150db2bd7cf55004f2fac64a2bed37ca14d8b
F20101204_AADCUU wilson_c_Page_170.tif
3c6892b712ffd41e7c5175f2040b074e
690677b81ba95be9c4837df200ff9e7d2abcc7cf
F20101204_AADBSG wilson_c_Page_178.tif
5a1f4685a42cb14e5d88a501e8541bc2
fc9d12e80be890195b660c6435c974bfb3226a00
F20101204_AADBRS wilson_c_Page_018.tif
c152368b7a11e29552339a1f9a812889
4cc215cb8ef43d66ab9b9459dd8eb6d2f5be66e3
1109 F20101204_AADDBE wilson_c_Page_157.txt
a246abdf2f5bcb11caa32e868d39481f
d95511ba85ed12bfea8d1a536b070cc1c93582e7
1748 F20101204_AADDAP wilson_c_Page_118.txt
d9734d44650132ee19263bf98c42328d
f66d5bed8843d4b1c4897af83e37dbaa9d13044d
31386 F20101204_AADCVJ wilson_c_Page_004.pro
c7267825249797204eb19db0f4383db7
7362e497c330e6c7470008c923cc294c68574a22
F20101204_AADCUV wilson_c_Page_171.tif
66e0e75cd9aeb123a39ba1fa109c9883
1e5014a4ab1f95fd57f9e7e342dfd8a01bc7c39d
10626 F20101204_AADBSH wilson_c_Page_140.QC.jpg
dc17edf239dac1f14ca4858fd6eb0a57
09dd451871b794264ca17bb70655b534a9be2479
15951 F20101204_AADBRT wilson_c_Page_092.QC.jpg
884fa7b7a3be3bc88bd3dd797ff800b4
98ac059e6aa8023881fa2e39de59d7dd368a87ce
1627 F20101204_AADDBF wilson_c_Page_158.txt
b71411b5d7ae4d82d4a6ee150b9379ad
d14b4c09aa6ad0b183fea3ad67f14a570bc43810
953 F20101204_AADDAQ wilson_c_Page_123.txt
a7d1c4fec9605026f8550f78d2ab5ad8
50034f69842f3e8f352260433d7bcfd8a2b03942
63673 F20101204_AADCVK wilson_c_Page_005.pro
bf20f6e7505e64acc9deb6fef55a3243
c624a97d542149f79e946ff45348004e6378defd
F20101204_AADCUW wilson_c_Page_172.tif
7f0c318a5ee43430635026e59ae0c838
932b2f1ef4731668c04f757c6601f8624367eadb
F20101204_AADBSI wilson_c_Page_042.tif
0a3e1a9e401d10dcc63ff6b96153e308
94d5fa7012e8bf2198ccfc421077cb95bce3b564
849 F20101204_AADBRU wilson_c_Page_126.txt
9cd1343929eaab034ca8aaee6d450fb5
a00daa4c76c3ee27712eeb3945ff0eac2aca88e2
730 F20101204_AADDBG wilson_c_Page_161.txt
d56b7e53550ef9de187a5b8c1f99579f
093c084d11fbe23f06b1afa3bf48972a3ffb2fc7
804 F20101204_AADDAR wilson_c_Page_128.txt
aa99ae33ff2c0bfdac20ad6a1dacb59d
06ed7c9bd8eed660f907dd4e145b0959111eba61
87401 F20101204_AADCVL wilson_c_Page_006.pro
4841ff464f13b6c90e13e1d632556617
9dc395a6fc5f4638a139b4a0cd34cec6a7a83b3d
F20101204_AADCUX wilson_c_Page_173.tif
aeab5a7305d973153043253e8d83a4bc
b705590f6d65df3724ccd859c23f8fd007c5a0ed
2149 F20101204_AADBSJ wilson_c_Page_008.txt
09ed155f22b37aef4dfc5ae2dd21f1eb
d32aa968571b7562b73478b95dbdf5de901c44a3
24665 F20101204_AADBRV wilson_c_Page_033.QC.jpg
28a593ac8fa555ca5c5b8685673eff3c
4fc32731500009584d0b86109f1a85ca5aa63d84
644 F20101204_AADDBH wilson_c_Page_162.txt
0f81bf691fc228a545e2f444c8525d6a
c193f2d41321bc665a2f7b15741f88b4341dd1b3
1438 F20101204_AADDAS wilson_c_Page_130.txt
f5995a00ec835214fa3ac3362223f2af
7f7ae20a1757b1eae0902ad28c7f86e5c4bfb6c5
49948 F20101204_AADCWA wilson_c_Page_029.pro
65c663ba079bca3bb2a8e409060f66a9
6d3393c6e4952e015eb88b3ce922d40a4fe72fb5
31590 F20101204_AADCVM wilson_c_Page_007.pro
23f5f12da11d275db67090afb89ed82a
b73d57b299549b079c221c22e35cfd6adcb6e0fd
F20101204_AADCUY wilson_c_Page_174.tif
d55a30dac8c38f673e71db0ac8ecae10
f0a7892fb78eaefc0681fa3ae33ff4c65905ad60
10879 F20101204_AADBSK wilson_c_Page_149.QC.jpg
78c78dfd14e9af7393960d1bf48ae865
59883d84babad003ba885272dcc0afe4333d9cca
11025 F20101204_AADBRW wilson_c_Page_009.QC.jpg
d93587d75a66062a9ce9268bab3dd8e1
da3760fd5a030bd05eec753f68c381e0de50da31
1970 F20101204_AADDBI wilson_c_Page_163.txt
c808ff13e90d4f0dc917d9a50b44df54
2e3ebf12f2d61b5f6c68c904374f24f830196db7
1708 F20101204_AADDAT wilson_c_Page_132.txt
60b0465675f8e576263ab4017730028d
8352bc2a908898c249942ca07336b08c30138e60
24225 F20101204_AADCWB wilson_c_Page_030.pro
bfc7a969cce1b58cd81f293d798d0064
74e1abd4c865b0a597ac42f337d9ab0a4207cbf1
52482 F20101204_AADCVN wilson_c_Page_008.pro
ae563ccd82e244e02d21d1b83a6f4c68
b4716a622caf31d4d8cddfbcaa3d8098aa209848
F20101204_AADCUZ wilson_c_Page_175.tif
b03c7244d903b0fb95be14dd520aa0e5
287e3cbf961c6a2297782c3b80c10fbfcb302d7f
F20101204_AADBSL wilson_c_Page_015.tif
26a7751b24ace1401a39867e527cbdc1
1fede118aba0166f297b2ebc6e02c8e2ae744f80
F20101204_AADBRX wilson_c_Page_040.txt
7d38029589f393e836e6bcdaee195027
3732b79e237cf5def98d06fe18e663c8cedecb2e
1267 F20101204_AADDBJ wilson_c_Page_165.txt
0c1075fd866d03322a1b065d619832b9
531cd626208aaf4e329321f8b68a58ec1fbf4b66
1898 F20101204_AADDAU wilson_c_Page_134.txt
ba3cebdac0ef9fd8fcd8f96d087ebc0f
41999ef7b791235fdf4f14567ec6f2c0ac1ba1cb
42341 F20101204_AADCWC wilson_c_Page_031.pro
be91d0fbbef075b7708876dd0f94ed01
884fd197c084c8f7d2e641aed6656da2c4562604
55134 F20101204_AADCVO wilson_c_Page_010.pro
621baad05deefd6f28b57f0793808287
8d6f1e62682eeb8aaf37aa2ea920547bcb91b2d2
1209 F20101204_AADBTA wilson_c_Page_062.txt
78c81e2b62dd5db8b1f49bbd18ccdf75
f875b419a2d20bfaa03c5110c493c4757f2d3f7f
27851 F20101204_AADBSM wilson_c_Page_013.QC.jpg
85b60068cf8acc5bd14a35730477a870
445304cd23b3548ce2c4cd46fd0d73de1cf25f29
27095 F20101204_AADBRY wilson_c_Page_051.pro
2007c4b3e7d8b7c7737316c0b3d14855
bf27dff75a91b953d17054dc865524e07798fe39
1328 F20101204_AADDAV wilson_c_Page_135.txt
4952b7899200bf00bd746290ce073dcf
c3864ab7d46bb56d7d529c7a342eb9ecb01c12df
48494 F20101204_AADCWD wilson_c_Page_036.pro
fdde5517e545f6189c44ffbb11488c2b
7de347cb1b1c65f98d55f5ebf1ef86e0a1960650
69028 F20101204_AADCVP wilson_c_Page_011.pro
1b419468b08bcf6377c3f2f35e313daf
4817a00df184bd0d511016397c73357d01552336
F20101204_AADBTB wilson_c_Page_032.tif
44d29ff9a16dde3373f486faa5e1bca4
da61c3e93683a7409c0dce9fa90ae83ff439cf26
54099 F20101204_AADBSN wilson_c_Page_035.jpg
40a6767a9d22db4da99ca46a8e8ea043
52b850f02458278ba346f72fc9893346637e0d5a
32734 F20101204_AADBRZ wilson_c_Page_162.jpg
7fe5b237ce44c3102d623e4e1fffbc52
70be63a52468b818c26ba7110cc2241a3f1c5d28
1960 F20101204_AADDBK wilson_c_Page_171.txt
95fb6b9b172fcd7f33973f755730cb29
f3a524483e71ff515f9acfd531b64fc1ee1a7424
706 F20101204_AADDAW wilson_c_Page_136.txt
01344d1727801be37436e792a81a3d3c
64aa15a2b5c2acdf9f4d76aac88380488a1df696
40454 F20101204_AADCWE wilson_c_Page_044.pro
0027da9937d29aa9d9cae86e6e90446a
c9e56d40658eeb30423ac8e2d45695baed482762
67708 F20101204_AADCVQ wilson_c_Page_013.pro
3bce925cae31a4b29626d16aca4c3deb
c116195d33a766311d04af27c4479bac468a60af
437 F20101204_AADBTC wilson_c_Page_168.txt
ca2a5f6fe8b6d21538744a7a34f2ffbe
730788e1a23f9173c8e309787f1581c825027ded
658390 F20101204_AADBSO wilson_c_Page_062.jp2
e22a83669e3a5cb276afbc6e88fbb529
1a625f240bb1cc6cb14c347bbb865a9f9f83b3b7
F20101204_AADDCA wilson_c_Page_001.QC.jpg
bc47b89d320dbece6dd0b202795c055a
836d10514de29e05133988d08f9dc83fcea1f319
1158 F20101204_AADDBL wilson_c_Page_173.txt
94a432e61a3115ddaaf74a91986c549a
af3441e2b9a8ce121d1fb81b801bb716c162696b
1958 F20101204_AADDAX wilson_c_Page_139.txt
bd3a23c43809a810ad46c310ffdfd8e8
6f2c4bfca6a400cc234a606fd4e2643531d72e21
24233 F20101204_AADCWF wilson_c_Page_045.pro
82c843c6dc7c437812e81ce01a72959d
c28d2187035e869300bd06555acd8626bbb9d74b
73867 F20101204_AADCVR wilson_c_Page_015.pro
ac7c701141bdca805da568c0ba5bfcc0
292ed19f955f273546c8f541f311ef7630416e7d
1884 F20101204_AADBSP wilson_c_Page_103.txt
5101587664019ac59bb25c6d55c2c8a7
7a3a816ab16b0c0ec85c928898701861208c8fd1
5079 F20101204_AADBTD wilson_c_Page_175thm.jpg
53cd6b720f42b004790927350594dc64
8cd72ef47d97b1ccdc29a2035532b33cac5b3cb2
553 F20101204_AADDCB wilson_c_Page_002thm.jpg
b5028288196aea2ba8b486d344cb15c7
96e61a963276be2851754997371e3d15a8cb36c0
2050 F20101204_AADDBM wilson_c_Page_174.txt
e90e6826c47edc531c0ee8cd07379170
6d71603d36f3073af46be779f6e999d26fb11438
654 F20101204_AADDAY wilson_c_Page_140.txt
14dfb21d3c45bc944be49fbda94030d7
1ce4c05971c7d04e5ca00564a82910578ce41150
22285 F20101204_AADCWG wilson_c_Page_046.pro
e5d0c404b234b65aa88636fab399176f
af25dcd1950ca68439f2d2d47526f95ad3765892
71377 F20101204_AADCVS wilson_c_Page_016.pro
1c09a5ca11ece587f7453dde269e6a54
2d3fc0f6ef21cdf3176ce760f5da68aba6a54ad7
40730 F20101204_AADBSQ wilson_c_Page_071.pro
0564ab76892f158c25ee75b70092d0a4
b35a8543762770d2cc7f34056dc44ebcf3bbc9c2
68882 F20101204_AADBTE wilson_c_Page_097.jpg
ff7f23cd09cf3222758059d769c5e17e
0762ae6f524e3b64a025789b7af023ce5166310a
2075 F20101204_AADDCC wilson_c_Page_003.QC.jpg
d8e99d32c001beb19198fe67a9190409
783b0fabdd171c81e2a824ab5c73b3d43d1fc5b5
439 F20101204_AADDBN wilson_c_Page_175.txt
bb68762689443aa6fa862eb4aa20990e
43d119602eedfc1aac34e14de2c2703961df7358
1990 F20101204_AADDAZ wilson_c_Page_146.txt
da57c3fed1cb1b20626a14ce0c2a3d0d
691f663ee1462ddb747c421bbd2602043bd121e8
13767 F20101204_AADCWH wilson_c_Page_059.pro
45daa59b97c438eb9f8b3ec7b3a0356f
d30051aa08bbd051e7e323df1d995ab85e3cafd4
11811 F20101204_AADCVT wilson_c_Page_017.pro
44af2e07743ae686d2714f5809106b32
7dd1cf6c3c5df3f6ff168691d989275de0db2243
4174 F20101204_AADBSR wilson_c_Page_113thm.jpg
18c696dfe57ebf48c17506646e206142
92ef9b2a23ee69c123cf5fe21bd25c0cd088b4f9
600809 F20101204_AADBTF wilson_c_Page_072.jp2
63785d2d3a5350f7406fd8e1c283bff6
e0cb6632a6b31b4aa3fd726d22a91e45a77291e7
801 F20101204_AADDCD wilson_c_Page_003thm.jpg
de72505aebd21c5eefbebb9a373a7464
f023df864730b471e922ec64325ccd9631777be5
324 F20101204_AADDBO wilson_c_Page_176.txt
d7b8d275d361447bedb2dac96bce03e5
2db99e71426f25bb99497f6677057ca32f6947e3
26733 F20101204_AADCWI wilson_c_Page_062.pro
85178bdfa47e60efe0d106cee525504f
4c0e9236b266f5b89383ad3d5499482db5f4b287
37637 F20101204_AADCVU wilson_c_Page_018.pro
5e88dee678a4d721c6b8e0caf9aa8abb
499f77025dbb4ab6c3e49fe4b28cfe0ca4600aff
5270 F20101204_AADBSS wilson_c_Page_158thm.jpg
4a8827a156abefe7912dff70c83c2d85
a688494cf04a9edeeb78f2a67ccaf07a61ff1711
45303 F20101204_AADBTG wilson_c_Page_163.pro
0e8b31b372f6f24a58149f262c726101
54c8dbdfb71fa466fe0eee313bc50723ead6d66a
4373 F20101204_AADDCE wilson_c_Page_004thm.jpg
73c76f7b10f09eb64047a708a146685b
ca3020391bee7360680b859a72db28a3a95f5e8d
231 F20101204_AADDBP wilson_c_Page_177.txt
1f7d1e1e0c89236f3a2e7d36407fb3ab
08059cfebeb50d957f42f69e8a614e09583ab312
30488 F20101204_AADCWJ wilson_c_Page_066.pro
f64e0ac7d4ac86d66060afe14cb94eec
bab4c58c7a918eae2c8b05bd026280a34d9d9a00
41928 F20101204_AADCVV wilson_c_Page_020.pro
772e9ca8ce9d42ad4998b9941df87058
dd6c4f2d0cddab7e62805c217a5b177beb993a2f
5403 F20101204_AADBST wilson_c_Page_133thm.jpg
a9180f420a5309e2a9cfdc4e47e80a75
51b29990b6d54f3931c35cfb58f695ea8a839b2f
F20101204_AADBTH wilson_c_Page_032thm.jpg
bce1212534b825da0c46dc22d1b3ff47
ecd93086bbe32c234298be4c91d71da2f1f2ec7c
4279 F20101204_AADDCF wilson_c_Page_005thm.jpg
bfd4a155323ffc9ddecbf02e292b4dbc
7ca66d6bdf41ff1e646eb91af8601dd016d0780b
376 F20101204_AADDBQ wilson_c_Page_178.txt
b8357b05690a9a3fe4ed6230d33de4cb
b503edd0fcbdb980f2823d2953a35ee4100cde13
47618 F20101204_AADCWK wilson_c_Page_067.pro
0835af0adfed053cf836149d2e63998a
d84275e5f44cf28a46052b3d601abd23763b36bc
40329 F20101204_AADCVW wilson_c_Page_023.pro
dea93e5ec198dd7b6230c523bf356702
80201e0112e89ee57739baadf541dc0b70d56905
F20101204_AADBSU wilson_c_Page_179.tif
a781928045c42991ceb5a1dcce9219c7
953f63a58e96cad44074713992b216400a312f95
1455 F20101204_AADBTI wilson_c_Page_138.txt
31df76fc4149b002d11b4ffbe6cc7aa0
9689b4ccc8d7ee155a6140644925edfaa3e9160b
2814 F20101204_AADDCG wilson_c_Page_007thm.jpg
036914d1dd2c2608b7fc42879513addb
aa9db1c2d5f3c3deb9d53102b10e7f5e4384cc91
1805 F20101204_AADDBR wilson_c_Page_181.txt
2b583a07ab509c8a2a6b5b13855661a1
4b4a7753f6f23cfa310ae9f3a78e2b88f60c1237
50217 F20101204_AADCXA wilson_c_Page_107.pro
63c2000d18b77452053ba7ee595d4cd8
cee7ce5a80c14b77a0fcd39a5ee4eb60eecfe346
23472 F20101204_AADCWL wilson_c_Page_072.pro
e3d300f1004c887df22ad2c6dec9863b
b549d17dda3d621eba47f6a2e85246d3d158d60d
46530 F20101204_AADCVX wilson_c_Page_024.pro
3dfbe15d6259444ec1aac6f671c06042
5cb29406e0f7719879d838f7803b57ddca87c06a
17077 F20101204_AADBSV wilson_c_Page_135.QC.jpg
eb4a7583e69b2d8753aa91347814e707
7efce09a4bf8f90f0493cd1b78d8e84047609a6f
111840 F20101204_AADBTJ wilson_c_Page_183.jp2
bec5650bd6f880d9dfab50acedaaf992
2363eb2e2ee366e888eed5b9a5d1dfc664b2e475
22335 F20101204_AADDCH wilson_c_Page_008.QC.jpg
4227b6c3cd99bf28d511db0d8d90eed6
857b961edee76fa500ac1c2f4da6e33e227f76ff
1374 F20101204_AADDBS wilson_c_Page_182.txt
cb71c79d60a7cbf6dc6ed16eec28e6cf
4b019bfc14563afa07a977a6411efa19ddfb90ad
6510 F20101204_AADCWM wilson_c_Page_076.pro
747107bfcc8e1872bdecd54f988935ac
0fe733ede8bab8e7a342a986674425ce3f573b20
30987 F20101204_AADCVY wilson_c_Page_025.pro
091fcc44b154a0535a02fde5436e0787
160cb5ffdf1caf7361df18e4d497a191910e2d53
71749 F20101204_AADBSW wilson_c_Page_086.jpg
cf2609b26d995c79686f0d2a34a19a98
2a648cab781847fa3003f487b2d9c26fa33142bd
6061 F20101204_AADBTK wilson_c_Page_057thm.jpg
ebc86019dcba2218e2f5f6de1273822c
a6ac8d1b1b08913aeeaf218d8e51b99f19ec958c
2817 F20101204_AADDCI wilson_c_Page_009thm.jpg
a4049987e18e10c3323fb4e3c17f6524
54e7b1d4626ac41eb0ca7ecf98dad9b5628aa438
2072 F20101204_AADDBT wilson_c_Page_185.txt
784607bedf7f7d270dfaee19df1236f5
aa638bac04d5d403f2bc3bae824ba6d2b4ccb83b
34269 F20101204_AADCXB wilson_c_Page_111.pro
17fbb94dab38ab1a1627c9d768fd5b4f
1c4f90b4b6139a1c7e4982eaea91d2011ec28f6b
11925 F20101204_AADCWN wilson_c_Page_077.pro
7dd2ae05d71739b94576079b63bb4ad2
8fed8273ebeac4a26a574fb82f04ac2ea0130d38
45449 F20101204_AADCVZ wilson_c_Page_026.pro
e1ecc6bf6afc590704f0a2450406e412
63ce194f3cb70bdeca8c83e1e835ff019fc6ce80
46034 F20101204_AADBSX wilson_c_Page_033.pro
4a3047d95ad5595fc5e1147a253837c6
7e4f1be53aeef8eb9ad9bb9260bb6ff7e65b626e
105831 F20101204_AADBTL wilson_c_Page_184.jp2
19c462a7da1c40bace670db98d39d659
1b58b869fc16162471559e7a80b04c4111eb89c7



PAGE 1

DESIGN, FABRICATION, AND CHAR ACTERIZATION OF LAMINATED HYDROXYAPATITE-POLYSULFONE COMPOSITES By CLIFFORD ADAMS WILSON II A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 By Clifford Adams Wilson II

PAGE 3

To my wife, Leslie, and my parents, Cliff and Marilyn, fo r your love and encouragement throughout this entire process

PAGE 4

ACKNOWLEDGMENTS I would especially like to thank Dr. John Mecholsky Jr. for his guidance and for always having his door open for students. I would also like to thank the other members of my supervisory committee, Dr. Anthony Brennan, Dr. Kenneth Anusavice, Dr. Wolfgang Sigmund, and Dr. Bhavani Sankar, for their input and advice, which helped guide me through this project. I need to acknowledge the contributions to this work made by Leslie Wilson, of the Department of Materials Science and Engineering, who prepared the initial PSu solutions used for solvent casting, Dr. Sukjoo Choi of the Department of Mechanical and Aerospace Engineering, who performed the finite element analysis presented in this study, and Gil Brubaker of the Particle Engineering Research Center, who performed the particle size analysis. I want to thank Allyson Barrett and Ben Lee of the Department of Dental Biomaterials for their assistance. Finally, I would like to thank the all other graduate students who helped made coming to lab everyday a pleasure, but especially, Dr. Tom Hill, and Dr. Alvaro Della Bona who helped to become acclimated when I first arrived at the University of Florida. iv

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ...........................................................................................................viii LIST OF FIGURES .............................................................................................................x ABSTRACT ...................................................................................................................xviii CHAPTERS 1 BACKGROUND AND RATIONALE.............................................................................1 Requirements for an Ideal Hard Tissue Regeneration Device......................................1 Materials Selection.......................................................................................................4 Research Rationale and Objective..............................................................................10 2 HYDROXYAPATITE....................................................................................................12 Background on Processing of Hydroxyapatite...........................................................12 Starting Materials........................................................................................................13 Processing Methods....................................................................................................13 Tape Casting.......................................................................................................13 Cold Pressing......................................................................................................14 Organic Burnout and Densification....................................................................14 Sintering..............................................................................................................17 Specimen Refinement.........................................................................................18 HA Characterization Methods....................................................................................18 Particle Size Analysis.........................................................................................18 X-Ray Diffraction...............................................................................................18 Density and Porosity...........................................................................................18 Elastic Modulus and Poissons Ratio..................................................................19 Optical Microscopy and Lighting Effects...........................................................20 Scanning Electron Microscopy...........................................................................20 Mechanical Testing Methods......................................................................................20 Establishing Baseline Properties for Monolithic Hydroxyapatite......................20 Indentation..........................................................................................................21 Biaxial Flexure....................................................................................................22 Work of Fracture and Toughness........................................................................22 Fracture Toughness.............................................................................................23 v

PAGE 6

Finite Element Analysis......................................................................................25 Statistical Analysis..............................................................................................26 Results and Discussion...............................................................................................26 Starting Powder Characterization.......................................................................26 Development of a Tape Casting Method for Hydroxylapatite............................28 Burnout and Sintering Process............................................................................35 Characterization and Mechanical Properties of Post-Burnout Hydroxyapatite..36 Problems with Specimens Fired to 1000C........................................................38 Mechanical Properties of 1000C Specimens.....................................................38 Constant Fracture Toughness..............................................................................41 Firing Study........................................................................................................43 Sintering Temperature Effect on Hardness.........................................................44 Sintering Temperature Effect on Biaxial Flexure Strength................................44 Sintering Temperature Effect on Density...........................................................44 Qualitative Determination of Hydroxyapatite Decomposition...........................46 Sintering Temperature Selection.........................................................................47 Optimization of Sintering...................................................................................48 Density and Elastic Modulus of 1200C Specimens..........................................49 Hardness of 1200C Specimens..........................................................................51 Biaxial Flexure Strength of 1200C Specimens.................................................52 Work of Fracture and Toughness........................................................................53 Fracture Toughness of 1200C Specimens.........................................................54 Monolithic Hydroxyapatite Specimens...............................................................55 3 LAMINATE FABRICATION........................................................................................59 Design and Nomenclature...........................................................................................59 Laminate Design.................................................................................................59 Laminate Geometry and Nomenclature..............................................................59 Solvent Casting of Polysulfone...........................................................................60 Materials.....................................................................................................................61 Methods......................................................................................................................62 Solvent Casting of Polysulfone...........................................................................62 Specimen Preparation.........................................................................................64 Characterization of PSu Films and Laminates....................................................71 Results and Discussion...............................................................................................72 Problems With the Solvent Casting Methods.....................................................72 Thermal Analysis of Polysulfone Layers............................................................79 4 LAMINATE THEORY..................................................................................................91 Background.................................................................................................................91 Methods......................................................................................................................92 Laminate Theory.................................................................................................92 Finite Element Analysis (FEA)...........................................................................95 Material Modeling Parameters............................................................................95 Results and Discussion...............................................................................................97 vi

PAGE 7

Comparison of FEA and Laminate Theory Models............................................97 Predicting Laminate Behavior...........................................................................102 5 LAMINATE BEHAVIOR............................................................................................111 Methods....................................................................................................................111 Mechanical Testing and Characterization of Laminates...................................111 Laminate Testing Variables...............................................................................112 Results and Discussion.............................................................................................114 Comparison of Laminates to Monoliths............................................................114 Testing of Laminate Parameters........................................................................118 Determination of Failure Mechanisms..............................................................135 Fracture Analysis of 400-200-800 Laminates...................................................139 Failure Mechanism of Laminates with Outer HA Layers < 400 m.................155 6 CONCLUSIONS...........................................................................................................162 LIST OF REFERENCES.................................................................................................170 BIOGRAPHICAL SKETCH...........................................................................................175 vii

PAGE 8

LIST OF TABLES Table Page 2.1 Mechanical data and fractography measurements for HA specimens fired to 1000C...............................................................................................................42 2.2 Biaxial flexure strength data for specimens sintered at 1200C............................52 2.3 Work of fracture values for specimens sintered at 1200C....................................54 2.4 Fracture toughness values for 1200C hydroxyapatite specimens........................55 3.1 Hardness of laminates made through the matching halves method.......................76 3.2 Weight percent solvent retained results for PSu layers fabricated at all three drying temperatures......................................................................................79 3.3 Comparison of laminates fabricated with polymer layers composed of both PSu and sPSu.........................................................................................................90 4.1 Initial comparison of laminate models to FEA for monolithic HA.......................98 4.2 Comparison of corrected laminate models to FEA for monolithic HA.................98 4.3 Comparison of maximum principal stress calculated through FEA and Laminate Theory for various laminate geometries................................................99 4.4 Contact stress field radii for each laminate geometry that yields the same results as the FEA................................................................................................100 4.5 Maximum stresses on the tensile surface of each layer for all laminate geometries experimentally tested in this study....................................................106 5.1 Comparison of the average mechanical property values for the monolithic HA versus the 400-200-800 laminates.................................................................116 5.2 Hardness data for initial flaw size laminates........................................................118 5.3 Mechanical property data versus indent loads......................................................119 5.4 Hardness data for the laminates with different PSu layer thicknesses..................122 viii

PAGE 9

5.5 Apparent toughness values versus increasing PSu layer thicknesses..................125 5.6 Hardness data for laminates with different outer HA layer thicknesses..............127 5.7 Apparent fracture toughness comparison for different outer HA layers..............131 5.8 Hardness data for the four laminate groups prepared with thicker middle HA layers.............................................................................................................131 5.9 Comparison of flaw sizes for monolithic HA and the outer HA layer for 400-200-800 laminates.........................................................................................139 5.10 Middle HA layer failure stress calculations for the 400-200-800 laminates.......144 5.11 Failure stress calculations for laminates with outer HA layers < 400 m...........161 ix

PAGE 10

LIST OF FIGURES Figure Page 1.1 A plot of fracture toughness versus Youngs modulus of biomaterials being developed for bone replacement adapted from Suchanek and Yoshimura [6]..........................................................................................................6 2.1 Weight % of each constituent in the tape casting slurry........................................15 2.2 Green tape processing, shown are punched green discs (top left), the 25 mm (1) knife edge punch (middle), and scrap tape (top right). The section of tape processed (bottom) is 30 cm (12) x 11 cm (4.5) and produced 36 green discs..............................................................................................................15 2.3 Organic burnout cycle performed on the consolidated discs produced by cold pressing hydroxyapatite green tape................................................................16 2.4 Oblique allowed surface features to be seen (bottom) in greater detail than with overhead lighting (top)...................................................................................21 2.5 Schematic representation of fracture markings that result from a brittle fracture adapted from Mecholsky, et al [53]..........................................................24 2.6 Particle size distribution for the starting hydroxylapatite powder.........................26 2.7 XRD spectrum for the as-received hydroxylapatite powder..................................27 2.8 Effects of applied pressure (top) and hold time (bottom) on the reduction in thickness of stacked green tapes during cold pressing......................................31 2.9 Optical micrographs of post-cold pressing edge delamination (top left), delamination at the center (bottom left), full consolidation (top right), and post-sintering center delamination.........................................................................32 2.10 The reduction in thickness of the stacked green tapes is plotted versus wt% binder in the tape casting slurries (top), and binder:plasticizer ratio (bottom)..................................................................................................................34 2.11 TG/DTA data curves for a pre-burnout green tape sample (top) and a postburnout hydroxyapatite specimen..........................................................................37 x

PAGE 11

2.12 XRD spectrum for HA specimens burned out at 1000C (top) and specimens sintered at 1200C (bottom).................................................................39 2.13 XRD spectrum for HA specimen fired to 1200C (top) and for naturally occurring HA (bottom) [8]. Circles () indicate peaks found on both spectra, the square () indicates the peak that is unique to the 1200C spectrum.................................................................................................................40 2.14 Optical micrographs of a Vickers indent for a 1000C specimen (left) and crack propagation through a Vickers indent..........................................................42 2.15 Log-log plot of failure stress versus indent load for HA specimens fired to 1000C...................................................................................................................43 2.16 A plot of hardness versus sintering temperature for the hydroxyapatite specimens.................................................................................................................45 2.17 A plot of failure stress versus sintering temperature for the hydroxyapatite specimens...............................................................................................................45 2.18 Plot of density versus sintering temperature for hydroxyapatite specimens..........46 2.19 Color comparison of specimens sintered at five different temperatures. Each row of specimens was fired at the temperature indicated to the right..........47 2.20 Plot of failure stress versus hold times at 1200C.................................................50 2.21 Plot of failure stress versus ramp rate sintering temperature of 1200C...............50 2.22 SEM micrograph of the microstructure of the hydroxyapatite specimens sintered at 1200C..................................................................................................51 2.23 Comparison of stress-strain curves for indented and non-indented HA specimens...............................................................................................................53 2.24 All specimens tested in biaxial flexure fractured into either two (left) or three (right) pieces.................................................................................................54 2.25 Plot of K C versus K SI for the hydroxyapatite specimens sintered at 1200C.........56 2.26 SEM micrograph of the initial flaw caused by an indent (black bar) for a 1200C HA specimen. The outer boundary of the initial flaw (white arrows) and a twist-hackle marking (black arrow) are shown...............................56 2.27 Optical micrographs of an entire fracture surface (top) and initial flaw (bottom) of a 1200C specimen indented with a 3.35 kg indent. The white arrows indicate the outer edge of the critical flaw.................................................57 3.1 Schematic diagram of HA/PSu laminate...............................................................60 xi

PAGE 12

3.2 Polysulfone............................................................................................................61 3.3 Sulfonated Polysulfone..........................................................................................62 3.4 Step-by-step schematic of laminate fabrication method 1: The Matching Halves Method.......................................................................................................68 3.5 Step-by-step schematic of laminate fabrication method 2: The Bottom Up Method...................................................................................................................69 3.6 Step-by-step schematic of laminate fabrication method 3: The PreFabricated PSu Layer Method...............................................................................70 3.7 Optical micrographs of (left) a large surface bubble formed during laminate fabrication, and (right) a small bubble (black arrow) within a PSu layer that caused a HA layer to fracture (white arrows)........................................73 3.8 Left: Optical micrograph of a PSu layer that peeled from the HA disc during drying. Right: Higher magnification image of the PSu layer (white arrow) that peeled due to fracture of the HA layer (black arrows) which is still bonded to the PSu layer..................................................................................73 3.9 Optical micrographs of a large hole that formed during indentation (left), and of a laminate with a defective PSu layer containing a large open cavity (right).....................................................................................................................76 3.10 SEM image of bubbles (black arrows) which formed during laminate fabrication using the bottom up technique.............................................................77 3.11 Schematic drawing of chipping failures that occur when fabricating thick PSu layers through the bottom up method.............................................................78 3.12 Complete TG curve for an as received PSu pellet.................................................81 3.13 A comparison of TG curves PSu films dried at the three different temperatures...........................................................................................................81 3.14 TG curves for six different PSu films dried at 70C..............................................82 3.15 TG curve for a PSu sample taken from a delaminated region of a laminate fracture surface.......................................................................................................82 3.16 DSC curves generated through three heating-cooling cycles for an as received PSu pellet.................................................................................................85 3.17 DSC curves generated through three heating-cooling cycles for a PSu film dried at 70C..........................................................................................................85 3.18 A plot of elastic modulus vs. drying temperature for PSu layers..........................86 xii

PAGE 13

3.19 A plot of break stress vs. drying temperature for PSu layers.................................86 3.20 A plot of elongation at break vs. drying temperature for the PSu layers...............87 3.21 A plot of failure loads and stresses vs. laminate fabrication method and drying temperature. Units of the y-axis are different for each series....................87 3.22 A plot of laminate failure loads and stresses vs. surface polishing medium. The ordinate axis has units are unique to each series plotted................................89 4.1 A schematic representation of a laminate indicating the mathematical variables required for laminate theory calculations...............................................93 4.2 The Microsoft Excel spreadsheet designed to calculate the laminate stress distribution for the 400-200-800 laminate.............................................................96 4.3 Graphical representation of the stress field resulting from a 1 N applied load on a 400-400-800 HA/PSu laminate..............................................................97 4.4 Stress distribution calculated using laminate theory resulting from a 1N load applied to a 400-200-800 laminate. Laminate layers are drawn to scale, with the 2.00 thickness representing the outer HA layer tensile surface..................................................................................................................102 4.5 Maximum stresses for the 400-200-800 laminate as (a) a plot of individual points and (b) a graphed smoothed curve............................................................104 4.6 Maximum stress curves for laminates as a function of varying polymer layer thickness......................................................................................................107 4.7 Maximum stress curves for laminates as a function of varying outer HA layer thickness......................................................................................................107 4.8 Maximum stress curves for laminates as a function of varying middle HA layer thickness......................................................................................................108 4.9 Maximum stress curves for various laminate geometries having a total laminate thickness of 2.0 mm..............................................................................108 4.10 A plot of maximum tensile stress values versus the ratio between the thicknesses of the outer HA and PSu layers........................................................109 4.11 A plot of maximum tensile stress versus the ratio of thicknesses between the PSu and middle HA layers.............................................................................109 4.12 A plot of maximum tensile stress versus the ratio of thickness between the outer and middle HA layers.................................................................................110 xiii

PAGE 14

5.1 Comparison of load displacement curves for the 400-200-800 laminates compared with monolithic HA.............................................................................116 5.2 Comparison of mechanical properties measured for monolithic HA as the 400-200-800 laminates. The units of the ordinate axis vary for each property listed with units designated in the column headings.............................117 5.3 Comparison of failure stress for 400-200-800 laminates versus both indented and non-indented monolithic HA specimens........................................117 5.4 Lateral cracking seen during indentation using a 9.35 kg load (left), and the resulting chip-out that occurs during loading (right).....................................119 5.5 A plot of failure stress versus loading rate...........................................................121 5.6 A plot of absorbed energy at failure (toughness) versus loading rate for 400-200-800 laminates.........................................................................................121 5.7 A plot of failure loads versus PSu layer thickness. Monolithic HA failure load is plotted as a PSu layer thickness of 0........................................................123 5.8 A plot of failure stress vs. PSu layer thickness. Monolithic HA is plotted as a PSu thickness of 0.........................................................................................124 5.9 A plot of stress vs. outer HA/PSu layer thickness ratio for a comparison of experimental values to laminate theory predictions.............................................124 5.10 A plot of work of fracture vs. PSu layer thickness..............................................125 5.11 A plot of absorbed energy at fracture (toughness) vs. PSu layer thickness.........126 5.12 A plot of failure loads versus outer HA layer thickness. The total thickness of the specimens is shown in parenthesis.............................................128 5.13 A plot of failure stresses versus outer HA layer thickness. Monolithic HA is plotted as a 2000 m thick HA layer................................................................129 5.14 A plot of work of fracture versus outer HA layer thickness................................130 5.15 A plot of toughness versus outer HA thickness...................................................130 5.16 Failure loads for laminates with thicker middle HA layers.................................134 5.17 Failure stresses for laminates with thicker middle HA layers.............................134 5.18 Experimental failure loads for all the laminate geometries having a total laminate thickness of 2.0 mm..............................................................................135 xiv

PAGE 15

5.19 An optical micrograph of 400-200-800 laminate fracture surface. The locations of the initial indent flaw (white arrow), middle layer flaw (black arrow), and loading piston contact are all indicated............................................136 5.20 Optical micrographs of a laminate still intact after being loaded to failure in biaxial flexure. Left: Failure occurred from the initial indent (indicated by the arrows) in the outer HA surface, Right: Examination of the edge of the specimen shows failure of the outer HA layer, delamination along the interface, and failure of the middle HA layer......................................................137 5.21 SEM image of punch through failure of a 400-400-800 laminate. The remnants of the indented outer HA layer are indicted by the black arrows, failure of the PSu layer were the loading piston breached all five layers is indicated by the white arrow................................................................................138 5.22 Optical micrograph of the initial flaw size (black arrows) produced by a 3.35 kg indent (white arrow) in a 400-50-800 laminate......................................140 5.23 SEM image of the initial flaw produced in a 400-100-800 laminate produced by a 3.35 kg indent load.......................................................................140 5.24 Load-displacement curves for two non-indented laminates. Outer HA layer failure loads are calculated either from the initial drop (top) or by extrapolating the initial slope (dashed line) and estimating when the slope change begins (bottom)........................................................................................142 5.25 Failure stresses of the outer HA layer calculated for three different initial flaw sizes..............................................................................................................143 5.26 Failure stresses of the outer HA a layer for laminates built with different PSu layer thicknesses...........................................................................................143 5.27 Optical micrograph of the fracture oringin of a 400-200-800 laminate. The center of the flaw is indicated by the white arrow...............................................145 5.28 Optical micrograph of a middle layer fracture origin a 400-100-800 laminate. The specimen has been sputter coated to make the surface marking stand out.................................................................................................145 5.29 SEM image of the middle HA layer fracture origin............................................146 5.30 SEM image of a 400-100-800 laminate showing contact damage (white arrows) from the loading piston which occurred prior to propagation of the primary fracture (black arrows) which was reinitiated near the HA/PSu interface (dashed arrow)......................................................................................148 5.31 Optical micrograph of secondary cracks (arrows) which occurred through punch failure after primary failure from bending................................................148 xv

PAGE 16

5.32 Optical micrograph of a PSu layer of a 400-200-800 laminate fractured at a crosshead displacement rate of 0.25 mm/min. The specimen was sputter coated prior to analysis........................................................................................149 5.33 Optical micrograph of a PSu layer of a 400-200-800 laminate fractured at a crosshead displacement rate of 25 mm/min......................................................149 5.34 Optical micrographs of the PSu layer features through overhead (top) and oblique lighting (bottom).....................................................................................150 5.35 SEM images of brittle failure of the PSu layer, the pore (indicated by the arrow in the left image) shows a twist hackle marking (arrow in the right image) characteristic of brittle fracture................................................................150 5.36 SEM image of ductile deformation of the PSu layer prior to failure at the center of the feature (black arrow).......................................................................151 5.37 SEM image of two side-by-side features within the PSu layer that failed in different manners, a twist hackle marking (solid arrow) indicates the left feature failed in a brittle manner, while the pullout deformation (dashed arrow) of the right feature indicates ductile failure.............................................151 5.38 Optical micrograph of a sPSu/HA laminate showing the depth of penetration of the casting solution during solvent casting...................................154 5.39 SEM images showing the penetration depth of PSu into the middle HA layer. The fracture surface is shown top left, the box indicates the area where the top right picture was taken, the bottom picture showing PSu pullout was taken from the area indicated by the box in the top right image......156 5.40 SEM images of PSu bridging of small cracks in the HA layers. These images were taken of the secondary interface crack shown in Figure 5.31.........157 5.41 SEM images showing PSu bridging of the HA/PSu interface. PSu fibrils are indicated by the arrow....................................................................................157 5.42 SEM image of PSu bridging of the HA/PSu interface.........................................158 5.43 Optical micrographs showing the pullout of the PSu layers during fracture of the laminates....................................................................................................158 5.44 SEM images of brittle PSu layer originating from the HA/PSu interface. The arrow indicates the location of the origin.....................................................159 5.45 SEM images of a PSu layer fracture origin. The location of the origin is indicated by the arrows on both images, failure occurred within the PSu layer and not from the interface...........................................................................159 xvi

PAGE 17

5.46 Optical micrograph of the fracture origin (solid white arrows) of a 100100-1600 laminate, the large damaged area is caused by the indent (dashed arrow)...................................................................................................................161 6.1 Addition of the HA/PSu laminates designed in this study to the graph of available biomaterials for bone replacement developed by Suchanek and Yoshimura [6]......................................................................................................163 xvii

PAGE 18

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DESIGN, FABRICATION, AND CHARACTERIZATION OF LAMINATED HYDROXYAPATITE-POLYSULFONE COMPOSITES By Clifford Adams Wilson II August 2005 Chair: John J. Mecholsky Jr. Major Department: Materials Science and Engineering There exists a need to develop devices that can be used to replace hard tissues, such as bone, in load-bearing areas of the body. An ideal hard tissue replacement device is one that stimulates growth of natural tissues, and is slowly resorbed by the body. The implant is also required to have elastic modulus, strength, and toughness values similar to the tissues being replaced. Hydroxyapatite (HA) is the primary mineral phase of bone and has the potential for use in biomedical applications because it stimulates cell growth and is resorbable. Unfortunately, HA is a relatively low strength, low toughness material, which limits its application to only low load-bearing regions of the body. In order to apply HA to greater load-bearing areas of the body, strength and toughness must be improved through the formation of a composite structure. The goal of this study to show that a composite structure formed from HA and a biocompatible polymer can be fabricated with strength and toughness values that are xviii

PAGE 19

within the range necessary for load-bearing biomedical applications. Therefore, Polysulfone-HA composites were developed and tested. Polysulfone (PSu) is a hard, glassy polymer that has been shown to be biocompatible. Composites were fabricated through a combination of tape casting, solvent casting, and lamination. Monolithic HA and laminate specimens were tested in biaxial flexure. A unique laminate theory solution was developed to characterize stress distributions for laminates. Failure loads, failure stress, work of fracture, and apparent toughness were compared for the laminates against monolithic HA specimens. Initial testing results showed that laminates had a failure stress of 60 10, which is a 170% improvement over the 22 + 2 MPa failure stress for monolithic HA. The work of fracture was improved by 5500% from 11 2 for the monolithic HA to 612 240 for the laminates. Work of fracture values gave the laminates an apparent fracture toughness of 7.2 MPam 1/2 compared to 0.6 MPam 1/2 for the monolithic HA. Laminates with different geometries were built and tested in an attempt to optimize the strength and toughness of the composites. Laminate behavior was characterized as a function of initial flaw size, HA layer thickness, PSu layer thickness, and stressing rate. The failure stress of the laminates was maximized at a value of 108 14 MPa, which is a 400% improvement over monolithic HA, and close to the 120 160 MPa range reported for bone. The work of fracture of laminates was maximized at 724 206 J/m 2 which is a 6400% improvement over monolithic HA, and yields an apparent fracture toughness value of 7.5 MPam 1/2 This apparent toughness value is within the 2-12 MPam 1/2 range for bone, and an 1100% improvement over the fracture toughness of monolithic HA. xix

PAGE 20

CHAPTER 1 BACKGROUND AND RATIONALE Requirements for an Ideal Hard Tissue Regeneration Device Tissue Response to Implanted Materials Any material implanted in the body will illicit a response from the surrounding tissues. Bioceramics, when implanted in the body will elicit one of four potential tissue responses [8]. If the material is toxic, the surrounding tissues will die. If the material is nontoxic and biologically inactive, the body will surround the material with a fibrous capsule but not experience any negative consequences due to the presence of the implanted materials. A material that elicits this type of tissue response is called biocompatible. A third tissue response occurs when an implanted material is nontoxic and biologically active. These materials are called bioactive because they form an interfacial bond with surrounding tissues. The final tissue response occurs when a nontoxic implanted material dissolves in the body. These materials are called bioresorbable [8]. Definition of the Gold Standard In order for successful use of a material or composite for hard tissue regeneration, the material of composite must fulfill some basic requirements. These requirements are not explicit rules that govern research and development; rather they are ideals that have been set for the eventual gold standard. Through the study of hard tissues and hard tissue replacements by Hench [1-3], Bonfield [4], and others [5-7], the gold standard has been defined as a material, which will possess a few basic traits. The ideal hard tissue implant 1

PAGE 21

2 must be biocompatible, bioactive, and mechanically identical to the tissue being replaced. Additionally, the ideal hard tissue replacement would be resorbed and replaced by the surrounding natural tissue. Biocompatibility Biocompatibility will be defined as the ability of a material to perform a desired function in the body without causing any adverse effect upon natural tissue with which it comes into contact [8]. A material is required that will not cause any type of inflammatory or destructive response in the body. No matter how well a material performs the required mechanical functions, if there is an adverse response to the material within the body, the material is unacceptable for use. Biocompatibility is the greatest limiting factor in biomaterial design. If the entire spectrum of known materials was available for use, artificial bones and hard tissues replacement devices would probably already exist. However, materials that are toxic or reactive in the natural environment (for example, corrosion of metals over time when implanted in the body) cannot be used due to adverse effects on surrounding tissues. Therefore, research is limited to materials that are biocompatible. Biocompatible materials currently used for biomedical applications include metals, such as titanium, stainless steel and cobalt chromium, bioinert ceramics, such as alumina, zirconia, calcium phosphates and inorganic glasses, and a polymers like polymethyl methacrylate (PMMA), polyethylene (PE) polylactic acid (PLA), and polysulfone (PSu) [8]. These are just a few examples of biocompatible materials used in the body and should not be considered a comprehensive list since new materials are continually being developed and approved for use in the body.

PAGE 22

3 Bioactivity and Bioresorbtion Bioactivity refers to the ability of a material to promote regeneration of the natural tissues [9]. Regeneration is favored over replacement due to the long-term degradation of replacement materials, which causes a need for repetitive procedures over the life of a patient. Another reason for bioactivity is that is can be a useful tool in the implantation of a hard tissue device. The conclusion on the biochemical reactions occurring at the tissue/implant interface of research by Hench [1-3] and Suchanek [6] is that activation of natural tissue generation and the subsequent growth in and around the implant will aid in anchoring the implant into place. Furthermore, if bioactive materials are broken down by the surrounding environment, the body may then resorb it. Bioresorbable materials are designed to degrade gradually over a period of time, and be replaced by the natural host tissue [1]. There is a very narrow range of materials that are both biocompatible, bioactive, and bioresorbable and can be used for hard tissue replacement. This range of materials is currently limited to bioceramics such as calcium phosphates and bioactive glasses [2]. Matching Mechanical Properties The third major requirement of these materials is that they have mechanical properties that are similar to those of the natural tissue in which they are in contact. Problems arise from stress shielding when there is a mismatch in properties from one material to the next [8]. For example, replacing bone with a higher modulus material leads to a lower strength healed bone around the implant due to the implant carrying more of the load than the surrounding bone. Since bone remodeling is a continuous process that the body uses it to adapt to changes in stresses. For example, if a normally,

PAGE 23

4 inactive person were to start exercising on a regular basis, the body would remodel bones to compensate for the greater stresses felt during exercise. In the case of a higher modulus material being implanted to replace bone, the body would lay down weaker bone in the areas surrounding a higher modulus implant because the implant would bear most of the applied loads. There is also a need for implanted materials to have toughness values greater than those of the surrounding tissues, such that mechanical failures will not occur within the implanted materials before they occur in the surrounding tissue. A failure of the implant would lead to increased loading of the weakened, surrounding tissue and compound the problems that are being resolved. Materials Selection Hydroxyapatite Tissue regeneration is the most limiting factor in materials selection as there are few materials capable of inducing bone growth. One of the better-studied materials is hydroxyapatite (HA). Hydroxyapatite is the primary inorganic component of all calcified tissues existing in the human body [8]. Synthetic HA powders can be prepared in a number of different ways. Processes like sol-gel [6], precipitation [10], and solid-state reactions [11] have all been used to produce to produce HA powder. HA powder fabrication is a common process such that HA is readily available through a number of commercial supply companies. Numerous studies have show that synthetic HA is bioactive and can be used as a bone replacement [1, 6].

PAGE 24

5 HA containing a Ca:P ratio of 1.67 is bioactive [2, 6] and theoretically can be used for bone tissue replacement. Unfortunately, characterization of synthetically made HA shows that it has mechanical properties that are inferior to the necessary properties to sustain load-bearing applications in the body. Dense forms of HA have been fabricated, and a wide range of mechanical data have been reported. Bending strengths from 38-250 MPa, tensile strengths from 38-300 MPa, and fracture toughness values in the range of 0.8-1.2 MPam 1/2 have been reported [6]. Youngs Modulus values have been reported in the range of 10-30 GPa [5, 6]. Porous versions of HA have mechanical properties that are much lower [12, 13]. Bending and tensile strengths for these materials have been reported in range of 2-11 MPa and around 3 MPa, respectively [6]. Fracture toughness values have been shown to decrease with increasing porosity [13]. These materials are capable of matching the bending and tensile strengths of compact human bone, which have been reported to be as great as 160 MPa and 124-170 MPa, respectively [6]. However, the fracture toughness of HA is well below that reported for compact bone, which is in the range of 2-12 MPam 1/2 [1, 13] HA, on its own, lacks adequate mechanical properties to be applied to major load bearing applications. HA Composites Since monolithic HA lacks mechanical properties sufficient to withstand load-bearing applications within the body, HA must be combined into composite structures with other materials in order to meet the load-bearing requirements for bone replacement. The goal of forming the composite is to increase both the strength and toughness of HA to levels more consistent with that of natural bone.

PAGE 25

6 HA was been formed into a composite with titanium, zirconia, and alumina [6], as well as being reinforced through the addition of particles [14], whiskers [15], and fibers [6]. All of these materials have significant limitations, which limit their usefulness. The greatest problem is the modulus mismatch with bone. Figure 1.1, recreated from literature [6], shows the relationship of the mechanical properties of these materials to the mechanical properties of dense HA and to bone. Clearly these materials will need to be further refined for use as a hard tissue replacement. Figure 1.1 A plot of fracture toughness versus Youngs modulus of biomaterials being developed for bone replacement adapted from Suchanek and Yoshimura [6]. The only other materials incorporated with HA to form composite materials with modulus and toughness values similar to bone are polymers. A few of the polymers that have been incorporated successfully into the HA matrix include polyethylene, poly(L

PAGE 26

7 lactide), polysulfone as well as some biopolymers such as collagen [6, 8, 9]. Bonfield and his group have reported extensive work on PE/HA composites [4, 16-18]. HAPEX is an HA/PE composite that is FDA approved for use in the inner ear. HAPEX and other HA/polymer composites are produced by adding HA as the reinforcement to the polymers. HA/PE composites have an elastic modulus in the range of 1-8 GPa [6] and fracture toughness values that overlap that of bone, see Figure 1.1. Since polymers cannot withstand the temperature required to strengthen HA through sintering, HA/polymer composites are confined to forming through addition of HA particles into the polymer matrix. The resulting composites have greatly improved toughness values over monolithic HA but the high polymer content lowers the strength below that required for load-bearing applications. Clearly, another composite structure composed of HA and polymer is required. Biological Structures Fabrication of strong, tough materials has already been achieved many times over. Nature, in many ways, has taken materials that on their own might seem useless and combined them to create structures with extraordinary mechanical properties. Structures like mollusk shells [19], arthropod cuticles [20], and bone [5] are fabricated starting at the atomic level and built up to the macroscopic level. There are orientations within orientations, and levels of organization within level of organization. While science cannot yet make these materials on the same molecular size scale as nature, the development of materials on the nanometer scale may lead to the development of biomaterials that closely mimic mechanical structures found in nature.

PAGE 27

8 Nature produces strong composites through the incorporation of polymers in the form of proteins into composite structure with ordinarily brittle materials. Much research is being done to develop materials, which mimic the structure and behavior of natural materials [21]. Work has been done with conch shells [19], which are composed of CaCO 3 and proteins, have strength and toughness values that are orders of magnitude greater than monolithic CaCO 3 Conch shell achieves its superior strength, and toughness through a hierarchical structure composed of a macroscopic three layer laminate structure, with each layer having a second and third order hierarchy. Bone itself is composed of a complex hierarchical structure [5] that could be compared to a laminate with a thin cortical bone layer at the outer surface, and spongy trabecular bone in the middle. Bone consists of collagen molecules mineralized with hydroxyapatite crystals, which are grouped into fibrils, which are grouped into fibers, which then form lamellar structures called osteons. Osteons align themselves parallel to the long axis of bones. Looking at each osteon as an individual layer, bone can be described as a laminate structure. While it is not currently plausible to build composites with the same degree of sophistication as conch shell and bone, it is possible to incorporate the premise of a laminate composite. Laminate Composites Lamination is an effective method of combining materials into composite structures. Common ways of creating laminate structures are pressing at room or elevated temperatures, using an adhesive to bond the lamina together or combining materials that are mutually reactive to form chemical bonds. The resulting composites

PAGE 28

9 will have laminas joined either by a chemical bond or by a mechanical interlocking of the lamina surfaces. Past studies deal with combining ductile and brittle materials together through lamination [22-24]. It has been shown that through the lamination of brittle ceramics with ductile metals the toughness of the brittle ceramic can be increased [22, 25]. Some systems that have been developed with laminate structures are Bioglass/Copper [22], and Alumina/Nickel [24]. The mechanical properties of laminates are directly determined by the interface that is formed between the laminated materials [25]. If there is a weak interface between the constituents of the ductile-brittle laminate, then in the case of crack propagation, there will be evidence of delamination along the interface between the two materials. The result will be an unusable, low strength composite. A ductile-brittle laminate that has a strong interface will have a significant increase in toughness resulting from crack bridging or cracks arrest at the interface between the two materials. Strength and Biocompatibility of Polysulfone Polysulfone (PSu) is a tough, thermoplastic. PSu is used extensively for engineering applications because it has good thermal stability with degradation temperatures in excess of 450C. PSu has greater strength and elastic modulus than PE, which has facilitated research into replacing PE with PSu as the matrix material of HA/polymer particulate composites [9]. PSu films have been used as reinforcements of carbon-fiber-reinforced epoxies [26], as well as for filtration membranes. PSu films are made through solvent casting, or through phase inversion techniques.

PAGE 29

10 Polysulfone has been shown to be biocompatible as it is used as for blood filtration membranes [8]. PSu has been combined with bioactive glasses for a bone fixation study [27] and with epoxy fibers implanted into rabbits [28]. Both studies demonstrated good long-term stability of the PSu composites in vivo. In vitro studies of polysulfone have show negligible cellular response to implanted polysulfone [52]. Research Rationale and Objective The goal of this study is to develop a composite structure composed of hydroxyapatite and polysulfone with modulus, strength, and toughness values similar to those reported for bone. Reaching this goal will require the achievement of three specific objectives. The first objective is to design a fabrication technique for combining hydroxyapatite and polysulfone into a laminate structure. Achieving this objective requires first producing monolithic hydroxyapatite layers through a combination of tape casting, burnout, and sintering of a starting hydroxyapatite powder, followed by lamination with polysulfone through solvent casting, stacking, and lamination. Chapter 2 of this document covers development and optimization of a tape casting, burnout, and sintering methodology for hydroxyapatite, as well as characterization and mechanical testing of the monolithic HA. Chapter 3 covers procedures for fabrication of HA/PSu laminates, as well as characterization and optimization of these laminate fabrication procedures. Strength and toughness of the laminates will be characterized through loading in biaxial flexure, which leads to the second objective. The second objective of this study is to derive a mathematical model for describing stresses resulting from applied loads during biaxial flexure of laminated circular discs. Such a solution exists to describe the flexural behavior of laminated beams, but not for

PAGE 30

11 circular discs. The derivation will be validated through a comparison with finite element analysis of the HA/PSu laminates before being applied to calculation of laminate strengths. Derivation of the laminate theory solution, along with a comparison to finite element modeling is the subject of Chapter 4. The final objective will be the characterization of the strength and toughness of the laminates. The strength and toughness will be tested as a function of flaw sizes and individual layer thicknesses. Fractography will be performed on fractured laminates to determine the failure mechanisms. The strength and toughness of the laminates will be compared with monolithic HA as well as to the reported values for bone to gauge the success of the project in terms of producing a composite with properties similar to bone. Mechanical testing data and failure analysis of the laminates is presented in Chapter 5.

PAGE 31

CHAPTER 2 HYDROXYAPATITE Background on Processing of Hydroxyapatite A number of important factors must be taken into consideration when working with hydroxyapatite. The most important factor is the stoichiometric ratio of calcium to phosphorus. The correct Ca:P ratio for bioactive hydroxyapatite is 1.67:1 [2, 6]. Ratios less than 1.67 can lead to the formation of secondary calcium phosphate phases, while ratios exceeding 1.67 can lead to CaO formation [13]. The presence of CaO and Ca(OH) 2 leads to cracking during cooling due to differences in the coefficients of thermal contraction. There is also an accompanying decrease in mechanical strength. Sintering temperatures for hydroxyapatite must be closely controlled because hydroxyapatite is susceptible to undergoing a decomposition reaction at temperature, which is only slightly higher than the sintering temperatures. Sintering temperatures for HA range from between 1000-1250C [13]. Hydroxyapatite undergoes dehydroxylation at ~800C, which leads to a deficiency of OH ions within the crystal structure, which is remedied through rehydration during cooling. Hydroxyapatite is best sintered between 1000-1200C to achieve the highest densities, with density and porosity being highly dependent upon starting particle sizes [13, 29, 30]. Decomposition of hydroxyapatite occurs at a temperature between 1250-1450C and results predominantly in the formation of tricalcium phosphate (TCP), tetracalcium phosphates (TTCP), or calcium oxide (CaO) [13, 41]. A significant loss in strength accompanies the decomposition reaction. 12

PAGE 32

13 Starting Materials All hydroxyapatite specimens used in this study were fabricated from the same commercially available hydroxylapatite (Alfa Aesar, Ward Hill, MA) starting powder. The MSDS accompanying the hydroxylapatite gives it a molecular formula of Ca 10 (PO 4 ) 6 (OH) 2 which is the same as naturally occurring hydroxyapatite. This powder was selected because the distributor advertises the powder as possessing the correct 1.67:1 ratio of Ca:P necessary for hydroxyapatite to be bioactive [3]. Processing Methods Tape Casting The starting hydroxylapatite powder was fabricated into green tape using an Incetek tape casting machine (Integrated Ceramic Technologies, Inc., San Marcos,CA). The composition of the tape casting slurries can be found in Figure 2.1. Slurries were combined into polypropylene bottles, beginning with the addition of the organic solvents. Three organic solvents make up the 35 wt% of the slurry: methyl ethyl ketone (MEK), toluene, and ethanol (EtOH). These are added in a ratio of 6.25:5.25:1. Plasticizer (Santicizer S-160), dispersant (Blown Menhaden Fish Oil), organic solvents, and 10-15 alumina milling balls are combined all at once. The combination is swirled until the dispersant is completely suspended. Hydroxylapatite powder and binder (Butvar B-98) are combined apart from the solvent mixture and then gradually added. The bottle containing all constituents was wrapped in paraffin wax to minimize solvent evaporation, and ball milled for a minimum of 12 h. After ball milling, the slurries are immediately tape cast. Slurries were not filtered prior to casting because a skin forms rapidly once the slurries are removed from

PAGE 33

14 the ball mill. Tape casting was performed at a speed of ~3.5 m/min, the best possible speed of the tape caster, and at ambient conditions. The doctor blade was set at a height of 200 m above the moving tape. Cast tapes were allowed to dry inside the tape casting machine for a period of 1-2 h to ensure complete drying, although after 15-20 min tapes are no longer tacky, and can be handled. Dried tapes had a thickness of 100-150 m. After 1-2 h, the green tape is removed from the tape casting machine and processed in to specimens for testing. Cold Pressing Green tapes are punched into discs using a 25mm diameter knife edge punch. 0.3 m (1 ft) of green tape yields 30-36 discs. Discs were stacked to a desired thickness and cold pressed in a graphite mold and die at a pressure of 3500 psi (24 MPa) using a hydraulic laboratory press (Model C, Fred S. Carver Inc., Menomonee Falls, Wis.). Pressure was applied in increments of 1000, 1000, 1500 psi, allowing the system to come to equilibrium after each pressure addition. Equilibrium was seen as being achieved once the applied pressure remained constant for a minimum of 30 seconds without any decrease in the pressure reading. Once the maximum pressure was applied, the system is allowed to come to a final equilibrium and held constant for 1 minute to ensure that equilibrium had been achieved. Organic Burnout and Densification Thermal analysis was performed on green tape samples using thermogravametric-differental thermal analysis (TG/DTA), (Seiko Systems, Model 302) in order to identify temperatures for removal of the organic constituents of tape casting slurries, Figure 2.3. Organic burnout was carried out in a muffle furnace (Model FA 1730, Thermolyne,

PAGE 34

15 051015202530354045HASolventsPlasticizerBinderDispersantWt % Figure 2.1 Weight % of each constituent in the tape casting slurry. Figure 2.2 Green tape processing, shown are punched green discs (top left), the 25 mm (1) knife edge punch (middle), and scrap tape (top right). The section of tape processed (bottom) is 30 cm (12) x 11 cm (4.5) and produced 36 green discs.

PAGE 35

16 0200400600800100001224364Time (Hours)Temperature (C) 8 Figure 2.3 Organic burnout cycle performed on the consolidated discs produced by cold pressing hydroxyapatite green tape. Dubuque, IA) with a programmable digital controller. The pressed hydroxylapatite discs were placed onto furnace plates that had been dusted with alumina powder to prevent any possible adhesion of the hydroxylapatite specimens to the furnace plates. The rate of heating of 1C/min was constant for all heating ramps, while the rate of furnace cooling was approximately 0.5C/min. Removal of the organics occurs during the 3 h hold at 200C and the 6 h hold at 450C. The 1 h holds at 800C and 1000C are required for initial sintering of the hydroxyapatite in order to give the specimens enough mechanical strength to be handled for further processing. Specimens cooled from 450C, and not fired through these upper two hold temperatures were too brittle to be handled.

PAGE 36

17 Sintering After organic burnout, specimens were removed from the burnout furnace and placed in a second furnace. The second furnace was required since the burnout furnace was not capable of reaching the desired sintering temperatures. Specimens were placed in a drop down furnace (Del Tech Inc.). A short study was conducted to find the sintering temperature that produced hydroxyapatite specimens with the best combination of strength and processability. Seven specimen groups were sintered at temperatures of 1000, 1050, 1100, 1150, 1200, 1250, and 1300C. The ramp rate of 5C/min and hold time of 90 min were constant for each firing temperature. The mechanical properties (flexural strength, hardness, fracture toughness) of the firing study specimens were characterized and compared for the seven sintering temperatures. The sintering temperature that yielded the best combination of strength and processability was 1200C. To optimize the sintering process, testing was done on the effect of both the ramp rate and hold time of the sintering process. The effect of hold time was tested by sintering two specimen groups at 1200C with different programmed hold times of 1 and 10 h, and using a constant ramp rate of 5C/min. Three specimens groups were tested for the effect of ramp rate on mechanical properties. The three ramp rates tested were 1, 5, and 20C/min to 1200C with a constant hold time of 1 h. The final sintering program consists of a ramp rate of 5C/min with a hold time of 90 min at 1200. This firing program was used to process all monolithic hydroxyapatite specimens as well as the hydroxyapatite specimens later used to fabricated composites. The thickness of specimens was measured after sintering. The total reduction in thickness from the stacked green tapes was, on average, 55%.

PAGE 37

18 Specimen Refinement The post-sintered specimens were polished by hand to a desired testing thickness using 45 and 15 m diamond polishing wheels. In some cases specimens were indented, these specimens were polished though the 15 m finish. Specimens that were not indented were further polished using 5, 3, and 1 m alumina pastes to remove as many large surface flaws as possible. HA Characterization Methods Particle Size Analysis The particle size distribution and specific surface area of the starting hydroxylapatite powder was characterized using a laser diffraction particle size analyzer (Beckman Coulter, LS 13 320 series). Analysis was performed on the starting powder as it was delivered by the manufacturer. X-Ray Diffraction X-ray diffraction (XRD) was performed using an APD 3720 automated powder diffractometer (Phillips Electronic Instrument, Inc., Mahwah, NJ). XRD was performed on the as delivered hydroxylapatite staring powder, post-organic burnout 1000C sample, and post-sintering at 1200C sample. Measurements were taken for a range of 2 = 20-70, with a step size of 0.05. Density and Porosity The density of the monolithic hydroxyapatite specimens was measured using the Archimedes principle of volume displacement [31]. Specimens were dried at 150C for 12 h and then weighed in air. Specimens were then placed in distilled water and placed in a vacuum chamber for 30 min to ensure saturation. Specimens were then weighed

PAGE 38

19 both saturated, m sat and suspended, m sus The density was then calculated by first calculating the volume of water displaced, V dis by the suspended specimen: OHsussatdismmV2 (2.1) where the density of water was assumed to be 1.0 g/cm 3 The density of the HA specimens was then calculated through: disairHAVm (2.2) The total porosity of the specimens was then calculated through: totalporosityVVPorosity% (2.3) where the pore volume is OHdrysatporositymmV2 (2.4) and V total = V porosity + V specimen Elastic Modulus and Poissons Ratio The elastic modulus of the hydroxyapatite specimens was characterized using an ultrasonic technique [32]. The elastic modulus, E, was calculated from the equation: 1/4322spsspVVVVE (2.5) where is the specimen density, V p is the longitudinal wave velocity, and Vs is the shear wave velocity. Poissons ratio, was then calculated from the equation [32]: 12 G E (2.6) where the shear modulus, G, is calculated from the equation [32]:

PAGE 39

20 2sVG (2.7) Optical Microscopy and Lighting Effects The majority of the fractography and other microscopic analysis were performed using optical microscopy on an Olympus optical microscope. This microscopy proved difficult at times due to the nature of the hydroxyapatite being analyzed. Hydroxyapatite is a white, polycrystalline material that reflects and absorbs light. The color of this material made it particularly difficult to see surface features necessary for fractographic measurements. Two methods were used to overcome these difficulties. The first was to sputter coat the specimens with Au-Pd alloy. This method was used primarily on the most difficult specimens since it requires both time and equipment that can be expensive to use. The second solution was to switch from using overhead lighting to oblique lighting. The oblique lighting caused surface features to stand out (Figure 2.4), and made analysis possible in a more cost effective manner. Scanning Electron Microscopy Scanning Electron Microscopy (SEM) was used to characterize the microstructure of the hydroxyapatite, and to perform fractography on the fractured specimens. All SEM work was performed on a Joel 6400 scanning electron microscope. Mechanical Testing Methods Establishing Baseline Properties for Monolithic Hydroxyapatite Strength tests were performed to establish the baseline properties of monolithic HA to compare with composite materials. Specimens were tested using a method of strength indentation in which specimens were first indented for hardness characterization, and then fractured in biaxial loading.

PAGE 40

21 Figure 2.4 Oblique allowed surface features to be seen (bottom) in greater detail than with overhead lighting (top). Indentation Monolithic hydroxyapatite specimens were indented with Vickers indentations. Indentation loads < 2 kg were applied using a Micromet3 microhardness tester (Buehler LTD, Lake Bluff, IL), loads > 2 kg were applied using an older indentation tester (Zwick Inc.) The hardness of the specimens was calculated using the equation: 20018544.0dPH (2.8) where P is the indent load in N, and d is the half diagonal length of the indent in mm [33].

PAGE 41

22 Biaxial Flexure The monolithic hydroxyapatite specimens were loaded to failure in biaxial flexure using a piston on three-ball testing fixture. A loading piston with a diameter of 2.2 mm was selected for this study, and an effective ring radius produced by the three-ball support varied depending on the specimens being tested. Hydroxyapatite specimens fired to 1000C were tested with a support diameter of 23.2 mm. Specimens fired to 1200C had a small diameter due to shrinkage compared with the 1000C samples and were instead tested with a support diameter of 15 mm. The failure stress, f from bending in a piston and three-ball fixture can be calculated using the equation developed by Wachtman [34]: 222222111ln21413RaabbatPf (2.9) Where P is the failure load, is Poissons ratio, t is the specimen thickness, a is the support ring radius, b is the loading piston radius, and R is the specimen radius. Monolithic specimens were tested using a tensile testing machine (Instron) at a loading rate of 0.25 mm/min. Work of Fracture and Toughness The toughness, or absorbed energy at fracture, of the monolithic specimens was calculated from the area under the stress-strain curves generated during loading in biaxial flexure [35]: dT (2.10) the stress, was converted from the measured loads using equation 2.9, and the strain,

PAGE 42

23 was calculated from the equation for elastic modulus, E=/. The units for the toughness are J/m 3 The work of fracture was calculated from the area under the load-displacement curves using the equation [36, 37]: A Pdwof2 (2.11) where P is the fracture load, is the measured crosshead displacement, and A is the projected cross-sectional area of the created fracture surface. Work of fracture has units of J/m 2 Even though circular discs were fractured in biaxial flexure the cross-section of the fracture surface was assumed to be rectangular. The area of the fracture surface was dependent on the number of pieces created during the fracture process. Projected fracture surface areas were calculated by measuring the lengths of fractures and multiplied by the specimen thickness. Fracture Toughness The fracture toughness, K C for the monolithic specimens was calculated using three different techniques. The first technique involved using the fracture mechanics equation [38]: 2/1cYKfC (2.12) where f is the failure stress calculated from equation 2.9, c is the critical flaw size, and Y is a geometric constant equal to 1.65 for indented specimens and 1.24 for nonindented specimens. The critical flaw size was measured using fractography and 2/1)(abc (2.13) where a is the length of the semi-minor axis of the critical flaw, and 2b is the length of

PAGE 43

24 the semi-major axis, see Figure 2.5. Figure 2.5 is a schematic representation of fracture markings resulting from a brittle fracture process and is taken from Mecholsky, et al [53]. A detailed explanation of the different fracture markings can be found in this reference. Since HA is a polycrystalline ceramic it is difficult to distinguish many of these features. For this reason, only the size of the critical flaw was measured for fracture toughness calculations. The fracture markings that were measured are designated a cr and b cr in Figure 2.5. Figure 2.5 Schematic representation of fracture markings that result from a brittle fracture adapted from Mecholsky, et al [53]. The second method for calculating fracture toughness was through the strength indentation technique [39]. This technique calculates the fracture toughness using the equation: 4/33/18/1PHEKfRVSI (2.14)

PAGE 44

25 where E is the elastic modulus of the specimen, H is the hardness, f is the failure stress, P is the failure load, and v R is a geometrical constant equal to 0.59 +/0.12. Anstis [40] demonstrated that the fracture toughness values K C and K SI should yield comparable results for brittle materials. The strength indentation technique was used to verify calculations using the fracture mechanics equation, which relies on measuring the critical flaw size, which in some cases proved difficult due to the fracture behavior of the hydroxyapatite and complexity of the observed cracks. The strength indentation technique was also used to ensure that the fracture toughness was constant across a range of indent loads, and thus for different flaw sizes. A log-log graph of failure stress versus indent load should yield a straight line with a slope of -1/3. Deviation from the -1/3 slope would demonstrate that fracture toughness is not constant for increasing flaw sizes and that phenomena such as microstructural effects or R-curve behavior was occurring. However, as will be demonstrated later in this chapter, fracture toughness values for HA were constant with increasing flaw size. The third method used to calculate the fracture toughness from the work of fracture and elastic modulus using the relation [38]: wofWoFEK2 (2.15) Finite Element Analysis Finite element analysis was also performed to analyze the monolithic HA specimens. A detailed explanation of the analysis and the results are discussed in the Chapter 4.

PAGE 45

26 Statistical Analysis All statistical analysis was performed using DOE Fusion Pro software. The software is running ANOVA statistics. All statistical analysis was run at = 0.05. Results and Discussion Starting Powder Characterization The particle size distribution for the staring hydroxylapatite powder is shown in Figure 2.6. The starting hydroxylapatite powder shows a bimodal distribution. The use of a starting hydroxyapatite powder with a bimodal distribution has been reported previously [51] with the bimodal distribution attributed to the presence of large agglomerates. The starting powder was used in its as-received state and with no refinement to the particle size distribution. Since the goal of this study was to create a 00.511.522.533.544.505101520253035404550Particle Diameter (microns)Volume Percent Figure 2.6 Particle size distribution for the starting hydroxylapatite powder

PAGE 46

27 porous material that could allow cell infiltration, the particle size was left large in order to prevent packing during the processing steps. If the starting powder was refined to a smaller, more uniform particle size distribution the net result would be denser hydroxyapatite specimens. While the denser material would have increased flexural strength and toughness, the ideal result for this study would demonstrate that hydroxyapatite with a large amount of porosity can be fabricated to have mechanical properties within the range of bone. XRD spectrum for the starting powder is shown in Figure 2.7. The spectrum shows a hydroxylapatite starting powder with low crystallinity, and is composed of mostly tri-calcium and tetra-calcium phosphate secondary phases. Figure 2.7 XRD spectrum for the as-received hydroxylapatite powder

PAGE 47

28 Development of a Tape Casting Method for Hydroxylapatite Due to the large particle size distribution of the starting powder, and the type of tape casting machine being utilized developing a tape casting process for the hydroxylapatite proved a long and tedious trial and error process. The Incetek Model 104 tape caster has one exhaust port through which evaporating solvents are evacuated to a fume hood and a number of open seams around its doors and edges. These openings make it difficult to control airflow through the tape caster and thus the drying rate of the tape. Initial attempts were made to fabricate slurries with ethanol as the only organic solvent. However, the lower boiling point of ethanol (~78C) leads to a rapid evaporation rate and thus cracking of the green tape. To slow the evaporation rate, toluene was substituted for the ethanol since it has a higher boiling point of 110C. The binder, B-98, is only partially soluble in toluene, which led to undissolved binder in the tape casting slurries and large tape defects upon drying. To remedy this problem MEK was added to aid in dissolving the binder. It has a boiling temperature (80C) that is slightly higher than EtOH, which would also slow the drying process. The final slurry composition, which contains all three organic solvents, is a mixture which balances the need to dissolve the binder completely with evaporation rates that allow for drying of the green tape free of cracking due to overly rapid drying. Once the drying problems were overcome, tapes were cast and consolidated through cold pressing. The two factors, which affected the cold pressing process, were applied pressure and hold time. A number of applied pressures (with a constant hold time) and hold times (with a constant applied pressure) were tested for their effect on the overall consolidation of the stacked green tapes during, Figure 2.8. Consolidation was

PAGE 48

29 quantified by the measuring the percent reduction of total thickness of the stacked HA discs was reduced during cold pressing. The consolidation of the stacked green tapes occurred through plastic deformation of the polymeric binder. This deformation determines the proximity of HA particles to each other. The organic tape casting constituents are removed during burnout leaving behind only the HA particles. With other factors such as phase structure, particle composition, wetting, and heating programs being constant, the particle spacing following cold pressing will control the final density of the HA specimens after sintering. As shown in Figure 2.8, it was found that the reduction in thickness of the stacked green tapes was not dependent upon either the applied pressure or the hold time. These results indicated that deformation of the binder occurs almost instantaneously upon the applied pressure reaching the threshold at which deformation occurs. The idea of the binder undergoing instantaneous deformation was supported by the equilibration of the applied pressure within a few seconds of the final pressure being applied and remaining constant for the duration of the hold time. Since the total amount of force applied by the press remains constant unless acted upon by an outside agent (i.e. additional ram strokes), a drop in the pressure of the system would indicate an increase in the area through plastic deformation. This deformation would result in an increase in the diameter of the discs. However, since the pressure comes to equilibrium after only a few seconds and remains constant all deformation must also occur during the initial seconds of applied pressure. The fact that the reduction in thickness remains constant for the pressure tested indicates that the threshold for deforming the polymer binder lies below the lowest pressure tested.

PAGE 49

30 During the early stages of this study, a large amount of binder was added to the tape casting slurries. Some of the first experimental slurries contained as much as 25-30 wt% binder. The strategy was to prevent cracking due to drying stresses through the addition of large quantities of binder. This represented a brute force attempt for overcoming the formation of cracks in the dried tapes by having more of the ductile binder material in the system to absorb the drying stresses and effectively toughen the green tapes. The large amount of binder was successful in preventing cracks; however, it caused the manifestation of even larger defects during burnout and sintering. The larger amount of binder increased the distance between HA particles in the dried tapes resulting in large defects during burnout and sintering. Blowholes, delamination (Fig 2.9), and large cracks resulted from particles not being in close enough proximity to bond during the firing processes. The survival rate of these early trials was at best 40-45%. Some experimental tapes had less than 10% of specimens survive the burnout process. Through numerous trials with experimental slurries and burnout processes it was deduced that poor survivability of specimens was due entirely to the HA particle separation being too great due to the large amount of binder. The solution to this problem was to reduce the HA particle separation thorough increased deformation of the binder during consolidation. To do this the starting slurry composition was altered to include more plasticizer, making the binder more ductile, and allowing for greater deformation of the binder during cold pressing. It was also discovered that with the binder in a more plasticized form, less of it was required to overcome drying stresses. The effect of both the binder:plasticizer ratio and total amount of binder were tested to identify the combination that provided the largest reduction in thickness, and thus the most

PAGE 50

31 Figure 2.8 Effects of applied pressure (top) and hold time (bottom) on the reduction in thickness of stacked green tapes during cold pressing.

PAGE 51

32 Figure 2.9 Optical micrographs of post-cold pressing edge delamination (top left), delamination at the center (bottom left), full consolidation (top right), and post-sintering center delamination. consolidation of the green tapes. The results of these tests, Figure 2.10, led to a decreased amount of binder within the tape casting slurries, and an increase in the amount of plasticizer to a level equal to a binder:plasticizer ratio of 1:3 by weight. The results were tapes with increased ductility, and greater deformation during cold pressing due to the now highly plasticized nature of the binder, which decreased the separation between HA particles to a level, which allowed for better densification of the specimens during firing. The net result was a survival rate of greater than 95% for the organic burnout firing process. Reduction in thicknesses were also measured before and after firing to see how great an effect firing had on the consolidation of the stacked green tapes. The reduction

PAGE 52

33 in thicknesses during both cold pressing and firing were then combined to quantify the total reduction in thickness from the starting thickness of the stacked green tapes to the post-fired thickness of the specimens. The total thickness reduction could then be used as a design tool to estimate the necessary amount of stacked green tapes required to achieve a desired post-firing specimen thickness. Two other aspects should be noted for the tape casting method developed here. First, the slurry composition in its final form, as described in the methods section of this chapter, can be used for any desired slurry volume. The slurry composition shown in Figure 2.1 can be used in small volumes for tape 1 m, moderate volumes for 3-5 m of tape, which is the limit of the research laboratory, or theoretically for large volumes necessary for mass production. Second, the addition of all slurry components prior to milling is somewhat unconventional. Typically, slurries would be produced by first completely dissolving binder in solvent, and then adding the plasticizers. Separately, the ceramic powder would be suspended in solvent, dispersant would be added, and the powder-dispersant system would be milled to allow the dispersant to disperse the ceramic particles. Finally, the binder solution would be added, and the entire system milled to allow the binder to infiltrate between the dispersant-coated particles. The binder would be bonded to the dispersant, and this is how the tape would be held together. This is the more conventional stepwise approach to slurry formation. However, in all the attempts made in the early portion of this study to utilize the conventional slurry formation process, a usable tape was never processed. The lack of success with the conventional process facilitated the development of the less conventional all at once approach applied here.

PAGE 53

34 Figure 2.10 The reduction in thickness of the stacked green tapes is plotted versus wt% binder in the tape casting slurries (top), and binder:plasticizer ratio (bottom).

PAGE 54

35 Burnout and Sintering Process The burnout and sintering process performed on the consolidated green tapes was determined after performing TG/DTA on samples of the green tape. TG/DTA was also performed on specimens following the burnout process to verify that all organic, tape casting materials were successfully removed. The resulting curve, Figure 2.11, showed that all the organic constituents burned out at temperatures at or below 450C. The green tape TG/DTA curve shows a large volume of material being lost between 200-250C. This represents the removal of binder and plasticizer as the binder has a glass transition temperature in the range of 140-200C and the plasticizer has a boiling point of 240C according to the material safety data sheets (MSDS) provided by the manufacturer. The remaining dispersant was removed at a temperature of 280C. The burnout program derived from the TG/DTA data contained two holds. The first 2 h hold at 250C was added to slow the burnout process and allow degradation products to diffuse out of the specimens. This first hold was essential to the survival of specimens during firing. Without it, the degradation products would build up too rapidly within the specimens building up internal pressure sufficient to cause blowholes to form, thus destroying the specimens. The additional 6 h hold at 450 ensures sufficient time for organics removal from the HA specimens. The TG/DTA curve for the post-burnout sample shows that the burnout process successfully removes the organic tape casting components. The small decrease in weight can be attributed to the loss of water bound within the HA crystal structure that was released at elevated temperatures [13].

PAGE 55

36 Characterization and Mechanical Properties of Post-Burnout Hydroxyapatite Following the burnout and partial sintering process used to remove the organic tape casting additives, samples were characterized using XRD to show the effect of the firing program on the crystallinity and existing phases in comparison to the stating powder, Figures 2.12-2.13. The XRD spectrum for the post-burnout HA shows that firing the starting powder to 1000C increases the degree of crystallinity and causes the transformation of the numerous secondary phases into a more homogenous material that is very similar to XRD spectra for naturally occurring form of HA [8], Figure 2.13. The only difference between the 1000C HA spectrum and naturally occurring HA spectra is a small peak at a 2 value between 37 and 38. This peak is not found in the XRD spectrum for the starting powder and is therefore formed during the firing processes. This peak represents the formation of a CaO during the firing process, and this is supported by a similar identification of the peak by another group [41]. Additional XRD was performed on specimens sintered at 1200C. The XRD spectrums for these specimens sintered at 1200C closely resemble the spectrums for the1000C specimens with the only difference being the intensity of a few of the peeks. The result shows that the second firing of the specimens does not cause any phase transformations that would result in the existence of secondary calcium phosphate phases. The result of the firing processes as explained through the XRD characterization is the conversion of a starting powder that contains secondary calcium phosphates into a more homogenous hydroxyapatite phase that closely resembles naturally occurring hydroxyapatite and is stable at the sintering temperatures used in this study.

PAGE 56

37 Figure 2.11 TG/DTA data curves for a pre-burnout green tape sample (top) and a post-burnout hydroxyapatite specimen.

PAGE 57

38 Problems with Specimens Fired to 1000C Initially it was proposed that the hydroxyapatite specimens would undergo only one firing process effectively minimizing the total energy required to produce testable specimens. A single firing at a relatively low temperature of 1000C would ensure a large amount of porosity within the system, which would be necessary for future use in biomedical applications. However, it became clear very quickly that a second sintering process was a necessity for specimen fabrication. Less than 10% of the burned out specimens survived the necessary polishing and refinement in order to be used for mechanical testing. Of the 24 samples that could be burned out at one time, only 1-2 samples usually survived to reach mechanical testing, the rest fractured during polishing, most often at the outset. Thus it was determined that in order to conserve starting materials, and produce the number of specimens necessary for completion of this study that the one firing route would have to be abandoned for a two firing system that would yield specimens with greater mechanical stability. Therefore, a short firing study was performed to establish the best temperature that would yield the most usable specimens. Usable was defined as having the greatest percentage of specimens survive the polishing process. Mechanical Properties of 1000C Specimens A total of 29 specimens were tested after burnout at 1000C. These specimens were polished to a thickness of 2 mm, and indented prior to loading in biaxial flexure. Indentation controlled flaw sizes were produced in order to lessen the number of specimens required for a statistical analysis. The six indent loads tested and the resulting mechanical properties are reported in Table 2.1. The diagonals of the Vickers

PAGE 58

39 Figure 2.12 XRD spectrum for HA specimens burned out at 1000C (top) and specimens sintered at 1200C (bottom)

PAGE 59

40 Figure 2.13 XRD spectrum for HA specimen fired to 1200C (top) and for naturally occurring HA (bottom) [8]. Circles () indicate peaks found on both spectra, the square () indicates the peak that is unique to the 1200C spectrum.

PAGE 60

41 impressions (Figure 2.14) were measured for hardness calculations prior to biaxial loading, with the hardness calculated using Equation 2.8. A statistical analysis ( = 0.05) showd indent load, as expected. Since the fracture toughness is constant for all specimks two son around the indents made flaw size measurements potentially inaccurate. An tween KC and KSI for the 1200C specimens later in this chaptthen the fracture toughness is constant for the range of indent loads tested. This is ed that there was no significant difference in the hardness values for the six tested indent loads. Stress at failure was calculated from the experimentally measured failure loads anEquation 2.9. A statistical analysis ( = 0.05) showed that failure stresses were statistically different for the range of indent loads used, i.e. a decrease in failure stresswith increasing ens, a larger indent load results in a larger initial flaw and consequently lower failure stress. Following failure in biaxial loading, fractures were examined to ensure that cracoriginated from the indent flaw Figure 2.14, and these flaw sizes were measured using fractography. Fracture toughness was calculated using both the strength indentation method, Equation 2.14, and direct flaw size measurement, Equation 2.12. Thesevalues should coincide with each other, but do not for the 1000C samples. The reafor the difference shown in Table 2.1 is flaw size measurements for the 1000C specimens were difficult due to the very low toughness of the material. Localized crushing 80% correlation will be shown be er. Constant Fracture Toughness Figure 2.15 is a log-log plot of the failure stress, of, versus indent load P. As discussed in the materials and methods section, if the slope of such a plot is equal to -0.33

PAGE 61

42 important because deviation from this slope indicates that other factors must be taken into consideration when calculating the fracture toughness. A best-fit line through the six da ta oints, factoring in standard deviation not shown on Figure 2.15, had a slope of -0.35. Table 2.1 Mechanical data aeats for HA specimens fired to 00Load n Hardness (GPa) F Load Stress (MPa) Initial KC(M2) KSI (M) p nd fractography mailure suremen 1 0C Indent (kgf) (N) Failure Flaw Size (m) Pa-m 1/ Pa-m 1/2 0.50 5 0.16 0.03 31 6 10.2 0.9 194 46 0.23 0.04 0.27 0.02 1.00 5 0.13 0.02 27 3 8.2 0.9 234 40 0.20 0.03 0.28 0.02 2.00 6 0.18 0.03 24 3 7.5 1.0 284 37 0.02 0.01 0.31 0.02 3.35 5 0.15 0.02 21 2 6.6 0.7 379 145 0.21 0.04 0.32 0.02 4.35 6 0.15 0.02 21 2 6.7 0.7 398 43 0.22 0.03 0.34 0.03 7.35 2 0.16 0.01 23 1 7.1 0.2 366 42 0.22 0.01 0.41 0.01 Figure 2.14 Optical micrographs of a Vickers indent for a 1000C specimen (left) and crack propagation through a Vickers indent.

PAGE 62

43 Figure 2.15 Log-log plot of failure stress versus indent load for HA specimens fired to 1000C Firing Study In order to improve the mechanical strength of the monolithic hydroxyapatite the specimens burned out at 1000C were then sintered at a higher temperature. A range of elevated sintering temperatures was tested to see their effects of the mechanical strength, density, and hardness. The selected sintering temperatures included 1050, 1100, 1150, 1200, 1250, 1300C. As with the 1000C treatment, all specimens were indented prior to loading in biaxial flexure, and for comparison purposes all specimens were indented with a 4.35 kg load. The 4.35 kg indent load was selected to ensure that fracture occurred from the indentation flaw. Six specimens were tested at each temperature except for 1300C. Only three of the specimens sintered at 1300C survived the fabrication processes. Most specimens fractured under the loads applied during polishing.

PAGE 63

44 Sintering Temperature Effect on Hardness The diagonals of Vickers indent impression were measured and the hardness calculated from Equation 2.8. Hardness values increased with increasing sintering temperature, Figure 2.16. The increase in hardness values associated with increases in sintering temperature has been shown to be the result of larger grains sizes [42]. Sintering Temperature Effect on Biaxial Flexure Strength The indented specimens were loaded in biaxial flexure and the failure load was recorded and used to calculate the failure stress of the specimens using Equation 2.9. The failure stress increased over the sintering temperature range from 1000C up to 1200C, as shown in Figure 2.17. The failure stress then remains statistically constant for a sintering temperature of 1250C. The 1300C specimens had an increased failure stress, however since only three specimens were tested it is difficult to identify the increase as a real occurrence or just as an anomaly resulting from the three strongest specimens surviving the fabrication process. Sintering Temperature Effect on Density The density of specimens fired at increasing sintering temperatures was measured using the Archimedes method described in the materials and methods section. The density of the hydroxyapatite increased with increasing sintering temperature up to a peak at 1200C, Figure 2.18, and then it decreased for the elevated firing temperatures. The decrease in density is a direct result of decomposition reactions, which lead to the formation of less dense calcium phosphates, calcium oxides, or oxyapatites and water as byproducts [13, 41].

PAGE 64

45 Figure 2.16 A plot of hardness versus sintering temperature for the hydroxyapatite specimens. 0510152025303595010001050110011501200125013001350Firing Temperature (oC)Failure Stress (MPa) 01234595010001050110011501200125013001350Temperature (oC)Hardness (GPa) Figure 2.17 A plot of failure stress versus sintering temperature for the hydroxyapatite specimens.

PAGE 65

46 Figure 2.18 Plot of density versus sintering temperature for hydroxyapatite specimens. Qualitative Determination of Hydroxyapatite Decomposition Hydroxyapatite has been shown to decompose into anhydrous calcium phosphates [13] at sintering temperatures between 1200-1450C depending on the characteristics of the starting HA powder. XRD analysis of specimens sintered at 1200C showed that decomposition had not occurred up to this temperature. However, a color change, most likely the result of some decomposition, starts to be seen at sintering temperatures at or above 1200C. A side-by-side comparison, Figure 2.19, of specimens showed that at 1200C specimens begin to take on a greenish hue, which darkens at 1250C. Specimens fired at 1300C, not pictured, take on a purple hue, indicating a further decomposition of hydroxyapatite into calcium phosphates.

PAGE 66

47 Figure 2.19 Color comparison of specimens sintered at five different temperatures. Each row of specimens was fired at the temperature indicated to the right. Sintering Temperature Selection After analyzing the mechanical properties of hydroxyapatite specimens, it was determined, that the hydroxyapatite specimen used for composite formation would be sintered at a temperature of 1200C. There were three reasons for this selection. The most important was that XRD performed at 1200C showed that hydroxyapatite, in a form closely resembling natural HA, had not undergone any decomposition significant enough to be show up in the spectrum. The XRD analysis showed that the hydroxyapatite was stable through both the organic burnout process and the additional sintering firing at 1200C. The second reason for selecting 1200C as the sintering temperature was that it provided greatest reproducible failure stress. While potentially the 1300C firing temperature would produce stronger specimens, the inability to fabricate intact specimens made using this temperature unfeasible. It would be possible to identify the reason for

PAGE 67

48 the brittle nature of the 1300C specimens, but it was determined that such an analysis would be costly in terms of time and materials. The third reason for selecting the 1200C is that it represents the lowest temperature that could produce specimens in bulk quantities. Being the lowest temperature, sintering would require less energy and consequently less heating and cooling time to produce the number of specimens required for this study. The only draw back of the 1200C specimens is their density being a relative maximum for the sintered specimens. The total porosity estimated from the Archimedes method calculations was 35-40% of the total volume. While the project rationale for this project called for producing HA specimens with the greatest possible pore volume, as this is desirable for biomedical application, using more dense specimens sintered at 1200C was necessary for increased processing ability. Optimization of Sintering Once 1200C was established as the sintering temperature for this study, the sintering process was optimized to reduce processing time. The mechanical properties were tested as a function of ramping rates, and hold times. Specimens were fabricated, indented, and loaded in biaxial flexure. The failure stress was then compared for two hold times, Figure 2.20, and three ramping rates, Figure 2.21. The results showed that the failure stress was constant for holds times of 1 and 20 h. This was confirmed by a statistical analysis ( = 0.05) that showed that there was not benefit to increased hold times. It was assumed that testing of additional specimens at hold times within in the range between 1 h and 20 h would produce strength values consistent with the 1 h, and 20 h hold strengths. Therefore, no further testing was performed.

PAGE 68

49 Failure stress was shown to decrease with increasing ramping rates. This is mostly likely the result of an inability of the Deltech furnace to keep up with the higher heating rates and consequently the temperature reached was lower than the desired 1200C. It was determined that 5C/min was the fastest ramping rate that the Deltech furnace could handle and still ensure reaching the desired sintering temperature at the programmed time. Density and Elastic Modulus of 1200C Specimens The density of the 1200C specimens calculated from the Archimedes method was found to be 2.85 0.06 g/cm 3 The theoretical density of hydroxyapatite is between 3.14 3.20 g/cm 3 The calculated density is 10% less than the theoretical density value indicated the existence of significant porosity. The porosity of the 1200C HA was calculated to be 30-35%. The strength values for the specimens sintered at 1200C are consistent with hydroxyapatite specimens verified to have 30% porosity through helium pycnometry by another study [13]. It is assumed therefore that the HA specimens produced for this study possess 30% porosity, and that the accuracy of the density measurement is limited due to the low precision of the Archimedes method. SEM imaging of the HA microstructure, Figure 2.22, show that the 30% porosity measurement is plausible. Mercury porosimetry was attempted in order to better quantify the density and porosity of HA specimens, however at the time of writing no data was available. The elastic modulus of 10 hydroxyapatite specimens sintered at 1200C was measured using an ultrasonic technique. The specimens were randomly selected from different sintering batches to gain a better statistical average for later us in laminate composites calculations. The elastic modulus would be assumed constant for all of

PAGE 69

50 Figure 2.20 Plot of failure stress versus hold times at 1200C Figure 2.21 Plot of failure stress versus ramp rate sintering temperature of 1200C.

PAGE 70

51 the HA used in composite fabrication. The elastic modulus was measured at 56 3 GPa, with the lowest measured value being 52 GPa. The 52 GPa value will be used for all composite calculations as it represents the minimum possible modulus for the HA specimens developed. Figure 2.22 SEM micrograph of the microstructure of the hydroxyapatite specimens sintered at 1200C. Hardness of 1200C Specimens The hardness of the 1200C specimens was measured for the firing study, Figure 2.16. The hardness was measured at 1.4 0.7 GPa for the size of the Vickers indent impression. This value represents an approximately 1000% increase from the hardness of the 1000C specimens. This increase in hardness as a result of increased sintering temperature has been well established for HA [30, 41, 43].

PAGE 71

52 Biaxial Flexure Strength of 1200C Specimens The mechanical properties of specimens sintered at 1200C were characterized in the same manner as the 1000C specimens. Groups of six specimens were indented and loaded in biaxial flexure to fracture. Four groups of 1200C specimens were tested, three of these groups were indented with 3.35 kg loads, and one group was tested without indents. The three indented groups were polished to three different thicknesses; these were 1.5, 2.0, and 2.2 mm. Three thicknesses were tested for a failure load comparison with composites having a similar range of thicknesses. The non-indented sample group tested the inherent strength of the hydroxyapatite. Failure stresses of the 1200C specimens, Table 2.2, showed an increase of approximately 200% over the values measured for 1000C specimens with the same indent load (Table 2.1). Table 2.2 Biaxial flexure strength data for specimens sintered at 1200C. Group n ITFFailure Stress ndent Load (kg) hickness (mm) ailure Load (N) (MPa) 163131 .35 .5 1 3 9 2 2 6 3.35 2.0 58 6 22 2 3632702 .35 .2 11 4 4 46N283 -I .0 9 19 9 8 Figure 2.23 is a comparison of stress-strain curves for an indented specimen tic versus a non-indented specimen. The comparison shows approximately constant elasmodulus as the curves have identical initial slopes, with the elastic modulus being 64 GPa. This is greater than the calculated value from ultrasound, but well within the acceptable range for hydroxyapatite.

PAGE 72

53 Work of Fracture and Toughness Work of fracture values were calculated using Equation 2.11, and are reported in Table 2.3. Specimens fractured into either two or three pieces as shown in Figure 2.24. The length of all cracks was measured from fracture surface. All work of fracture integrations and toughness calculations were performed in Microsoft Excel spreadsheets. Work of fracture and toughness values were not calculated for Group 1 because most of the loading data was lost with only the greatest measured load and loading curves being saved on paper. Calculating the area under load-displacement curves by hand would have resulted in inaccurate data being included in the study. Figure 2.23 Comparison of stress-strain curves for indented and non-indented HA specimens. 010203040506000.00050.0010.00150.002Strain (mm/mm)Stress (MPa) Non-Indented 3.35kg Indent

PAGE 73

54 Figure 2.24 All specimens tested in biaxial flexure fractured into either two (left) or three (right) pieces. Table 2.3 Work of fracture values for specimens sintered at 1200C Group n Indent Load (kg) Work of Fracture (kJ/m 2 ) Absorbed Energy at Failure (kJ/m 3 ) 2 6 3.35 0.009 0.002 5 2 3 6 3.35 0.011 0.002 5 1 4 6 N-I 0.020 0.006 17 8 Fracture Toughness of 1200C Specimens Fracture toughness values were calculated using the three methods described in the Methods section of this chapter. The facture toughness values measured using all three methods are reported in Table 2.4. The measured values for the fracture toughness calculated from the size of the critical flaw, K C and from strength indentation, K SI are in reasonable agreement. It has been demonstrated that these two values should be equal, and a plot of these values against one another should yield a line with a slope of 1.0. A plot of the fracture toughness values measured in this study, Figure 2.25, produces a best-fit line with a slope of 0.8. The discrepancy most likely arises from the direct measurement of the flaw sizes and Equation 2.12 which assumes a particular flaw geometry resulting from indentation

PAGE 74

55 Table 2.4 Fracture toughness values for 1200C hydroxyapatite specimens Indent Loads (kg) Initial Flaw Size (m) K C (MPa-m 1/2 ) K SI (MPa-m 1/2 ) K wof (MPa-m 1/2 ) 3.35 313 39 0.59 0.66 1.0 Non-Indented 98 43 0.63 N/A 1.5 yielding a constant value of 1.65 for the geometric constant, Y. Fracture surfaces generated for the HA were often complex due to the porous nature of the HA microstructure and loading in biaxial flexure, and thus there is some experimental error involved in the fracture toughness calculation. This error accounts for the discrepancy between the two values for fracture toughness. Critical flaws for two of the 1200C specimens are shown in Figures 2.26 2.27. The fracture toughness value calculated from the work of fracture, K WoF is statistically greater ( = 0.05) than the other two values due to an underestimation of the fracture surface area generated during the fractures process. The projected area of the fracture surface was used for calculation of K WoF however an area of the fracture surface calculated on smaller length scales is required for an exact calculation of fracture toughness. Because of the area estimation, K WoF should be viewed as an apparent toughness value and not a direct measurement of fracture toughness. Monolithic Hydroxyapatite Specimens The focus of this study is on the ability of a composite structure to enhance mechanical properties. To this end, the role of the monolithic testing was to establish baseline mechanical properties for comparison to composite properties. The mechanical properties of the monolithic hydroxyapatite produced here were considered acceptable regardless of how they compared to previous work done with hydroxyapatite. The firing

PAGE 75

56 Figure 2.25 Plot of K C versus K SI for the hydroxyapatite specimens sintered at 1200C. Figure 2.26 SEM micrograph of the initial flaw caused by an indent (black bar) for a 1200C HA specimen. The outer boundary of the initial flaw (white arrows) and a twist-hackle marking (black arrow) are shown.

PAGE 76

57 Figure 2.27 Optical micrographs of an entire fracture surface (top) and initial flaw (bottom) of a 1200C specimen indented with a 3.35 kg indent. The white arrows indicate the outer edge of the critical flaw.

PAGE 77

58 study and analysis performed to ensure that hydroxyapatite specimens could be produced in sufficient quantities for composite formation and that fabrication would be both efficient and reproducible. There are many possible ways to further refine the hydroxyapatite and increase the mechanical properties. The optimization of the hydroxyapatite prior to composite formation is a tangential aspect of this work that was left to future researchers.

PAGE 78

CHAPTER 3 LAMINATE FABRICATION Design and Nomenclature Laminate Design Symmetric laminates were designed with five total layers. The five layers consisted of two thin outer ceramic layers, each bonded to a polymer layer, and then a thick middle ceramic layer. This design is based on the principle that the stress at which brittle failure occurs in a ceramics is controlled by the fracture toughness of the ceramic and size of the initial flaw as shown by fracture mechanics [38]. Based on this principle it was hypothesized that thinner ceramic layers on the laminate surfaces would limit the potential size of the initial flaw and increase the failure strength of the laminate. The thicker middle ceramic provide the mechanical strength of the laminate. Polymer layers are positioned to increase laminate toughness through crack arrest and reinitiation. Similar toughening mechanisms have been shown to occur in ceramic/metal laminates [24]. Laminate Geometry and Nomenclature A laminate design with a five-layer structure of alternating HA and Polysulfone (PSu) layers in the order HA/PSu/HA/PSu/HA was selected for fabrication. Laminates with various geometries were fabricated for studying how different variables, such as individual layer thicknesses, influence mechanical behavior. The initial laminate design was 2.0 mm in total thickness, with a symmetric geometry such that individual layer 59

PAGE 79

60 thicknesses were 400 m/200 m/800 m/200m/400m. These laminates will be designated as 400-200-800. This nomenclature represents the individual layer thicknesses in microns in the order of outer HA layer/PSu layer/middle HA layer. Even though not explicitly designated through this naming scheme, all laminates for this study have five total layers. In later chapters, laminates with different geometries will be discussed using this same designation scheme. Outer HA LayerOuter HA LayerMiddle HA LayerPSu LayerPSu Layer Outer HA LayerOuter HA LayerMiddle HA LayerPSu LayerPSu Layer Figure 3.1 Schematic diagram of HA/PSu laminate Solvent Casting of Polysulfone Solvent casting, a.k.a. film casting or solution casting, as described by Allcock [44] is a simple technique requiring only a polymer and a solvent, which readily dissolves the selected polymer. Once the polymer has been dissolved, solutions can be poured onto any surface on which a polymer film or coating is desired. Drying occurs through evaporation of the solvents out of the polymer. The result is a polymer film coating on the desired surface. Casting of polysulfone by immersion precipitation was shown to produce porous membranes used for filtration [28]. This technique is analogous to the solvent casting technique, and indicates a high probability that solvent casting of PSu will produce films with some degree of porosity.

PAGE 80

61 Materials The polymer selected for this study was commercially available polysulfone (Udell 1700, Solvay Advanced Polymers Inc.). Polysulfone has the repeat unit shown in Figure 3.2. Data sheets provided by the distributor for the polysulfone listed it as having a molecular weight of 35000 g/mol, and a glass transition temperature of 180C. Molecular simulations and experimental values have demonstrated that polysulfone is an amorphous polymer with little long range order [45]. PSu is distributed in the form of small pellets. The term PSu pellet seen at times in this chapter refers to characterization performed on PSu in as received form. Figure 3.2 Polysulfone A small number of composites were also fabricated out of a sulfonated version of the polysulfone (sPSu), Figure 3.3. sPSu was prepared through a sulfonation reaction performed by the research group of Dr. Anthony Brennan of the Materials Science and Engineering Department at the University of Florida. sPSu was incorporated into this study because it can be used in association with coupling agents and therefore can be chemically bonded to the HA to increase the mechanical integrity of the PSu-HA interface. However, studying the effect coupling agents have on the mechanical behavior of the laminates has been left to future work. The only results of any significance gained from the sPSu will be shown in discussion on interfaces in Chapter 5.

PAGE 81

62 Figure 3.3 Sulfonated Polysulfone Methods Solvent Casting of Polysulfone Casting Solution As received PSu pellets were dried at 100C for 24 h before solution preparation. Solvent casting of polysulfone was performed by first preparing 10 wt% solutions of polysulfone (PSu) in trichloroethane (TCE). Solutions were prepared by dissolving 10 g of PSu pellets per 100 mL of TCE. Solutions were typically prepared in volumetric flasks and had a total volume of 250 mL. Once PSu pellets were added to the required volume of TCE the volumetric flasks were placed on a hot plate with magnetic stirring capability and the solutions were stirred on host plates for a period of 24 h, or longer as needed to ensure all PSu was dissolved. The 10 wt% solution was used for all polysulfone layer fabrication in this study. The original casting solutions used were furnished by the research group of Dr. Anthony Brennan of the Materials Science and Engineering Department at the University of Florida. Successful early attempts at casting the 10 wt% PSu solution on glass slides, coupled with the fact that higher weight percent solutions are difficult to produce due to

PAGE 82

63 solubility limits between the PSu and TCE, were the reasons that only one wt% casting solution was used in this study. Solvent casting of the sPSu was performed using the same techniques as the PSu, but instead of TCE, the sPSu was dissolved in dimethyl formamide. (DMF) Casting of Polysulfone Casting of PSu films was performed by pipeting fixed volumes of the 10 wt% PSu solution onto surfaces. The volume of solution deposited on the surface was controlled to within 0.05 mL. Through early tests, it was established that 0.0012 mL of casting solution was required per mm 2 of surface area being coated with the PSu film. This ratio of casting solution volume to surface area was sufficient to produce a 100 m thick PSu film on a surface. For laminate formation, the casting process was repeated as many times as necessary to build up the desired PSu layer thickness. After each PSu solution addition, the solution was dried at 70C for a period of 4-6 h, or until the surface of the PSu was no longer tacky to the touch. Great care was taken to ensure that specimens were dried on a level surface to avoid any variation in the thickness of the PSu layers being cast. Drying of PSu Films Drying of the PSu films was studied at temperatures of 70C, 150C, and 190C. 70C was the drying temperature recommended by the research group that provided the initial solutions, and 150C and 190C are temperature above and below the glass transition temperature of the PSu. Solvent removal through drying at these temperatures was characterized using TG/DTA and DSC. The elastic moduli, break stresses, and elongations at break for PSu films dried at each temperature were characterized through

PAGE 83

64 tensile tests conducted according to ASTM D638-02. The 70C drying temperature was used for the majority of laminates fabricated, with one group of laminates being fabricated using a 190C drying temperature. 70C was selected because characterization of the dried films showed no statistical variation between the elastic modulus and failure stress for the three drying temperatures. In addition, the group of laminates fabricated through the 190C drying temperatures was lower strength that laminates dried at 70C, as will be shown in the results section of this chapter. The 70C drying is also more time efficient because the higher drying temperatures require drying to be done in drying oven with controlled ramping rates to avoid rapid solvent diffusion out of the solutions. Theoretically drying at a lower temperature should take more time, but when the added time necessary to set up drying ovens and slowly ramp to a higher drying temperature is factored in, the lower drying temperature becomes more efficient. Specimen Preparation Tensile Testing Specimens PSu films for tensile tests were prepared by casting 1.0 mm thick films onto glass slides. From each glass, slide two dogbones were punched using a steel die with dimensions conforming to ASTM D1822-68. Laminate Specimen Preparation HA-PSu laminates were fabricated using three methods. These methods will be designated: the matching halves method, the bottom up method, and the prefabricated polymer layer method. Each of these methods was used to fabricate laminate specimens for this study. The matching halves method was the original method used to fabricate laminates. Drawbacks associated with this method lead to the development of the bottom

PAGE 84

65 up method, which was used to fabricate laminates designed with PSu layers 200 m. Again, this method has some drawbacks, which facilitated the development of the prefabricated polymer layer method. This third method was used to fabricate laminates designed with PSU layers > 200 m. All HA specimens used for laminate fabrication had been processed according to the procedure described in the Methods section of Chapter 2 and polished to a thickness of 0.80 mm. Laminate Fabrication Method 1: Matching Halves Method Figure 3.4 shows a schematic of this laminate fabrication technique. This technique begins by building PSu layers of equal thickness onto HA layers. In the case of the 400-200-800 laminates, a 100 m PSu layer was built onto two HA specimens. Layer thicknesses were controlled by measuring the change in thickness that resulted after each casting. Once the difference in thickness from the starting HA disc thickness to the new total thickness was measured at 100 m the matching halves are ready for fabrication. PSu layers were dried in order to remove as much of the TCE solvent as possible. Once dried, the PSu surfaces were re-wet with a small volume (< 0.1 mL) of the casting solution. The two wetted surfaces were then placed together and a 300 g weight was placed on top of the three-layer sandwich structure. The sandwich structure was then dried at 70C for 4-6 h. The next step was to build PSu layers on a third HA specimens and on one of the outer surfaces of the sandwich structure. Again, PSu layers were built to 100 m and dried at 70C. The dried layers were wetted, placed together, and the weight re-applied to the now five-layer structure. The specimen was then dried

PAGE 85

66 for a final time at 70C. Once dried the specimen was ready for polishing and mechanical testing. Laminate Fabrication Method 2: Bottom Up Method As shown in the schematic for the bottom up method, Figure 3.5, processing of laminates begins by building a PSu layer on a single HA disc, which forms one of the two outer layers of the laminate. Once the PSu is built to the designed thickness, it is wet with casting solution (< 0.1 ml), and the second HA layers in placed on the wetted surface, a weight is applied and the sandwich structure is dried at 70C. After drying, the second PSu layer is built onto what will become the middle HA layer of the finished laminate and the four-layer structure is dried at 70C. Following drying, the second PSu layer is wetted with casting solution and a third HA discs in placed onto the wetted surface. Weight is applied to the five-layer structure, and the entire laminate is dried at 70C. Once dried the laminate is ready for polishing and mechanical testing. Laminate Formation Method 3: Pre-fabricated PSu Layer Method As the name implies, the third method involves fabrication of PSu layers with the designed thickness prior to building laminates. A schematic of this method is shown in Figure 3.6. A casting solution in the required volume is pipetted into a 1 deep aluminum pan with a diameter greater than that of the HA discs. The increased volumes of casting solutions necessary to produce thicker PSu layers requires extended drying times to ensure that as much solvent as possible was removed from the PSu films. Subsequently, PSu films were dried for 48 h at 70C. The dried films were then trimmed to be the same diameter as the HA discs.

PAGE 86

67 Thin PSu layers (<50 m) were cast on two HA discs, and dried at 70C. Following drying, the two PSu layers were wetted with casting solution, a thick pre-fabricated PSu layer was placed between the two wetted surfaces, and a weight was applied to the sandwich structure. The structure was then dried at 70C. Again, thin PSu layers (< 50 m) were cast on a third HA disc, and on one of the HA layers of the sandwich structure. These layers were dried, wetted, and a second thick PSu layer was placed between them. A weight was again applied, and the five-layer structure was dried at 70C. Once dried, the finished laminate was ready for polishing and mechanical testing. Salvaging of HA Discs As with any processing techniques, there exists the potential for defect formation during the solvent casting process. The ease of the casting process means that any PSu layers with defects could be discarded and a new defect free layer cast. However, the HA discs used for laminate formation require many days of processing and refinement. On the rare occurrence that solvent casting was performed poorly, the valuable HA discs were left coated with a defective PSu layer and could not be used in such condition for continued laminate processing. The remedy for this situation was to reclaim the HA disc by dissolving away the PSu layer. Defective specimens were placed in chloroform until all visible PSu was dissolved away, then the HA discs were heated to 600C to burnout any residual PSu. HA discs were inspected for any macroscopic flaws, and were then reincorporated into laminate processing.

PAGE 87

68 Wt Wt Begin With 2 Polished HA Discs Casting Solution PipettedOnto HA Surfaces and DriedPSu Layers Wetted, Matching Halves Stacked with PSu Layers Touching, Weight Applied, and Sandwich Dried Step 1Step 2Casting Solution PipettedOnto HA Surfaces of Sandwich and a Third DiscStep 3Step 4Step 5Outer HA Surfaces Polished to Designed Thickness111112223223 PSu Layers Wetted, Matching Halves Stacked with PSu Layers Touching, Weight Applied, and Sandwich Dried Wt Wt Begin With 2 Polished HA Discs Casting Solution PipettedOnto HA Surfaces and DriedPSu Layers Wetted, Matching Halves Stacked with PSu Layers Touching, Weight Applied, and Sandwich Dried Step 1Step 2Casting Solution PipettedOnto HA Surfaces of Sandwich and a Third DiscStep 3Step 4Step 5Outer HA Surfaces Polished to Designed Thickness111112223223 Wt Wt Wt Wt Wt Wt Begin With 2 Polished HA Discs Casting Solution PipettedOnto HA Surfaces and DriedPSu Layers Wetted, Matching Halves Stacked with PSu Layers Touching, Weight Applied, and Sandwich Dried Step 1Step 2Casting Solution PipettedOnto HA Surfaces of Sandwich and a Third DiscStep 3Step 4Step 5Outer HA Surfaces Polished to Designed Thickness111112223223 PSu Layers Wetted, Matching Halves Stacked with PSu Layers Touching, Weight Applied, and Sandwich Dried Figure 3.4 Step-by-step schematic of laminate fabrication method 1: The Matching Halves Method

PAGE 88

69 Begin With 1 Polished HA DiscCasting Solution PipettedOnto HA Surface and Dried, Step Repeated Until PSu Layer Reaches Designed ThicknessPSu Layers Wetted, Second HA Disc Stacked on Top, Weight Applied, and Sandwich Dried Step 1 Step 2Casting Solution PipettedOnto HA Surfaces and Dried, Step Repeated Until PSu Layer Reaches Designed Thickness Step 3 Step 4Step 5Second PSu Layer Wetted, Third HA Disc Stacked on Top, Weight Applied, and Laminate DriedOuter HA Surfaces Polished to Designed Thickness 1 1 Wt 12 12 Wt 123 Begin With 1 Polished HA DiscCasting Solution PipettedOnto HA Surface and Dried, Step Repeated Until PSu Layer Reaches Designed ThicknessPSu Layers Wetted, Second HA Disc Stacked on Top, Weight Applied, and Sandwich Dried Step 1 Step 2Casting Solution PipettedOnto HA Surfaces and Dried, Step Repeated Until PSu Layer Reaches Designed Thickness Step 3 Step 4 Step 4Step 5Second PSu Layer Wetted, Third HA Disc Stacked on Top, Weight Applied, and Laminate DriedOuter HA Surfaces Polished to Designed Thickness 1 1 1 1 Wt 12 Wt Wt Wt 12 12 12 Wt 123 Wt Wt Wt 123 Figure 3.5 Step-by-step schematic of laminate fabrication method 2: The Bottom Up Method

PAGE 89

70 Figure 3.6 Step-by-step schematic of laminate fabrication method 3: The Pre-Fabricated PSu Layer Method Begin With 2 Polished HA Discs, and 2 Pre-Fabricated PSU Layers Casting Solution PipettedOnto HA Surfaces and DriedPSu Layer Wetted, Pre-Fab. PSu Layer Stacked On Top and Wetted, Second HA Disc -PSu Layer Side Down-Stacked on Top, Weight Applied, and Sandwich Dried Step 1Step 2Casting Solution PipettedOnto HA Surfaces of Sandwich and a Third HA Disc, and DriedStep 3 Step 4Step 5Second PSu Layer Wetted, Second Pre-Fab. PSu Layer Stacked On Top and Wetted, Third HA Disc PSu Layer Side Down-Stacked on Top, Weight Applied, and Laminate DriedOuter HA Surfaces Polished to Designed Thickness12 132 12 3 Wt 123 Wt 12345 5 4 Begin With 2 Polished HA Discs, and 2 Pre-Fabricated PSU Layers Casting Solution PipettedOnto HA Surfaces and DriedPSu Layer Wetted, Pre-Fab. PSu Layer Stacked On Top and Wetted, Second HA Disc -PSu Layer Side Down-Stacked on Top, Weight Applied, and Sandwich Dried Step 1Step 2Casting Solution PipettedOnto HA Surfaces of Sandwich and a Third HA Disc, and DriedStep 3 Step 4 Step 4Step 5Second PSu Layer Wetted, Second Pre-Fab. PSu Layer Stacked On Top and Wetted, Third HA Disc PSu Layer Side Down-Stacked on Top, Weight Applied, and Laminate DriedOuter HA Surfaces Polished to Designed Thickness12 132 132 12 3 Wt 123 Wt Wt 123 Wt 12345 Wt Wt 12345 5 4

PAGE 90

71 P At the start of fabrication, all HA discs have been polished to a thickness of 800 microns using a 600 grit diamond polishing wheel. After the final drying, the laminates possess three HA layers of equal thickness. The thickness of the entire laminate is measured, and then each outer HA layer is polished to the designed thickness. In the case of the 400-200-800 laminates, half the thickness of the outer HA layers is polished away. Polishing to the final thickness was done through 600 (45 m) and 1200 (15 m) grit diamond polishing wheels if they were to be indented with non-indented laminates polished through a 5 m alumina paste. To test the effect of surface finish on bonding, a set of laminates was prepared for which the surface of the middle HA discs was polished through the 5 m alumina paste. Failure loads and stresses of these laminates where compared to the same properties for middle HA discs polished with the 600 grit finish. Failure loads and stresses were determined using methodology that will be described in detail in the following two chapters. Characterization of PSu Films and Laminates Thermal Analysis of PSu Films TG/DTA was performed using the same equipment described in Chapter 2. As delivered pellets and solvent cast PSu films were tested at a heating rate of 4C/min to a temperature of 800C. Differential Scanning Calorimetry (DSC) was performed using a differential scanning calorimeter (Seiko Systems). Heating profiles were established through a temperature programs that included three heating and cooling cycles. The cycled olishing of Laminates

PAGE 91

72 temperature range was from 0C to 350C and heating was constant at a rate of 10C/min. Mechanical Testing of Laminates for laminates fabricated through each of the three ethodr Problems With the Solvent Caent casting technique leads to bubble formation within the ciated with solvent casting was drying stresses arising from shried ic oved difficult as the PSu layers would hold the fractured pieces together causing the crack to Failure loads were measured ms described using the biaxial flexure test described in Chapter 2. The process fothe converting these loads to stresses is described in detail in Chapters 3 and 4. Results and Discussion sting Methods General Casting Problems The nature of the solv dried PSu layers. In some cases while diffusing out of the drying layers, the solvent is trapped in pockets. When this occurs on the surface, these pockets can expand to form large bubbles, see Figure 3.7. Solvent pockets forming within a PSu layer between two HA discs form locations of stress concentration, which occasionally caused fracture of anHA layer, see Figure 3.7. The other problem asso nkage of the PSu layers due to solvent loss. The drying stresses were of sufficient magnitude to cause fracture of the HA discs. The result was either a completely fracturHA disc or peeling of the PSu layer off the HA discs (Figure 3.8). In one case of complete fracture of the HA disc, a critical flaw was identified through fractographanalysis of the fracture surface. The measured size of the flaw showed that a failure stress comparable to the biaxial flexure strength of the HA was reached. The major problem with this particular failure mode was that detection of the failed HA discs pr

PAGE 92

73 Figure 3.7 Optical micrographs of (left) a large surface bubble formed during laminate fabrication, and (right) a small bubble (black arrow) within a PSu layer that caused a HA layer to fracture (white arrows). Figure 3.8 Left: Optical micrograph of a PSu layer that peeled from the HA disc during drying. Right: Higher magnification image of the PSu layer (white arrow) that peeled due to fracture of the HA layer (black arrows) which is still bonded to the PSu layer.

PAGE 93

74 be undetectable through a macroscopic inspection. Once this failure mode was identified, all HA discs were inspected microscopically at each step of the laminate formation process to ensure that no fractured specimens were incorporated into laminates. The second failure mode, peeling of the PSu layer from the HA discs, was the more prevalent of the two types of failure cause by solvent casting. These failures also occurred by fracturing the HA. However, in this case the fracture tended to propagate in the direction of the PSu/HA interface. Figure 3.8 shows that this failure was not of the interface itself, but of the HA as evidenced by the thin HA layer still bonded to the PSu post failure. Although no direct evidence was found, it is most likely that these failures occurred along within the HA along the interface of what was formerly two green tape layers that were not fully bonded during sintering. Most of the post-firing delamination was seen closer to the center of the HA discs rather than near the outer surfaces. Through polishing of the specimens these delamination flaws were moved from the center to the surface of the HA discs, marking these defects the most plausible explanation for the peeling failure seen during solvent casting. These failures were prevented through casting of multiple thin PSu layers instead of one thick layer. The drying stress associated the thin layers was of lesser magnitude and consequently prevented failures of this type from occurring. Problems with the Matching Halves Method The matching halves method for fabricating laminates was the first method developed for laminate fabrication. This method is the most time efficient of the three methods, but turned out to have the largest drawback. This method requires half the processing time of the other methods because each PSu layers can be built

PAGE 94

75 simultaneously on matching HA specimens and then combined to form the designed PSu layer. This is a simple two-step process to producing each polymer layer. However, problems arose from the inability of gases, trapped when combining the two halves, to escape through the already dried layers. The result was the formation of a large open cavity at the center of the PSu layers. The cavity formed at the center because the wetted PSu layers dry the quickest at the outer edge where diffusion of the solvent into the ambient environment occurs almost immediately resulting in a less permeable skin around the edge. As the rest of the PSu layer dries, evaporating solvent and gases migrate to the center of PSu layer as this is where drying occurs the slowest and the PSu is still the most permeable. The resulting pressure forces separation of two halves of the PSu layer. The two layers dry separately and can trap a large pocket of evaporated solvent and air. This processing problem was discovered when specimens were indented prior to mechanical testing. Instead of indenting the outer HA surface, the indenter proceeded to penetrate through the entire outer HA layer resulting in a large hole, Figure 3.9. Shortly after indentation of these specimens, the distinct odor of TCE could be detected in the environment surrounding the indenter, indicating the release of the trapped solvent. Specimens were fractured despite these huge flaws, and analysis of the fracture surfaces revealed the large cavities at the center of the specimens, Figure 3.9. Thirty laminates prepared through this method were indented with 3.35kg Vickers indents. Hardness measurements, shown in Table 3.1, indicated that cavity formation occurred in 43% of the laminates. This percentage represented an unacceptable number of defective

PAGE 95

76 specimens produced through this method and facilitated the development a better laminate formation method. Table 3.1 Hardness of laminates made through the matching halves method. n ( N total = 30) Hardness (GPa) % (n/N total ) 17 1.20 0.05 57 13 0.05 0.03 43 Figure 3.9 Optical micrographs of a large hole that formed during indentation (left), and of a laminate with a defective PSu layer containing a large open cavity (right) Problems with the Bottom Up Method The bottom up method was developed to prevent the formation of a large cavity in the PSu layers. Laying down the complete PSu layer onto one HA disc provides a greater surface area for solvent diffusion. Drying of laminates fabricated through this method was always done with the newest added layer positioned at the top to allow for solvent

PAGE 96

77 evaporation through the new layer prior to drying. Hardness for 19 laminates made through this method indented with 3.35 kg Vickers indents was measured at 1.33 0.20 GPa. This value includes 100% of the laminates fabricated demonstrating that there was no longer the potential for large cavity formation during processing. SEM, Figure 3.10, taken after mechanical testing revealed that instead of one large open cavity at the middle the solvent formed many small cavities or bubbles throughout the layer. The drawback to this method was that it could only be used to fabricate laminates with PSu layers with thicknesses 200 m. The reason for this is that when the casting solution is pipetted onto the HA disc surface it forms a meniscus. As the PSu layer dries Figure 3.10 SEM image of bubbles (black arrows) which formed during laminate fabrication using the bottom up technique. it retains the rounded edge formed by the meniscus. Subsequent PSu castings to build up the thickness of the PSu layer result in an increase in the distance between the outer edge of the HA disc and the outer edge of the PSu surface layer. When the next HA disc was

PAGE 97

78 added, the round edge of the PSu layer led to an open area between the surfaces of the HA discs. When the outer HA layers are polished these open areas act a pivot points and concentrated stresses which leads to chipping of the outer HA layer (Figure 3.11). For PSu layers 200 m. The chipping occurs outside the diameter of the biaxial flexure Figure 3.11 Schematic drawing of chipping failures that occur when fabricating thick PSu layers through the bottom up method. support ring and thus does effect mechanical testing. However, with thicker polymer layers chipping can occur inside the diameter or the three-ball support. Chipping inside the support diameter was deemed unacceptable was this could result in lower values for failure loads and thus jeopardize the statistical integrity of the study. Problems with the Pre-fabricated PSu Layer method The method of pre-fabricating PSu layers was developed because it prevented the chipping problem associated with the bottom up method. The pre-fabricated PSu layers were trimmed to the exact diameter of the HA discs and thus there are no gaps formed at the edges. There are two drawbacks to this method. The first is the 48 h required to dry the PSu layers is equivalent to the time required to completely fabricate laminates through the other methods. Therefore this method is more time consuming that the other methods.

PAGE 98

79 The second drawback to this method is alignment of the stacked five layers during drying of the laminate. Due to both sides of the thick PSu layer being on contact with wetted surface the layers were susceptible to sliding if stacking and drying were not done on a completely level surface. The slightest angle would cause laminate layers to slide off-center and as such are unusable. When this occurred the defective laminates were placed in chloroform, the PSu layers dissolved, and the HA discs reused in other laminates. Thermal Analysis of Polysulfone Layers TG/DTA of Polysulfone A full TG/DTA curves generated for the as received PSu is shown in Figure 3.12. TG data was plotted as a percentage of the starting weight versus temperature. The full curve for the as received PSu shows the onset of degradation occurring at approximately 450C, with all remnants of the PSu completely burned away by 625C. TG curves for the solvent cast layers dried at the three different temperatures area shown versus the as received PSu in Figure 3.13. The TG curves were used to measure the amount of solvent retained by the PSu films after drying at different temperatures. The amount of solvent is quantified in Table 3.2. The retained solvent data shows that a Table 3.2 Weight percent solvent retained results for PSu layers fabricated at all three drying temperatures. Drying Condition Wt% Solvent Retained PSu, As Delivered < 1% 70 C 20% 150C 10% 190C 4%

PAGE 99

80 significant amount of solvent is trapped within the PSu films until the films are heated past the T g of the PSu at which point softening occurs and allows for diffusion of the remaining solvent out of the film. Tensile testing of this films which will be shown later in this chapter, proved that there was no benefit to increasing the drying temperature above 70C. Therefore, it is reasonable to assume that the PSu layers of the laminates contain approximately 20 wt% solvent. Consistency of Solvent Casting Once 70C was selected as the drying temperature of the solvent cast PSu films, TG/DTA was run on six sample films cast at different times to see, qualitatively, how much variation there is in this solvent casting process. The six curves are shown in Figure 3.14. The solvent casting process shows good consistency with there being only a 4% difference between extremes. The TG data suggests that the solvent casting technique has sufficient reliability to assume that trends which arise during mechanical testing are a function of laminate behavior an not due to variations in the solvent casting technique. TG/DTA Analysis of a Fractured PSu Layer One of the 400-200-800 laminates tested according to the parameters that will be discussed in the subsequent chapters has a fracture surface which demonstrated a considerable amount of delamination. Enough of a PSu layer was exposed to use for a TG/DTA run. Figure 3.15 is the curve that resulted from this run. The curve seems to confirm the idea that as the PSu layers dry, solvent segregated into small cavities or bubbles. The numerous drops in weight seen in the curve could possibly coincide with

PAGE 100

81 Figure 3.12 Complete TG curve for an as received PSu pellet. parison of TG curves PSu films dried at the three different temperatures. PSu Pellet 190C150C70C PSu Pellet 190C150C70C Figure 3.13 A com

PAGE 101

82 Figure 3.14 TG curves for six different PSu films dried at 70C. Figure 3.15 TG curve for a PSu sample taken from a delaminated region of a laminate fracture surface.

PAGE 102

83 the release of solvents from different cavities. As the PSu sample is being heated in the filmfrfailure mechanisms of the laminates, and for ore layers were not tested in TG/DTA. The tested layer could possibly be nothing more than an anomaly, but the TG laminate layers suggest that solvent segregation is occurring as has been described. DSC was performed on a PSu pellet and on a PSu film dried at 70C. Three PSu pellet are shown in Figure 3.16. The curves confirm that the as received PSu has a gorphous as there is no melting peak. DSC curves generated for the PSu films are shown in Figure 3.17. The initial TG/DTA, the outer surface of the PSu softens first with the middle softening last due to the principles of heat flowing into the system. The result is solvent pockets closer to the outer film surfaces are released first, followed by the pockets closer to the center. Thisbehavior was not seen in prior testing because those films were cast onto glass slides, which produce more homogenous films. Solvent pockets would be more prevalent within the laminate layers because there is a second HA disc stacked on top of the drying which acts as an additional barrier preventing solvent from escaping into the environment. Unfortunately, this data could not be reproduced as only this one sample was taken om an actual fracture surface. The value of the fracture surfaces is in explaining the this reason, m curve coupled with the SEM image, Figure 3.10, of numerous bubbles within the PSu DSC Analysis of PSu Films heating and cooling cycles were performed over the temperature range of 0 350 C, with the polymer specimens in crimped aluminum pans. DSC curves generated for the T around 180C and is am heating show an initial drop beginning at 50C, which is consistent with the PSu pellet,

PAGE 103

84 and a second drop that begins between 100-150C. This second peak is the result of solvent evaporation that increases the pressure inside the crimped pan, which is read asan increase in heat. This has been shown to occur else where when additives burnou t of a the inability to get an accurate reading of the Tg of the PSu films.ulus, stress at break, and 20. ted nsile odulus, and a greater elongation at break er. polymer matrix. The result is Once the solvent has completely evaporated, the second heating and cooling curves are consistent with the as received PSu. Tensile Testing of PSu Layers Tensile tests were performed on twenty-four total PSu specimens with eight specimens fabricated for each drying condition. The elastic mod elongation at break were compared for the three drying temperatures, Figures 3.18 3.There was no statistical difference (=0.05) between the mechanical properties of PSu films prepared at each of the three drying temperatures. The tensile testing demonstrathat strength of the PSu layers were not directly influenced by the amount of residual solvent contained within the layers. These results are an indication of segregation between the polymer chains and solvent molecules. If the solvent molecules were dispersed within the PSu chain then they would affectively plasticizer the PSu and tetesting would show the 70C to be of a lower m would be seen. Laminate Fabrication Effect on Biaxial Flexure Strength Laminates fabricated through each method were tested in biaxial flexure to establish which method and drying temperature would provide the greatest values for failure loads and failure stress. The process for calculation failure stress is detailed in the next chaptFigure 3.21 shows the loads to failure measured during biaxial flexure and the

PAGE 104

85 Figure 3.16 DSC curves generated through three heating-cooling cycles for an as received PSu pellet. Figure 3.17 DSC curves generated through three heating-cooling cycles for a PSu film dried at 70C.

PAGE 105

86 Figure 3.18 A plot of elastic modulus vs. drying temperature for PSu layers. Figure 3.19 A plot of break stress vs. drying temperature for PSu layers.

PAGE 106

87 Figure 3.20 A plot of elongation at break vs. drying temperature for the PSu layers. Figur and drying temperature. Units of the y-axis are different for each series. e 3.21 A plot of failure loads and stresses vs. laminate fabrication method

PAGE 107

88 corresponding failure stresses for the three laminate fabrication methods and for drying at 70C and 190C. Statistical analysis (=0.05) showed no difference between the failure load and failure stress values for fabrication methods 2 and 3 dried at 70C. Statistics showed that the load and failure values for these two methods were statistically different from the values for both method 1, the matching halves method, and for drying at 190C. The biaxial flexure results indicate that fabrication through methods 2 and 3 yields laminates with similar failure loads and stress and are superior to those of laminate fabricated through method 1. The methods 2 and 3 failure loads and failure stresses, which are 175% of those for method 1. The lesser properties of the method 1 laminates areal support of the PSu, the HA layers around the cavity regions are unable to sustain loads a result of the large cavities, which form within the PSu layers. Without the structur equivalent to those where the PSu is present. This is why only methods 2 and 3 were used for laminate fabrication in this study. Results also suggest that drying at 70C is preferable to drying at 190C, with failure loads and stresses for the 70C laminates being 130% of those for the 190C laminates. This suggests that the greater residual solvent content of layers dried at 70C drying is potentially beneficial to the mechanical properties of the laminates. The reason for better properties is that 70C is below the boiling temperature of TCE (b.p. = 110C) the drying process will be slow. As the TCE evaporates the viscosity of the PSu layer increases. With the slower drying process allows the PSu to maintain a lower viscosity longer, resulting in a greater depth of penetration into the HA discs. The greater penetration depth establishes a strong interface between the PSU and HA which are bonded through a

PAGE 108

89 mechanical interlocking. These stronger interfaces allow for greater stress transfer to the PSu layers during loading and result in the increased failure loads over the 190C drying.Middle Layer Surface Finish and Mechanical Properties Biaxial flexure testing of six laminates prepared with middle HA layers finely polished prior to lamination were compared with data gained from testing of laminates which had been prepared with the rougher 600 grit finish. A comparison of failure land failure stresses is shown in Figure 3.22. Statistical analysis (=0.05) showed thvalues for failure loads and stresses for the two polishing methods were different, with oads at the The ordinate axis has units are unique to each series plotted. polished specimens delaminated occurred along entire HA/PSu interfaces which Figure 3.22 A plot of laminate failure loads and stresses vs. surface polishing medium the 600 grit finishing having failure loads that were 150% of those laminates having the fine polished middle surfaces. Analysis of the fracture surfaces showed that the finer

PAGE 109

90 compromised the mechanical stability of the laminate. Initial debonding along the HA/PSu interface was sufficient to cause the PSu layer to bow out, or buckle; opening a large gap between the PSu layer and the middle HA layer. Th is result indicates that ere would be no benefit to using the sPSu without coupling agents, which could potentially strengthen the interface and change the mechanical behavior of the laminates. Since addition of coupling agents is not part of this study, no further testing of laminates with sPSu was performed. Table 3.3 Comparison of laminates fabricated with polymer layers composed of both PSu and sPSu Polymer Failure Load (N) Failure Stress (MPa) bonding of the laminate is through mechanical interlocking and not through chemical means. Comparison of PSu Laminate to sPSu Laminates A set of six laminates were fabricated with sPSu instead of PSu. Failure loads and stresses from loading in biaxial flexure are compared in table 3.3. The results show that th PSu 150 22 56 8 sPSu 135 15 50 5

PAGE 110

CHAPTER 4 LAMINATE THEORY Background nt of model. The strain gauge method requires experimental testing of specimens, and the finite element modeling requires access to modeling whiDevelopment of a Laminate Theory Solution minate solue developed fore discs loaded in biaxial flexure. Accomplishing this will be done through combining laminate theory derived for the bending of beams with bending moments derived for simply supported circular plates loaded at the center. Laminate theory as presented by Mallick [48] describes the bending mechanics for fiber-reinforced cross-ply laminate beams. This laminate theory can be applied to laminates that do not contain fibers by assuming that all fibers are of the same modulus as the surrounding matrix, and are aligned in the 0 direction. Making these two assumptions allows the laminate theory to describe the bending mechanics for laminate Stress Analysis for Laminated Discs No mathematical solution currently exists for converting experimental failure loads measured through biaxial flexure tests into stresses for laminated discs. There are only two used to calculate stresses for laminated discs. These are through attachmestrain gauges prior to testing [46] or through finite element modeling [47]. While these are effective ways for calculating stresses, they are more complex and time consuming than having a mathematical software, which can be expensive. For these reasons, a purely mathematical model, d predict fa ch coul ilure stresses and failure loads, is desired. A la tion will b laminat 91

PAGE 111

92 beams made out of the materials selected for this study. However, since the experimental design calls for biaxial flexure of laminated discs, not beams, the bending moments used in the laminate theory must be changed to etry. Bending moments for simple supded at the center as described by Timoshenko and Woinowsky-Krieound to be the most applicable to the ate solution for the laminated discs using these nt ation reflect this change in loading geom ported circular plates loa ger [49] were f desired solution. Derivation of a lamin two theories is presented in the following section. The final solution requires a six-stepcalculation to generate a stress profile for a laminate. Verification of Laminate Theory The validity of the derived laminate solution will be verified through a comparison with finite element analysis run on identical laminate geometries. Agreemebetween the laminate theory and finite element models will demonstrate that a propermathematical solution was derived. Methods Laminate Theory A schematic of a laminate showing the variables necessary for laminate theory calculations can be found in Figure 4.1. The first step of the laminate theory is calculof the stiffness matrix [Q] using equation 4.1 4.2[48]: 2221QQQQ 1211Q (4.1) 222111 EQQ, (4.2a)

PAGE 112

93 Laminate Midplane Z Figure 4.1 A schematic representati on of a laminate indicating the mathematical variables required for laminate theory calculations. z hk2k3k4k5 k1k6 Laminate Midplane Z z h Laminate Midplane Z z hk2k3k4k5 k1k6 2 2112 1 is Poissons ratio of the particular lamina being nalyzed. From this matrix, the bending matrix [D] can then be calculated through equation 4.34.4[48]: EQQ (4.2b) where E is the elastic modulus and a 1211DD 2121D DD (4.3) kkkzhQDD)()(112211 (4.4a) NLkkkhQ12)()(1112 kkkkkzhQhQ2)()(12)()(12 (4.4b) NLkDD1211212 and z = distance froigure 4.1. An equivalent value for the Poissons ratio of the composite can be calculated through equation 4.5: m the midplane of the laminate to the midplane of the kth layer, see F

PAGE 113

94 1112DDeq (4.5) The next step in the solution is calculating the moments of inertia, Mnd 4.6b respectively [49]: r and Mt, through equations 4.6a a r aPMrlog14 (4.6a) 1log14raPMt (4.6b) where P is the applied load, is Poisons ratio for the composite equal to eq a is the From the moments of inertia and the bending matrix, the curvatures, Kr and Kt, are calculated through rrMDK1 (4.7) ion 4.8[48]: trtrKKZwhere Z is the distance from the laminate mand differs from the previous value for z. Finally, from the strain values and the stiffness matrix [Q], the stresses, and can be calculated through matrix multiplication, equation 4.9 [48]: Q (4.9) support radius, and r is the radial distance from the center of loading. the matrix multiplication of equation 4.7 [48]: ttMK From the curvatures, the strains, r and t are calculated through the matrix multiplication of equat (4.8) idplane to the tensile surface of the kth layer, r t trtr

PAGE 114

95 The entire lamress values could be gained immediately if any parameters needed to be changed. A sample Finite element aankar in the Department of Mechanical and Aerospace Engineering of the University of Florida. FEA was performed using in the thickness heory calculations. The data gained from the FEA was a calculation of the maximum principal stress at the center of the tensile surface res being placed on a 400-400-800 HA/PSu laminate. The region showing the darkest red coloring on the bottom surface represents the area on the tensile surface of the laminate where the highest principal stress can be found. arameters were made constant for modeling. 5 verified through ultrasound, while performing modulus measurements, and assumed to be 0.33 for PSu. Biaxial loading fixture dimensions for the support ring radius of 7.5 mm, and the loading piston radius of 1.1 mm were constant for both models. inate theory calculation was setup in a Microsoft Excel spreadsheet so that st sheet for the 400-200-800 laminate can be found in Figure 4.2. Finite Element Analysis (FEA) nalysis was performed by the research group of Dr. Bhavani S the computer program ABAQUS, standard version 6.4-1. The model had 1533 nodes and 1440 total elements with 5 integration points throughout the element thickness. Each laminate layer had 4 elements direction. The type of elements used was axissymmetry 3-D solid 4 node elements. The FEA was performed to verify the accuracy of laminate t ulting from a 1 N load being applied, as it would be during loading in biaxial flexure. Figure 4.3 is an example of the graphical output gained fromFEA analysis of a 1 N applied load Material Modeling Parameters A number of materials and testing p Elastic modulus for HA and PSu were verified through previously described experiments at 52 GPa and 2.7 GPa respectively. Poissons ratio for HA was 0.2

PAGE 115

96 distribution for the 400-200-800 laminate Figure 4.2 The Microsoft Excel spreadsheet designed to calculate the laminate stress

PAGE 116

97 Initial Model Comparison The initial comparison of the laminate models, Table 4.1, indicated there was a discrepancy between the FEA solution and the laminate theory solution. The laminate theory underestimated the FEA value by 35%. Figure 4.3 Graphical representation of the stress field resulting from a 1 N applied load on a 400-400-800 HA/PSu laminate. Results and Discussion Comparison of FEA and Laminate Theory Models Monolithic HA Modeling Modeling was performed for a monolithic HA specimens because this is the simplest design for comparison. The monolithic HA specimen was modeled with a thickness of 1.6 mm. The monolithic specimens could be modeled numerous ways through laminate theory. To demonstrate that the laminate theory solution developed was valid, the monolithic HA was modeled as both a symmetric bilayer and a 200-200-800 laminate. The solution for both models should yield the same value and be comparable to FEA. Thus, the FEA results would be used to validate applying the laminate model to the HA/PSu laminates.

PAGE 117

98 Table 4.1 Initial comparison of laminate models to FEA for monolithic HA. Model Maximum Principal Stress Resulting From a 1 N Load FEA 0.51 MPa Laminate Theory Bi-layer 0.33 MPa Laminate Theory 400-200-800 0.33 MPa discrepancy was ultimately in thcalculated by the minate theory. It was assumed that the variable r in the moment of inertia equations, 4.6a and 4.6b, was the contact raas it is for biaxial flexure, from the center of loading at culated. Therefore, the radius of the loading piston in equate of Maximum Principal Stress Resulting Correction of Laminate Theory Many reasons for the discrepancy between the two models were analyzed. The e position where the stress was being la dius of the loading piston equation 2.8. However, r is actually the radial distance which the stress value is cal ions 4.6a and 4.6b was converted to a value of 0.2 mm, which proved to be thvalue that brought the laminate theory and FEA into agreement. A new comparison the models, Table 4.2, yielded values that were identical for all three models. Table 4.2 Comparison of corrected laminate models to FEA for monolithic HA. Model From a 1 N Load FEA 0.51 MPa Laminate Theory Bilayer 0.51 MPa Laminate Theory 400-200-800 0.51 MPa Moder racy in describing the behavior of ling of Laminates The corrected laminate model was shown to be accurate for calculating stresses fothe monolithic HA. It was then tested for its accu

PAGE 118

99 HA/PSu laminates. Eight laminate geometries were modeled through FEA fo r comparison with the laminate theory. The laminate would all be fabricated and testentally. These particular geere selected because they represented a wide range of total laminate thicknesses, ratios between the icknesses of the individual layers, and composite moduli, which were calculated using field radius was kept constant at 0.2 mm, as this Geometry (MPa) Difference geometries chosen ed experim ometries w th the rule of mixtures. The contact corrected value was established through the monolithic HA calculations. The results forthe selected geometries are shown in Table 4.3. A negative % difference indicates an underestimate by the laminate theory, while a positive % difference means an Table 4.3 Comparison of maximum principal stress calculated through FEA and Laminate Theory for various laminate geometries. Laminate Composite Modulus (GPa) Total Thickness (mm) FEA Laminate Theory (MPa) % 400-200-800 42 2.0 0.41 0.38 7% 400-50-800 44 1.8 0.48 0.47 2% 100-200-800 37 1.4 1.09 1.14 + 5% 200-200-800 40 1.6 0.69 0.70 +1 % 100-100-1600 47 2.0 0.41 0.41 0% 350-400-500 32 217% .0 0.52 0.43 400-400-800 36 2.4 12% 0.34 0.30 200-400-800 32 2.0 0.55 0.56 2% overestimation. On the average, laminate theory underestimated the stress from a 1N load by -3% from the FEA stress value. However, considerable differences exist for the 350-400-500, and 400-400-800 laminate geometries.

PAGE 119

100 In order for each laminate theory calculation to agree with FEA the radial contact distance, r, was varied until the correct value was established which brought the lamitheory into agreement with the FEA. The corrected radial distance for each lamgroup is shown in Table 4.4. The results show that 0.2 mm is a good approximation the contact radius for all laminate groups except the 350-400-500 and 400-400-800 groups. Table 4.4 Contact stress field radii for each laminate geometry that yields the same Laminate Geometry Composite (GPa) Total (mm) (MPa) Laminate (MPa) Corrected (mm) nate inate for results as the FEA Modulus Thickness FEA Theory Radial Distance Monolithic HA 52 2.0 0.51 0.51 0.2 0 400-200-800 42 2.0 0.41 0.41 0.14 400-50-800 44 1.8 0.48 0.48 0.18 100-200-800 37 1.4 1.09 1.09 0.25 200-200-800 40 1.6 0.69 0.690.22 100-100-1600 47 0.41 0.410 2.0 .19 350-400-500 32 2.0 0.52 0.52 0.07 400-400-800 36 2.4 0.34 0.34 0.10 200-400-800 32 2.0 0.55 0.55 0.22 The origin for the discrepancy of the 350-400-800 and 400-400-800 group is not kno No relaship or eqn could ived fr compomodulus, total thickness, or laminate geoies, whiuld yielradial dthatFEA sol to the lamte theoryion. Oothesis it the FEA does not properly model the viscoelastic nature of the PSu. The FEA assumes a well-bonded material and looks at each layer only as they affect the overall global wn for certain. tion uatio be der om the site metr ch wo d the istance correlates the ution ina solut ne hyp s tha

PAGE 120

101 behavior of the system. It would therefore be possible that the FEA over estimates the deflection resulting from the 1N load. Greater deflection would result in a greater principal stress at the center of the tensile surface. Th e laminate theory calculation deals with th lculations because the average discrepancy is only -3% and, at worst, this pproximation leads to an underestimation of the failure stresses for the laminates with Laminate Stress Distri tagete abo rewithin. This is done by adjusting the value of Z, which is the distance from the midplane of the laminate to the location where the stress value is desired, making it possible to derive a picture ofut the entire lamstressr the 400-800 ldistribat thelied forcenear aall th layers a two PSu layers, but is discontinuous at the iace be the Hd PSu. Thss distribution results from the lower elastic modulus of PSu compared to HA. A 1N load applieate results in deflection of the ate. Tflection itified terms of a uniform strain on the laminate. If each layer is assumed to behave e stiffness of each layer separately taking into account the contribution of each layer on a more localized scale. The localization of the laminate theory versus the global properties predicted by the FEA may be the reason for the discrepancies that arise withthe thicker PSu layers. Without a proven way of calculating the adjusted value for the contact radius of the loading piston the value of 0.2 mm was applied to all laminate theory ca a thicker PSu layers. bution of lamina One advan theory is the ility t quantify st sses at any point the laminate the stress distribution througho inate. The distribution fo 00-2 aminate is shown in Figure 4.4. The stress ution shows th app is li cross ree HA nd the nterf tween A an is stre d to the lamin lamin his de s quan in

PAGE 121

102 according to Hookes Law [35], = E, the same strain results in a greater stress felt within the higher modulus layers. applied to a 400-200-800 laminate. Laminate layers are drawn to scale, with the Figure 4.4 Stress distribution calculated using laminate theory resulting from a 1N load 2.00 thickness representing the outer HA layer tensile surface. Predicting Laminate Behavior Maximum Stresses Laminate stress distributions were refined for comparison of different laminate geometries. Six data points were plotted for each laminate geometry modeled using the laminate theory, Figure 4.5 (a). Three points were for the maximum tensile stresses found for the outer HA layer on the indent side of the laminate, the PSu layer on thetensile side of the laminate, and for the tensile side of the middle layer. The other three points were for the maximum compressive stresses of the middle HA layer, the second

PAGE 122

103 PSu layer, and the outer HA layer in contact with the loading piston. These points werethen connected, generating a smooth curve, Figure 4.5 (b), which shows how the stress values changed from layer to layer. To avoid confusion it should be pointed out is not the same as the stress-distribution curve shown in Figure 4.4. This graph showshow stress values change between th that this e individual layers of the laminate. These two plots contain entirely different information. A graph like this one was generated for all lam mamto see wvalues of all of the laminate geometries tested experimentally are shown in Table 4.5. It was predicted that the laminate with the minimum stress value at the middle layer would rimental failure loads. Trends the stress curves will be pared with experim inate geometries. Selected curves were drawn on the same graph in order to compare the stresses for different laminate geometries. Comparison of Maximum Stress Curves Stress curves were graphed for all laminates in order to see if the laminate theory tched the experimental behavior of the laminates, and could thus be used as a predictive tool. Experiments were designed to test the effect of each layer on the echanical behavior for the laminates. Stress curves were generated to show the effect of changing the thickness of the PSu layers (Figure 4.6), the outer HA layers (Figure 4.7), and the middle HA layer (Figure 4.8). Laminate theory was also used to generate stress curves for different laminate geometries with a constant thickness of 2 mm (Figure 4.9), hich laminate geometry produced the minimum stress value. The six stress result in the highest expe comental data in the laminate behavior chapter.

PAGE 123

104 (a) Figure 4.5 Maximum stresses for the 400-200-800 laminate as (a) a plot of individual points and (b) a graphed smoothed curve. (b)HAHAPSuPSu HAHA (a) (b)HAHAPSuPSu HAHA

PAGE 124

105 Maximum Tensile Stress and Lamina Ratios Laminate theory could also predict the effect of individual la yer ratios on the aximum stress at the tensile surface. Total laminate thickness was held constant at 2 mm. One additional layer was held constant while the thicknesses of the remaining two layers were varied. Stresses were plotted versus the ratio of the two varied layers. With the middle HA layer thickness held constant at 800 m, the thicknesses of the outer HA and PSu layers were tested over a range of 0 600 m. Layer thicknesses were varied according to the equation (2 x Outer HA thickness) + (2 x PSu layer thickness) = 1200 m. The resulting plot, Figure 4.10, showed the greatest stress values for the thinnest HA layers with stress values decreasing towards an asymptotic value of 0.32 MPa. 0.32 MPa value represents the stress from a 1 N load applied to a 2.0 mm monolithic HA specimen. This plot will be compared with experimental data in the next chapter to show the predictive capabilities of the laminate theory. If the PSu layer is given a constant thickness of 200 m, and the thickness of the HA layers varied according to the equation (2 x Outer HA thickness) + Middle HA thickness = 1600 m. The thickness of the outer HA layer ranged form 0 800 m corresponding to the middle HA layer thickness of 1600 0 m. A plot of the outer HA thickness to middle HA thickness ratio versus stress, Figure 4.11, shows decreasing stress values towards the asymptotic value of 0.32 MPa. This plot will also be compared with experimental data in the next chapter. Finally, the outer HA layer was held constant at 400 m, and the thickness of the middle HA layer and the PSu layer were varied according to the equation (2 x PSu thickness) + Middle HA thickness = 1200 m. The thickness of the PSu layer ranged m

PAGE 125

106 from 0 600 m corresponding to middle HA layer thickness of 1200 0 m. The plot the mbe comp Table 4.5 Maxim Laminate Geometry ression of PSu:Middle HA ratio versus tensile stress, Figure 4.12, showed an initial increase fromonolithic HA stress value of 0.32 to constant value of 0.41 MPa. This plot will also ared to experimental values in the next chapter. The ratio comparison show that contribution of the PSu layer to mechanical properties is dependent upon the thickness of the outer HA layer. When the HA layer was held constant at 400 m there is only a 9% difference in the maximum tensile stress um stresses on the tensile surface of each layer for all laminate geometries experimentally tested in this study. Outer HA Layer -Tensile Side PSu Layer -Tensile Side Middle HA Layer Tensile Side Middle HA Layer -Compression Side PSu Layer -Compression Side Outer HA Layer -CompSide 400-200-800 0.38 0.01 0.15 -0.15 -0.01 -0.38 400-50-800 0.47 0.01 0.22 -0.22 -0.01 -0.47 400-100-800 0.44 0.01 0.19 -0.19 -0.01 -0.44 100-200-800 1.14 0.06 0.65 -0.65 -0.06 -1.14 200-200-800 0.70 0.03 0.35 -0.35 -0.03 -0.70 100-100-1600 0.41 0.02 0.33 -0.33 -0.02 -0.41 100-200-1400 0.51 0.03 0.36 -0.36 -0.03 -0.51 200-100-1400 0.39 0.02 0.27 -0.27 -0.02 -0.39 200-200-1200 0.56 0.02 0.31 -0.31 -0.02 -0.56 400-400-800 0.30 0.01 0.10 -0.10 -0.01 -0.30 35 0-400-800 0.43 0.02 0.11 -0.11 -0.02 -0.43 200-400-800 0.56 0.03 0.22 -0.22 -0.03 -0.56

PAGE 126

107 Figure 4.6 Maximum stress curves for laminates as a function of varying polymer layerthickness. layer thickness. Figure 4.7 Maximum stress curves for laminates as a function of varying outer HA

PAGE 127

108 Figue 4.8 Mam stress curves for laminates as a function of varying midhigure 4.9 Maximum stress curves for various laminate geometries having a total inate thickness of 2.0 mm. r ximu dle HA layer t ickness. F lam

PAGE 128

109 Figure 4.10 A plot of maximum tensile stress values versus the ratio between the thicknesses of the outer HA and PSu layers. Figure 4.11 A plot of maximum tensile stress versus the ratio of thicknesses between the PSu and middle HA layers.

PAGE 129

110 outer and layers. value for the most extreme PSu layer thicknesses. However, when the outer HA layer to PSu layer ratio is < 2 the PSu layer has a significant impact on the maximum tensile stress value. An impact becomes increasingly meaningful as this ratio is decreased further. Thus, the ratio analysis reveals that there is a finite range over which the PSu will have the most significant impact on mechanical behavior Figure 4.12 A plot of maximum tensile stress versus the ratio of thickness between the middle HA

PAGE 130

CHAPTER 5 LAMINATE BEHAVIOR Methods Mechanical Testing and Characterization of Laminates Indentation As with the monolithic HA specimens, laminates were indented with Vickers indenprior to loading. The Vickers indent impressions were measured for calculating hardness ing equation 2.8. Hardness values similar to the monolithic HA would indic ts values usate goA layer due to the lower modulus polymer below. For laminates with very thin outer HA layers, larger Vickers impressions were expected as indentation was effectively taking p The failure sf od mechanical stability of the laminate, as there would be no deflection of the outer H lace on the low modulus PSu. Biaxial Flexure All laminate specimens were loaded to failure using the piston-on-three ball biaxial flexure fixture described previously in the Methods section of Chapter 2. tress, f, for the laminates was determined using the equation: LTP where P is applied load in Newtons measured experimentally, and LT is the maximpal stress (MPa/N) calculated using laminate theory. Since the laminate theory solution calculates both the stresses and strains for the laminates, stress-strain curv um princies were generated from the appliedlexure tests. loads measured during biaxial f 111

PAGE 131

112 Work of Fracture Work of fracture for the laminates was calculated using Equation 2.11. This samprocess was used for calculating the work of fracture for the monolithic HA samples. Projected fracture surface areas were calculated assuming rectangular cross-section, and depended on the number fracture pieces generated as described in Chapter 2. The work of fracture has units of J/m2. Toughness Toughness calculated from the areas under the stress-strain curves, and Equation 2.10, could be called the absorbed energy at fracture. In order for this value no to be confused with fracture toughness, the toughness calculated from the area under the stresse strain curve will be referrtress-strain curves used were generated from the experimentally measured applied loads through the lamithe mlamire tested to aid in explaining the mechanical behavior of the laminates. Loading Parameters Except where it is specified to the contrary, laminates were indented with the same VExcept for the loading rate testing, all biaxial flexure loading was performed at a crosshead displacement rate of 2.5 mm/min. ed to as the absorbed energy at failure. The s nate theory solution that has been developed. This is the same calculation used for onolithic HA specimens. Absorbed energy at fracture has units of J/m3. Laminate Testing Variables Testing Rationale After an initial comparison of mechanical properties between 400-200-800 nates and monolithic HA, a number of different laminate geometries and design variables we ickers indent load prior to loading in biaxial flexure. The indent load used was 3.35 kg.

PAGE 132

113 InitiaLamient indented using a 3.35 kg Vickers biaxial flexure. Four roups consisting of 5-7 laminates were tested at crosshead displacement rates of 0.025, s 00 and 800 m respectively. Laminate groups were fabricated with four different PSu layer thicknesses. The selected thicknesses of the PSu layers were 50, 100, 200, and 400 m. Outer HA Layer Thickness The effect of the outer HA layer thickness on the mechanical behavior of the laminates was tested by fabricating laminate groups with three different outer HA layer thicknesses. The PSu and middle HA layers were fabricated with constant thicknesses of l Flaw Size The effect of the initial flaw size of the outer HA layer was tested by indenting four groups of laminates with a range of indent loads prior to loading in biaxial flexure. nates used for these testing groups had geometry of 400-200-800, with testing groups consisting of 6-7 laminates. The four Vickers indent loads selected were 0, 1.35, 3.35, and 9.35 kg. Laminates were tested in biaxial flexure at a crosshead displacemrate of 2.5 mm/min. Loading Rate The effect of the viscoelastic PSu layers on laminate behavior was tested by loading groups of lam inates at different crosshead displacement rates. All laminates were indent load prior to loading in g 0.25, 2.5 and 25 mm/min. Laminates had a geometry of 400-200-800. Polysulfone Layer Thickness The effect of the PSu layer on the strength and toughness of the laminates watested by fabricating laminates with a range of PSu layer thicknesses. The outer and middle HA layers were fabricated to constant thicknesses of 4

PAGE 133

114 200 and 800 m respectively. The three outer HA layer thicknesses tested were 100, 200, and 400 m. Middle Layer Thickness The effect of middle HA nical behavior of the laminates was tested by fabricating four roups with varying middle layer nstant at 2.0 mm. The inate groups were 100-100-1600, 200-200-1200, 200-100-1400, and 100-200-1400. on the assumption that crack propagation the HA/PSu interface. The loads required initiate crack penough to initiate crack propagation. laminates ith geometry of 400-200-800. The mechanical properties of these laminates are layer thickness on the mecha laminate g thicknesses. The total thickness of the laminates was held co geometries of the four lam Highest Predicted Failure Load Specimens Two additional laminate groups were fabricated and tested. These laminate groups had geometries of 350-400-500, and 200-400-800. These geometries represented the laminate groups that were predicted to have the highest failure loads by the laminate theory model and the maximum stress curves for laminates with a total thickness of 2.0 mm, see Figure 4.8. The prediction is based through the outer HA layer will be arrested at to reropagation through the middle HA layer is predicted by the stress/N load at the tensile surface of the middle HA layer calculated using laminate theory. The lesser the stress/N value the greater the applied load must be to produce stresses large Results and Discussion Comparison of Laminates to Monoliths Initial testing of laminate behavior was performed on a group of seven w

PAGE 134

115 compared to monoli thic HA specimens of the same thickness (see Chapter 2). All r hardness value is the result of indenting on top of the us PSu layers. This is the same phenomenon, which produced large holes thic r red in this inates compared with the monolithic HA. The values for stresses, work of fracture, and toughness all dramatically increased for thies for e value for the laminates shown in Table 5.1 and equation 2.15. This r indentation and biaxial flexure testing were performed using the same parameters for the laminates as were used for the monoliths. The hardness of the laminates was calculated at 1.32 0.17 GPa. Statistical analysis ( =0.05) showed this value to be significantly lower than the hardness value of the monolithic HA. The lowe lower modul when indenting laminates made through the matching halves technique, see Chapter 3. The difference here is that PSu provide greater support than air, but still allowed for increased deflection at the HA surface being indented, and thus larger impressions are made leading to lower hardness values. A comparison of load displacement curves for the laminate and the monoliHA, Figure 5.1, shows the dramatic increase in load bearing capacity of the laminate ovethe monolithic HA. Hardness values were the only mechanical property measu study that decreased for the lam failure loads, failure e laminates, see Figure 5.2. The percentage increase of each of these propertthe 400-200-800 laminates over the monolithic HA are shown in Table 5.1. Apparent fracture toughness, K app value of 7.2 MPa-m 1/2 was calculated from the average work of fractur epresented a 620% increase from the K app value of 1.0 MPa-m 1/2 for the monolithic HA. K app for the laminates is well within the 2-12 MPa-m 1/2 range, which is reported for bone.

PAGE 135

116 The presen ce of the PSu resulted in an increase in strength and toughness of the HA. Table 5.1 Comparison of the average mechanical property values for the monolithic HA Load Stress Fracture Absorbed Failure The toughening mechanisms responsible for these increases are characterized through fractography, and will be discussed in detail later in this chapter. versus the 400-200-800 laminates. Failure Failure Work of Energy at Monolithic HA Indented 63 N 22 MPa 11 J/m25 kJ/m3 400-200-800 Laminates 150 N 56 MPa 620 J/m260 kJ/m3 Perc ent Increase 140% 154% 5500% 1100% Figure 5.1 Comparison of load displacement curves for the 400-200-800 laminates compared with monolithic HA.

PAGE 136

117 Figure 5.2 Comparison of mechanical properties measured for monolithic HA as the 400-ith units designated in the column headings. Figure 5.3 Comparison of failure stress for 400-200-800 laminates versus both indented monolithic HA specimens. 200-800 laminates. The units of the ordinate axis vary for each property listed w and non-indented

PAGE 137

118 Testing of Laminate Parameters Initial Flaw Size Effect Four groups of 400-200-800 lamina tes were tested with a range of initial flaw size produc r n the three indent loads, but that the hardness values Indent Load (kg) Diagonal length (m) Hardness (GPa) ed using indentation. Larger indent loads produced larger impressions, whichbecome fracture origins upon loading in biaxial flexure. Usually the length of radial cracks emanating from the corners of the indent would be measured to show that largeflaws were introduced into the material. However, due to the porous nature of the HA radial cracks were not visible through optical microscopy prior to loading in biaxial flexure. Diagonal lengths and hardness values calculated for the indented groups areshown in Table 5.2. Statistical analysis (=0.05) showed that there was no significant difference in the hardness values betwee were all statistically different from the hardness value of the monolithic HA. While the 9.35 kg indents showed the same hardness values as the smaller indent loads, there was significant lateral cracking associated with this largest indent load, Figure 5.4. Table 5.2 Hardness data for initial flaw size laminates. 1.35 130 10 1.4 0.2 3.35 213 14 1.3 0.2 9.35 377 29 1.2 0.2 It was hypothesized that the magnitude of failure loads for these four groups of laminates would be controlled by the initial flaw size of the outer HA. Therefore, the non-indented group would fail at the greatest loads, while the laminates with the largest initial fl aws, the 9.35 kg group, would fail at the lowest loads. Mechanical data for the four different flaw size groups is found in Table 5.3. Statistical analysis (p = 0.05)

PAGE 138

119 showed as hypothesized, and that the mechanical behavior of the laminates is dependent on other factoer laminate testing and analysis, which will be discussed later in this chapte rel sho laminate with the 400-200-800 geometry the strength of the laminate is determined by the flaw size of the mi that there was no difference between failure loads, failure stresses, work of fracture, or toughness for any of the four initial flaw sizes. These results show that laminate strength was not controlled by the initial flaw size of the outer HA surface as w rs, that were explained through furth er. Thes sults wil w that for ddle HA layer. Figure 5.4 Lateral cracking seen during indentation using a 9.35 kg load (left), and the resulting chip-out that occurs during loading (right). Table 5.3 Mechanical property data versus indent loads Indent Load Failure Loads (N) Failure Stress (MPa) Work of Fracture (J/m2) Absorbed Energy at Failure (kJ/m3) Kapp (MPa-m1/2) NI 134 33 50 13 586 99 52 21 7.0 1.35 165 22 62 8 672 377 68 26 7.5 3.35 150 23 57 9 618 240 60 21 7.2 9.35 152 17 58 6 609 142 56 19 7.2 NI = Nonindented Laminates

PAGE 139

120 Loading Rate Effect Four groups of 400-200-800 laminates were tested at four different loading rates to test what effect the viscoelastic polymer had on laminate behavior. Failure loads were statistically equivalent (=0.05) for each of the four loading rates. The failure stress (Figure 5.5) and absorbed energy at failure (Figure 5.6) versus loading rate curves best demonstrate how the viscoelastic nature of PSu affects laminate behavior. Statistical analysis (=0.05) shows that there is no statistical difference between the failure loads for the four loading rates. There is a statistical difference between the toughness values for the different loading rates. However, there is no statistical difference between the failure loads of the 2 fastest loading rates of 2.5 and 25 mm/min. The rfailure strength of the laminates, but does affect the toughness. The toughness increase is e odulus l. lamiared to HA lamiLaminate geomet4. As with the previous two testing parameters, the hardness values were not statistically different esults show that the viscoelastic nature of the PSu does not directly influence the a function of the viscoelastic nature of the PSu. PSu is a viscoelastic material, and as such, behavior is time dependent. The behavior of the PSu depends of the relaxation timof the polymer. The relaxation time is dependent upon the viscosity and elastic mof the PSu and reflects how the polymer deals with applied loads at a molecular levePolysulfone Layer Thickness Effect on Laminate Behavior Since the viscoelastic nature of PSu was shown to influence the toughness of nates, it was hypothesized that the amount of viscoelastic material compwould have a significant effect on laminate behavior. This was the rationale for testing nates fabricated with different PSu layer thicknesses. ries and hardness values are shown in Table 5.

PAGE 140

121 Figure 5.5 A plot of failure stress versus loading rate. Figure-200-800 laminates. 5.6 A plot of absorbed energy at failure (toughness) versus loading rate for 400

PAGE 141

122 (=0.05) for the four laminate geometries, but were statistically lower than the hardnof the monolithic HA due to increased deflection resulting from the presence of the lmodulus PSu. The failure loads, Figure 5.7, increased with increasing polymer thickness, whicwould be expected since the thickness also increases as the PSu layer thickness was increased. However, the 50 m PSu laminates 1.7 mm thick failed at the same loads asthe monolithic HA specimens, which were 2.0 mm thick. This demonstrates an increasedload bearing capacity of a laminate over the monolithic material. Since stresses are proportional to P/t ess ower h 2, where P is the load and t is the thickness the thinner laminates fail at a greater ssystemsameTGeometry tress than the thicker monoliths, Figure 5.8. The 400-100-800 laminates with a thickness of 1.8 mm have even greater failure loads and stresses compared to the monoliths. These results were the first demonstration of the potential of the laminate as a thinner material could be used in place of the thicker monolith and have the load-bearing capacity. able 5.4 Hardness data for the laminates with different PSu layer thicknesses. Laminate n Total Thickness (mm) Hardness (GPa) 400-50-800 4 1.7 1.04 0.51 400-100-800 6 1.8 1.1 1 0.36 400-20 0-807 .30 0 2.0 1 0.16 400-400-800 6 2.4 1.24 0 .19

PAGE 142

123 Figure 5.7 A plot of failure loads versus PSu layer thickness. Monolithic HA failure load is plotted as a PSu layer thickness of 0. The stress values were also plotted versus the ratio of the outer HA layer thickness to the PSu layer thickness, Figure 5.9, for comparison to same plot generated using laminate theory, Figure 4.9. The laminate theory stress/N values were multiplied by an arbitrary 100N load to get the stress values shown on Figure 5.9. This adjustment was made in order to compare the trend curves on the same scale. The actual stress/N values are < 1 MPa and the predicted curve could not be seen on the scale of the experimental values. Comparison of the predicted curve versus the statistical trend for experimental data shows that the laminate theory can be used to predict failure stresses for laminates with decent accuracy.

PAGE 143

124 Figure 5.8 A plot of failure stress vs. PSu layer thickness. Monolithic HA is plotted as a Figu PSu thickness of 0. re 5.9 A plot of stress vs. outer HA/PSu layer thickness ratio for a comparison ofexperimental values to laminate theory predictions.

PAGE 144

125 Additionally, the work of fracture (Figure 5.10) and absorbed energy at failure (F5.11) increase as the PSu layer thickness increases. Kapp also increased as the PSu layer thickness is increased, Table 5.5. However, the absorbed energy at failure values are statistically (p=0.05) equivalent for the two thickest PSu layers. This is a result of a different failure mechanism for the 400-400-800 laminates, which will be described later in this chapter. Table 5.5 Apparent toughness values versus increasing PSu layer thicknesses Laminate Geometry Kapp (MPa-m1/2) igure Monolithic HA 1.0 400-50-800 3.7 400-100 -800 5.4 400-200-800 7.2 400-400-800 7.2 Figure 5.10 A plot of work of fracture vs. PSu layer thickness

PAGE 145

126 Figure 5.11 A plot of absorbed energy at fracture (toughness) vs. PSu layer thickness Outer HA Layer Thickness Effect on Laminate Behavior Testing of laminates with the same outer layer thickness indented with different indent loads showed no differe nce in the failure loads, failure stress, work of fracture, or toughness. The mechanical properties measured were independent of the crack size w size produced by the selected indent loadrefore, laminth outer HA layers < 400 m were proposed to study the effect when the outer flaw size produced by indentation is larger than the thickness of the outer HA layer. Laminates were produced with outer layef 400,0, and 100 100 m represents the thinnest outer HA yer that could be produced through hand polishing. The hardness values for the specimens built with thinner outer HA layers showed a significant effect the of the PSu layers. Laminate geometries and hardness values are shown in Table 5.6. The hardness values are not statistically different (=0.05) for the placed in the outer layer; however, the outer layer was thicker then the initial fla s. The ates wi r thicknesses o 20 m. la

PAGE 146

127 100 and 200 m outer HA layers, but are statistically different from the 400 m thickness. The impression showed visible plastic deformation around the edges of the indents indicating that contact stresses were penetrating to depths greater than the thickness of the outer HA layer. A discussion of laminate failure mechanisms later in this chapter will show that indentation of outer HA layer < 400 m causes the laminates to fail by a different mechanism than laminates with outer HA layers 400 m thick. Table 5.6 Hardness data for laminates with different outer HA layer thicknesses. Laminate Geometry n Total Thickness (mm) Diagonal Length (mm) Hardness (GPa) 100-200-800 6 1.4 459 135 0.33 0.16 200-200-800 6 1.6 397 179 0.57 0.36 400-200-800 7 2.0 213 14 1.32 0.17 Thinner outer HA layer had a significant effect on the mechanical behavior of the lam inates. The failure loads for these laminates are shown in Figure 5.12. The failure loads for the 100-200-800 lamouter layers and into the middle HA layer of the laminate. The net effect, as will be shown, is a small flaw within the middle HA layer being bridge by PSu, which affectively increases the fracture toughness of the area inate geometry, which have a total thickness of 1.4 mm, is a 200% improvement over monolithic HA specimens of a comparable 1.5 mm thickness, and a 60% improvement over monolithic HA specimens having a greater thickness of 2.0mm. Failure stresses for the outer HA layer specimens are shown in Figure 5.13. The stress value for the 100-200-800 laminates is the highest of all laminates tested in this study. The reason for this, as will be described in greater detain in this chapter, is that the indentation damage zone is larger than the outer HA layer thickness, and proceeds to make one large flaw that penetrates the two

PAGE 147

128 around the critical flaw, with the result being greater than failure loads and thus the failure sresulting PS Figure 5.12 A plot of failure loads versus outer HA layer thickness. The total thickness of the specimens is shown in parenthesis. greatest calculated stress values. The result is an almost 400% improvement over the tress of the monolithic HA Along with the high strength values, the thinner outer HA laminates have the greatest toughness values compared to other laminate geometries. While the work of fracture values, Figure 5.14, are not statistically different for the three outer HA thicknesses tested, the absorbed energy at failure (toughness) values increase with decreasing outer HA layer thickness, Figure 5.15. This is the result of the increased toughness around the crack tip produced by the indentation damage zone, and the u bridging of the initial flaw.

PAGE 148

129 Figure 5.13 A plot of failure stresses versusnolithic HA is plotted as a 2000 m thick HA layer. frome lowest m outer HA layer thickness. Mo Even though the thinner outer HA specimens had the greatest toughness calculated the area under the stress-strain curve, apparent fracture toughness values do not follow the same trend, see Table 5.7. The reason for the difference in trends is that the apparent fracture toughness is calculated from the work of fracture, and takes into account the composite elastic modulus, see Equation 2.15. Since the laminates with different outer HA layers had comparable work of fracture values, it follows that thodulus material would have the lowest apparent toughness value.

PAGE 149

130 Figure 5.14 A plot of work of fracture versus outer HA layer thickness. Figure 5.15 A plot of toughness versus outer HA thickness.

PAGE 150

131 Table 5.7 Apparent fracture toughness comparison for different outer HA layers. c app Laminate Geometry E(GPa) K (MPa-m1/2) Monolithic HA 52 1.0 100-200-800 37 6.4 200-200-800 40 6.3 400-200-800 42 7.2 Thick Middle Layer Effect on Laminate Behavior The results from the study of the thin outer HA layers showed that the greatest ick midould uThe principle is that if laminates with 100 m oute failed agh stress levels, thicker mcrease the load because stress is proportional to P/t2. FGeometry (mm) stress was obtained with thin layers. Thus, laminates with thin outer layers and th dle layers sh res lt in increased strength and increased load-bearing capacity. r HA layers t the same hi iddle layers w ould in our laminate geometries with thicker middle layers were built. These geometries, along with hardness values calculated for each group are shown in Table 5.8. Table 5.8 Hardness data for the four laminate groups prepared with thicker middle HA layers. Laminate n Diagonal Length Hardness (GPa) 200-2 00-1200 6 250 29 0.98 0.21 100-200-1400 6 292 127 0.89 0.60 200 -100-1400 6 388 72 0.43 0.17 100-100-1600 6 267 125 0.97 0.47 The mechanical property data for the thicker middle layer specimens show that no advantage was gained by increasing the thickness of the middle HA layer. The failureloads for the thicker middle HA layer laminate, Figure 5.16, were surprisingly lower than

PAGE 151

132 expected. The laminate theory model predicted that the 100-100-1600 group should fail have one of the highest failure loads. However, the experimental failure loads are equivalent to the failure loads of the non-indented monoliths, and to the 200-200-800laminates from the prior section, which are thinner and therefore should have smaller loads. The poor failure loads of the 100-100-1600 specimens are the first direct indica00 he a -100-e effect that laminate geometry has on the mechanical behavior of the laminate system, and shows that the designed placement of the PSu layer is a critical component that influence the behavior of the laminates. While the 100-100-1600 has the lowest stress/load value calculated using laminate theory at the outer tensile surface it has the smallest difference between this stress values at the outer HA layer and the stress values calculated at the tensile surface of the middle HA layer. For the calculated values, see Table 4.5. The difference is only 0.08 MPa between the two values. What this signifies is that only a small energy increase in the form of increased loading is necessary to reinitiate crack propagation through the middle layer. The 400-200-800 laminate from previous sections had higher failure loads of 150 N with the same total thickness. The difference between the outer and middle stress values for the 400-200-800 specimens is 0.22 MPa. This signifies that tion that the ultimate failure loads are not controlled by the outer HA layer. Failure stresses for the thicker middle layer laminates were dependent upon the thickness of the PSu layers. In the cases of the 100-100-1600, and the 200-100-14laminate groups there was no significant difference between the failure stresses of tlaminates versus the failure stress of non-indented monolithic HA. However, there issignificant increase in the failure stress of the 100-200-1400 laminates over the 2001400 laminates. This fact demonstrates th

PAGE 152

133 iddle Idealiz geomlami the miile propagation through the mminate geometries were modeled using the laminate theory, Figure 4.9. With these criteria in mind, the 350-400-500 geometry was selected. The difference in outer and middle layer stresses for this geometry was 0.32 MPa. Figure 5.18 is a plot of the failure loads gained for the 350-400-500 laminates versus other laminates having a total thickness of 2.0 mm. While there was not a dramatic increase in failure loads as expected, the 350-400-500 group statistically (p=0.05) yields the greatest failure loads of any of the laminates fabricated to a thickness of 2.0 mm. The 171 N average is a 175% increase in the load-bearing capacity for these laminates over indented monoliths, and a 110% increase over the non-indented monoliths. This is another demonstration of the predictive abilities of the laminate theory as the 350-400-500 laminate geometry was selected solely on laminate theory calculations. it will require a greater increase in load to reinitiate crack propagation through the mHA layer for the 400-200-800 laminates then for the 100-100-1600 laminates. ed Laminate Having already shown the greatest stress values for the laminates can be obtained by reducing the thickness of outer HA layer, the goal became to design the laminate etry that would produce the greatest failure load. The total thickness for the nate would be held constant at 2.0 mm. However, an improved load bearing capacity through a change in laminate geometry was desired. The principle developed fromddle layer testing showed that the greater the difference between the predicted tensstress of outer and middle HA layers the greater the load has to be to reinitiate crack iddle HA layer. A number of theoretical la

PAGE 153

134 Figure 5.16 Failure loads for laminates with thicker middle HA layers. Figure 5.17 Failure stresses for laminates with thicker middle layers

PAGE 154

135 Figure 5.18 Experimental failure loads for all the laminate geometries having a total mcommon failurending. The laminate thickness of 2.0 mm Determination of Failure Mechanisms Through mechanical testing the laminates were shown to have superior mechanical properties. It now becomes important to explain why the properties are achieved and through what mechanisms. The goal is to use the laminate theory to improve design and predict actual failure loads. This requires a better understanding of the failure echanisms. Laminate Failure Modes Loading in biaxial flexure produced three laminate failure modes. The most e mode was complete separation of all five layers due to the b

PAGE 155

136 result was that laminates fractured into two or three pieces similar to monolithic fa ilures shown in Figure 2.25. Figure 5.19 shows the fracture surface of a 400-200-800 laminate. The second failure mode is the same as the first except specimens remain in one piece fo in biaxial flex Optical microschowed that fracture through both outer and the middle HA layers occurred, but at least one of the PSu layers is still intact for these specimens, Figure 5.20. Loading was automatically stopped by the Figure 5.19 An optical micrograph of 400-200-800 laminate fracture surface. The f the inl inrow), (black arrow), and loading piston contact are all indicated. testie theplied loow 10 N or machine reached the cem alloweding fixture. If loading could be continued, then eventually the PSu layers would have undergone continued deformation until final failure. Intact samples are a result of experimental limitations that arise from llowing failure ure. opy s locations o itia dent flaw (white ar middle layer flaw ng machine onc ap ads fell bel if the testing maximum displa ent d by the loa

PAGE 156

137 Figu in en of or thickne with gh -PSu layer. The loading piston actually penetrated through all five layers of the laminate. re 5.20 Optical micrographs of a laminate still intact after being loaded to failurebiaxial flexure. Left: Failure occurred from the initial indent (indicated by thearrows) in the outer HA surface, Right: Examination of the edge of the specimshows failure of the outer HA layer, delamination along the interface, and failurethe middle HA layer. loading ductile materials in a bending test, as there is a finite displacement range that can be tested. Most polymers are loaded in pure tension to allow for large displacements fthis reason. The third failure mode was penetration of the loading piston through the entire ss of the specimens. This failure mode occurred primarily in the laminatesthe thickest PSu layers. Analysis of the fracture surfaces showed that the punch throuwas a secondary failure as the laminates failed first through bending in the HA layer before the PSu layers failed; at this point failure loads became sufficient to drive the loading piston through the middle of the laminates. The fractured laminates have a large hole through the center, but are otherwise intact. Figure 5.21 is a picture of a 400-400800 laminate that failed via the punch through mode. The indented outer HA layer has been shattered and is completely removed from the image, revealing the partially failed

PAGE 157

138 Figure 5.21 SEM image of punch through failure of a 400-400-800 laminate. The the PSu layer were the loading piston breached all five layers is indicated by the white arrow. he ers lated Fracture analyses are divided into two categories: analysis of the 400-200-800 laminates and analysis of all other laminate geometries. The reason for this division is remnants of the indented outer HA layer are indicted by the black arrows, failure of Fractography of Laminates Optical microscopy and SEM were used to analyze the fracture surfaces of tlaminates. Fractography was used to find fracture origins of the indent side HA layand of the middle HA layer. These fracture origins were used to calculate the failure stress of each layer through Equation 2.12. The fracture toughness of each HA layer was assumed a constant value of 0.6 MPa-m 1/2 This was the fracture toughness calcufrom testing of the monolithic HA specimens (see Table 2.4).

PAGE 158

139 that the 400-200-800 laminates have the only geometry with an outer HA layergreater than th thickness e flaw size produced by the 3.35 kg load. The average flaw size for monolithic HA specimens indented with the 3.35 kg load was 313 m. Analysis of those of HA flaws aret a 3.35 kg load produced larger indents in the laminates than in the monoliths, Table 5.9, most likely due to the presence of the less stiff polymer layers. Table 5.9 Comparison of flaw sizes for monolithic HA and the outer HA layer for 400-200-800 laminates. Non-Indented Flaw Size (m) Flaw Size From 3.35 kg Indent (m) fracture surfaces revealed that laminates with an outer HA layer thickness < 313 mshowed flaws from indentation that bridged the PSu layer, and therefore will be analyzed differently than the 400-200-800 m laminates. Fracture Analysis of 400-200-800 Laminates Failure of the Outer HA Layer Fractography performed on all laminates showed that laminates with indented outer HA layer thicknesses of 400 m contained slightly larger initial flaw sizes than monolithic HA with the same indent loads. Typical flaws within the outerlayer produced through indentation are shown in Figures 5.22-5.23. These larger initial consistent with the results of hardness measurements, which showed tha Monolithic HA 98 43 313 39 Outer HA Layer of Laminates 125 37 374 75

PAGE 159

140 Figure 5.22 Optical micrograph of the initial flaw size (black arrows) produced by a 3.35 kg indent (white arrow) in a 400-50-800 laminate. Figure 5.23 SEM image of the initial flaw produced in a 400-100-800 laminate produced by a 3.35 kg indent load.

PAGE 160

141 Outer layer flaw sizes for the non-indented laminates were not statistically different (=0.05) from the flaw sizes of the non-indented monoliths. With equivalent flaw sizes at the outer tensile surfaces, the failure stress and failure loads should be the same for the monolithic HA specimens and the outer HA layer of the laminates if the outer layer controls failure. Since the ultimate failure loads of the laminates have been shown to result from failure of the middle HA layer, the load at which the outer HA layer fractures had to be estimated from the load displacement curves, Figure 5.24. In some cases there was a measurable drop in loading that occurred when the outer HA layer failed making measurement of the failure load relatively straightforward. However, in most cases the failure load had to be estimated from the slope change of the load is dTo show that the outer HA layerayer, failure loads and tresses were compared for nonindented laminates, and for the 400-200-800 laminates re 5.25. From Equation 2.13, when fracture toughness is constant, an increased flaw size e fracture process begins with the failure of the outer HA layer. compared for laminates with different PSu layer thicknesses. As seen placement curve. failed first prior to the middle l s indented with 1.35 and 3.35 kg loads, Figu results in a decreased failure stress. The ultimate failure of the 400-200-800 laminates has been shown to be independent of the initial flaw size, Table 5.3, and therefore the laminates all failed at the same stress. However, the fact that the failure stress of the outer HA layer is dependent upon initial flaw size indicates that th The outer HA layer failure stresses were also in Figure 5.26, outer HA failure stresses are statistically (=0.05) independent of the thickness of the PSu for the laminates with 400 m outer

PAGE 161

142 layersfailure loads are calculated either from the initial drop (top) or by extrapolating the This result demonstrates that the initial failure of the laminates is dependent on the flaw size of the outer HA layer, and is not influenced by the PSu layer. Figure 5.24 Load-displacement curves for two non-indented laminates. Outer HA layer initial slope (dashed line) and estimating when the slope change begins (bottom).

PAGE 162

143 Figure 5.25 Failure stresses of the outer HA layer calculated for three different initial flaw sizes. Figure 5.26 Failure stresses of the outer HA a layer for laminates built with different PSu layer thicknesses.

PAGE 163

144 Failure of the Middle HA Layer Fracture origins were found and measured for the middle HA layers in order to calculate the stress at which the middle HA layers failed. Figures 5.27 5.29 are the fracture origins that were found within the middle HA layers. Failure iddle HA layers were calculated using Equation 2.12, and assumghness of the middle HA layer to be the same as for the monolith of the middle layer was also calculated by multiplying the ex by the stress/1 N load calculated by laminate theory for the tensile surface of ddle HA layer. For the 400-200-800 laminates, the middle HA layer has a stress/1N load of 0.16 MPa. The comparison of these stresses shown in Table 5.10, onstrates that there is good agreement between the two methods of calculation. The two stress values are consistent with each other, and inditheory can be used to calculate the failure stress of the middle layer for a lam examples of stresses for the ming the fracture touic HA. The failure stressperimental failure loadsthe midem cate that the laminate inate structstress of the materials is relatively constan average failure stress by the the predicted failure load for the specimen. Geometry FS, Calculation FS, Calculation ure. The stress/1 N load could also be used to estimate failure loads if the failure nt. Dividing a stress/1N calculation produced Table 5.10 Middle HA layer failure stress calculations for the 400-200-800 laminates. Laminate n Experimental Failure Load (N) Middle Layer Flaw Size (m) Flaw Size (MPa) LT (MPa) 400-200-800 13 155 22 254 41 30 2 28 6 n = number of tested speciemens LT = laminate theroy FS = Failure Stress

PAGE 164

145 Figure 5.27 Optical micrograph of the fracture oringin of a 400-200-800 laminate. T he center of the flaw is indicated by the white arrow. Figure 5.28 Optical micrograph of a minate. The specimen has been sputter coated to make the surface marking stand out. iddle layer fracture origin a 400-100-800 lam 400 m 400 m Outer HA Layer 500 m PSu Layer Outer HA and PSu Layers Middle HA Layer 500 m 500 m Outer HA Layer PSu Layer Outer HA and PSu Middle HA Layer Layers

PAGE 165

146 Figure 5.29 SEM image of the middle HA layer fracture origin. Failure of the Outer HA Layer, Loading Piston Side Fracture surface analysis also revealed contact damage resulting from the loading piston. The fractures markings are consistent with contact damage from blunt indenters that produce Hertzian cone cracks [38]. Cone cracks form due to localized tensile stresses located at the edge of the indenter and propagate at an angle of 22 from the surface. Damage from the loading piston was seen in the form of surface damage prior to tensile failure of the laminate in biaxial flexure, Figure 5.30, or as secondary fractures as the piston penetrates the laminate layers once tensile failure has occurred, Figure 5.31, although not all specimens showed damage from the loading piston. Determination of when piston damage had occurred was done by matching fracture markings on either side ith each other, as in Figure 5.31, then the loading piston crack occurred after primary failure. However, if fracture markings on either side of the of cracks, which were produced by the loading piston. If fracture markings on either side of the crack are consistent w

PAGE 166

147 p iston cracks differ in their patterns, as in Figure 5.30, then the piston damage occurred prior tos yers, tures seen are a result of failure mechanisms not crystallization ; f the bending failure. Polysulfone Layer Failure Examination of the polymer portion of the fracture surface of the 400-200-800 laminateshowed that the PSu layers failed in an unusual manner. Figures 5.32 and 5.33 are images of the PSu layer fracture at two different crosshead displacement rates. At first inspection, the images appear to show spherulitic structures that signify a crystalline polymer. However, PSu is known to be an amorphous polymer with little long range order. Comparison of the PSu layers fractured at different crosshead displacements showed the greater the displacement rate the smaller the features within the PSu laindicating that the fea during the drying process. Oblique lighting of the PSu layers causes the features to become invisible to the optical microscope when compared to the overhead lighting, Figure 5.34. PSu layers appear to show crazing because of the change in lighting anglehowever, the features are much larger than would be expected for typical crazing, andcrazing is usually associated with a coalescence of the crazes into one united crack front from which catastrophic failure occurs. The surface features seen in the PSu layer olaminates occur throughout the entire length of the fracture surfaces, and never coalesce into one crack front. Closer examination of the features using SEM shows that the features are failing in both brittle, Figure 5.35, and ductile manners Figure 5.36. In some cases, two PS features located side-by-side to one another, Figure 5.37, fail in two different manners.

PAGE 167

148 Figure 5.30 SEM image of a 400-100-800 laminate showing contact damaarrows) from the loading piston which occurred prior to propagation of the primfracture (black arrows) which was reinitiated near the HA/PSu interface (dashed arrow). Figure 5.31 Optical micrograph of secondary cracks (arrows) which occurred through ge (white ary punch failure after primary failure from bending.

PAGE 168

149 Figure 5.32 Optical micrograph of a PSu layer of a 400-200-800 laminate fractured at a crosshead displacement rate of 0.25 mm/min. The specimen was sputter coated prior to analysis. Figure 5.33 Optical micrograph of a PSu layer of a 400-200-800 laminate fractured at a crosshead displacement rate of 25 mm/min. 200 m 200 m 200 m 200 m 200 m 200 m

PAGE 169

150 FiguFigure 5.35 SEM images of brittle failure of the PSu layer, the pore (indicated by the n the left im a twarkin in the rie) terise fractu re 5.34 Optical micrographs of the PSu layer features through overhead (top) and oblique lighting (bottom). arrow i age) showscharac ist hackle mtic of brittl g (arrowre. ght imag

PAGE 170

151 Figure 5.36 SEM imr to failure at the Figure 5.37 SEM imdifferent me left feature age of ductile deformation of the PSu layer priocenter of the feature (black arrow). age of two side-by-side features within the PSu layer that failed in anners, a twist hackle marking (solid arrow) indicates thfailed in a brittle manner, while the pullout deformation (dashed arrow) of the right feature indicates ductile failure.

PAGE 171

152 Analysis of the PSu fracture surface feature leads to no definitive explanation as to the origin of the featuresost stressing rae mthe amevaporation che sensitivity of the evaporation process results in a range of evaporation rates at the ing polymer layer will have Microscopic analysis demonstrated that these features are mlikely not spherulites, or crazes. However, the size dependence of the features on te leads to the possibility that the features are the result of phase segregation that occurs during drying of the PSu films As the TCE solvent begins to evaporate during drying of the PSu layer, the volumof the polymer/solvent mixture decreases. The evaporation process is constant on the acroscopic level, and controlled by a number of factors, which include the conditions bient environment, and viscosity of the polymer/solvent system. The rate of an be significantly affected by altering any of these conditions. T microscopic level. Therefore, different regions of the dry different concentrations of solvents, and therefore, different regions will dry at different rates. While there is a range of solvent concentrations throughout the regions of thedrying PSu layer, there is also a range in the size of PSu chains within each of these regions. Since the solubility parameter of the PSu does not change, a reduction in solvent due to evaporation causes the PSu chains to precipitate out of solution. The solubility parameter is dependent upon molecular weight [54], and thus the largest PSu chains precipitate out of solution first, followed by the smaller chains. However, due tochain interactions, and diffusion, when the larger chains precipitate out of solution, some of the smaller chains become entangled with the larger chains and are forced to precipitate out of solution. The net result of this drying process are phase separated

PAGE 172

153 regions, the size of which is dependent on solvent concentration and the molecular weighdistributions within speci t fic each region of the drying PSu layer. loaded in biaxial flexure, the failure of the PSu layers s s are ement rate of 0.25 mm/min. A ated ler the porosity of the HA discs, forming the mechanical bonding that holds thand When the laminates are becomes dependent upon at what loading rate that the specimens are tested. At a slower loading rate, the PSu chains have greater time for relaxation processes to occur. Thiallows for activation of the largest polymer chains within the failure process, and the larger chains will then control failure of the PSu regions. These larger failure regionseen along the fracture surface shown in Figure 5.32, which is for the slower crossheaddisplac t increased loading rates, there is insufficient time for chain relaxation prior toultimate failure. The faster loading rate means activation of only the smaller polymer chains within each phase segregated region since there is only sufficient time for the relaxation and orientation of the shortest PSu chains prior to chain scission occurringThe result is that failure is controlled by the small PSu chains within the phase segregregions. The fracture surface of laminates tested at a crosshead displacement rate of 25 mm/min, Figure 5.33, shows an increased number of regions that are significantly smalthan the regions seen at the slower loading rate. Interfacial Failure and Toughening Mechanisms As was described in Chapter 3 on laminate fabrication, prior to drying, the casting solution infiltrates into e layers together. This penetration can be seen macroscopically as a darkening color change in HA prior to drying. After drying the PSu layer becomes translucent could not be seen due to light reflection off the white HA layers. However, the

PAGE 173

154 sulfonated-PSorange hue to the filmcroscopy, Figure 5.38. Optical mm. Figure 5.38 Optical m easured through optical mifound at different ore than half the thicknessd on both surfaces of the middle HA lathrough the HA. Figures 5.41-5.42 shows interfacial bridging that occurs in between the HA and PSu layers. u, which was used briefly in this study as described in Chapter 3, has an s when dry and can been seen though optical miicroscopy showed that sPSu penetrated to depths between 50 and 100 icrograph of a sPSu/HA laminate showing the depth of penetration of the casting solution during solvent casting. SEM revealed the depth of penetration was much deeper than mcroscopy. Figures 5.39-5.42 are SEM micrographs of PSu penetration depths into the HA. Figure 5.39 shows PSu penetration across m of the middle HA layer. Since casting is performeyer, there is a strong probability that PSu penetrates the entire thickness of e middle HA layers. Figure 5.40 show that the PSu bridges cracks that propagate th

PAGE 174

155 The ability of the PSu to bridge cracks is only one of the mechanisms that results Delami e formrein strength and toughness of the 400-200-800 laminates compared to the monolithic HA. Failure Mechanism of Laminates with Outer HA Layers < 400 m different mproduced fromdeformFpropagation occurs during loading in biaxial flexure. easurements were made on only a small percentage of the tested specimage from indentation and subsequent loading left little of the outer HA layers intact for flaw size measurements. Average flaw sizes and the failure stresses in the increased toughness of the laminates compared with the monolithic HA. nation causes crack arrest and leads to the required energy increase necessary toreinitiate cracks through the middle HA layer. Evidence of delamination can be seen through pullout of the PSu layers, Figure 5.43, and through PSu failure origins that occur as the crack travels along the HA/PSu interface. Delamination can also be seen in th of crack deflection, Figure 5.20, as cracks propagate down the interface and then itiate into the middle HA layer. Failures of the PSu layers as cracks propagate along the interface are shown in Figures 5.44-5.45. The combination of these toughening mechanisms leads to the increased failure Laminates fabricated with outer HA layers thinner than 400 m failed via a echanism than the 400-200-800 laminates due to the size of the cracks which are produced through indentation of HA with a 3.35 kg load. The average flaw size a 3.35 kg Vickers indent was 313 m as measured for fracture toughness calculation of the monolithic HA specimens. For the thinner outer HA layer contact stresses from the 3.35 kg indent were sufficient to cause crushing of the outer HA layer, ation of the PSu layer beneath, and introduce cracks into the middle HA layer, see igure 5.46. The resulting damage zone acts as the initial flaw from which crack Flaw size m ens as the dam

PAGE 175

156 yer. Figure 5.39 SEM images showing the penetration depth of PSu into the middle HA la The fracture surface is shown top left, the box indicates the area where the top right picture was taken, the bottom picture showing PSu pullout was taken from the areaindicated by the box in the top right image.

PAGE 176

157 Figure 5.40 SEM images of PSu bridging of small cracks in the HA layers. These images were taken of the secondary interface crack shown in Figure 5.31. Figure 5.41 SEM im ages showing PSu bridging of the HA/PSu interface. PSu fibrils are indicated by the arrow.

PAGE 177

158 Figure 5.42 SEM image of PSu bridging of the HA/PSu interface Figure 5.43 Optical micrographs showing the pullout of the PSu layers during fracture of the laminates.

PAGE 178

159 Figure 5.44 SEM images of brittle PSu layer originating from the HA/PSu interface. Tharrow indicates the location of the origin. e e occurred within the PSu layer and nterface. Figure 5.45 SEM images of a PSu layer fracture origin. The location of the origin is indicated by the arrows on both images, failurnot from the i

PAGE 179

160 associated with these flaws are shown in Table 5.11. The predicted failure stress from laminate theory is also shown in Table 5.11. The failure stress for these laminates was calculated by assuming that the flaws simila r to the one shown in Figure 5.34 were only HA, just as with the middle layer last section. The predicted fantal failure loads by the stresseometries. easured stress for the laminates. The lack of agreement is th the thinner HA layer lamitobridgeenon similar to this bridges the critical flaw effectively increasing the toughness of the material around the flaw. Fs and g the greatest ratio of PSu layer thickness to outer HA thickness. A 2:1 ratio for these laminates means a greater amount of PSu in the damaged area failure calculation for the 400-200-800 laminates discussed in the ilure stresses were calculated by multiplying the experime/N values given by the laminate theory for each of the laminate gLesser agreement exists between the predicted and mnates with outer HA layers < 400 m than for the 400-200-800 lamie result of the PSu that is present within the flaws ofnates. As was shown for the 400-200-800 laminates, PSu has the ability cracks and interfaces within the laminate system. A phenom is occurring within the initial flaws of the thinner outer HA laminates. The PSu layer or the same flaw size, an increased toughness results in a greater failure stress according to the fracture mechanics (See Equation 2.12). This toughening mechanism isimilar to crack-tip shielding [50] which has been shown to increase the toughness though the deposition of thin films onto indented surfaces prior mechanical testing. Further evidence of the toughening around the initial flaw is the increased discrepancy between the compared failure stresses in Table 5.11 for the 100-200-800the 100-200-1400 laminates. The greater discrepancy for these two laminate groups arises from their havin

PAGE 180

161 compar effect 00-) n Failure Load ed with the two other laminate groups in Table 5.11 that have a 1:1 ratio of PSu toouter HA thickness. An increased amount of PSu results is a greater tougheningfrom PSu bridging, and correspondingly greater than expected failure loads. Figure 5.46 Optical micrograph of the fracture origin (solid white arrows) of a 100-11600 laminate, the large damaged area is caused by the indent (dashed arrow Table 5.11 Failure stress calculations for laminates with outer HA layers < 400 m Laminate Geometry Experimental (N) Flaw Size (m) Middle Failure Stress, FS (MPa) Middle Failure Stress, LT (MPa) 100-200-800 1 77 314 27 44 200-200-800 3 112 18 178 30 36 3 39 6 100-200-1400 2 131 28 190 4 35 0 47 11 100-100-1600 6 107 19 289 81 29 4 35 6 FS = calculation from a measurement of flaw size; LT = Laminate Theory

PAGE 181

CHAPTER 6 CONCLUSIONS Primary Conclusion The primary goal of this project was to demonstrate that two biocompatible terials, a brittle hydroxyapatite (HA) and a ductile polysulfone (PSu), could be bined into a composite structure with strength and toughness values similar to bone. In order to achieve this goal a novel combination of processing techniques was developed to produce HA/PSu laminate structures. Laminates were shown to have strength values that are 400% greater, and apparent fracture toughness values 1100% greater, than monolithic HA. In Chapter 1, a graph developed by Suchanek and Yoshimura[6] showed the available biomaterials for bone replacement, Figure 1.1. The apparent fracture toughness and composite moduli for each laminate group designed and tested are showin comparison to bone and other HA composites in Figure 6.1. The data points plotted for the HA/PSu laminates represent the average of the upper and lower bounds of the rule ixtures. The box represents the range of values covered by the upper and lower bounds of the rule of mixtures. The strength, toughness, and modulus of the laminates fabricated here are closer to the properties of bone than any known composites. macomn of mgreateof the HA/PSu laminates aramic composites. owever, the composite moduli are much more comparable to bone than the all ceramic omposites. The net result is an improvement over both types of existing composite The apparent fracture toughness and elastic modulus of the HA/PSu laminates are r than the values for existing HA/polymer composites. Apparent fracture toughness re within the same range as the HA-based ce H c 162

PAGE 182

163 Figure 6.1 Addition of the HA/PSu laminates designed in this study to the graph of available biomaterials for bone replacement developed by Suchanek and Yoshimura [6]. It has been shown that the toughnesickness. The resulting HA/PSu laminates would have strength and toughness values overing the entire range of bone. Secondary Conclusions A number of beneficial results were gained through the course of achieving the biomaterials. However, the HA/PSu laminate geometries tested in this study still do not cover the entire range reported for bone. Thus, further refinement of the laminate structure is necessary for the HA/PSu laminates to be fully compliant with the full range of elastic modulus and fracture toughness of bone. Future work should focus on fabricating laminates with even thicker PSu layers than the 400 m tested in this study. s of the laminates increases with increased PSu layer th c

PAGE 183

164 primary goal of this project. Porous hydroxyapatite specimens were fabricated through a tape casting process that iconstituents insteg presence of secondary calciumd into hydroxyapatite with an XRD spectrumhydroxyapatite except for one smand had strength and fracture through me Monolithic HA and PSu were formed into laminate composites through one of three lamination methods developed specifically for this study. Each of the three methods (matching halves, bottom up, and pre-fabricated PSu layer) involved solvent 70C, and created a more time efficient less. Surface finish of the HA discs prior to casting was shown to reduce the strength of the laminates, Figure 3.22. Since bonding between the PSu and HA occurs through a mechanical interlocking, finer polishing of the HA discs prior to lamination reduces the amount of surface area available for the interlocking to take place. The result was delamination along the interface to the nvolved formation of slurries through a one-step addition of ad of a more traditional multi-step approach. The startinhydroxylapatite starting powder having an XRD spectrum, Figure 2.7, showing the phosphate phases was transforme that is identical to naturally occurring all CaO peak, Figure 2.13. The sintered hydroxyapatite contained 30% porosity toughness values consistent with hydroxyapatite prepared thods that are more complex. casting, stacking, and drying with all three methods successfully producing laminates. The benefits and limitations associated with each method were examined, and it was determined that for PSu layers 200 m the bottom up method works best, and for PSu layers > 200, the pre-fabricated PSu layer method works best. It was also demonstrated that the strength and toughness of the PSu layers are independent of retained solvent content and drying temperature. This fact allowed for drying of solvent cast PSu films at amination proc

PAGE 184

165 extent that the mechanical stability of the laminates was compromised, and failure occurs at a lower load. A mathematical model was developed to describe the stress distributions that results when applying load to the HA/PSu laminates. This laminate theory model was derived by combining existing laminate beam theory using bending moments for simply supported circular plates. The validity of the theory was confirmed though a comparison with finite element analysis of the HA/PSu laminates. A comparison, Table 4.3, showed agreement of the two models to within 3%. This agreement validated the use of the laminate theory model to calculate stress distributions from experimental loads. The laminate theory model was used to predict which laminate geometry with a total laminaThe failure mechanisms of the laminates were characterized to gain a better understanding of the toughening mechanism. Initial testing of 400-200-800 laminates showed a drastic improvement of strength and toughness over monolithic HA, Figures 5.1 5.2. In order to identify the design parameters that most influenced laminate behavior, laminates were tested with flaw sizes both smaller and larger than the thickness of the outer HA layer. When the initial flaw size is smaller than the outer HA layer, thickness of 2.0 mm would produce the greatest failure loads in biaxial flexure tests, Figure 4.9. Design optimization led to the fabrication and testing of laminates with a 350-400-500 geometry. A comparison of the failure loads of the 350-400-500 laminate geometry versus other laminates having the same 2.0 mm total thickness, Figure 5.18, showed that the 350-400-500 geometry did indeed have the greatest experimental failure loads. Thus, the lam inate theory can be used for predicting failure strengths for different te geometries.

PAGE 185

166 crack propagation was shown to begin at the tensile surface of the outer HA layer, crackpropagation is arrested at the HA/PSu interfac s e, and then reinitiated into the middle HA layer.lent the lamates ee for t of this toue ness, The critical flaw size of the middle HA layer is what ultimately determines the failure strength of the laminate. Middle HA layers have a fracture toughness equivato that of monolithic HA, and therefore they fail at approximately the same stress assuming similar flaw sizes. However, the presence of the PSu layers shields the tensile surface of the middle HA layer for direct loading. Thus, greater failure loads are seen for inven though the middle HA layer and the monolithic HA specimens fail at approximately the same stress. Flaw sizes and middle layer failure stresses were determined using fractography, and compared to the failure stress of the middle HA layer predicted by the laminate theory solution, Table 5.10. When the initial indentation flaws are larger than the thickness of the outer HA layer the result is a bridging of the critical flaw by the PSu layer and a greater fracture toughness of the area around the critical flaw. Instead of crack propagation starting at thouter HA layer, arresting at the interface, and reinitiating into the middle HA, the entiredamage area formed through indentation acts as the initial flaw. Failure of the entire laminate occurs once the stress reaches a magnitude sufficient to propagate a crack the initial flaw despite the additional toughness gained from PSu bridging. The resul ghening mechanism was the greatest failure stresses calculated for any of thlaminates tested in this study, Figure 5.13. The real benefit of this mechanism was shown through a comparison of failure loads for laminates with thin outer HA layers to monolithic HA, Figure 5.12. The 100-200-800 laminate, which has a total thickness of 1.4 mm, failed at loads that were 200% greater than monoliths of a comparable thick

PAGE 186

167 and 60% greater than monoliths that were 2.0 mm thick. These results show that a thinner laminate has greater load bearing ability than a thicker monolithic specimen. Thus, less material could be used to fulfill the same load-bearing requirements

PAGE 187

LIST OF REFERENCES 1. Hench L., Bioceramics: From Concept to Clinic. Journal of the American Ceramic Society 74 [7]: 1487-1510 (1991) 2. Hench L., Bioceramics. Journal of the American Ceramic Society, 81(7): 1705-1728 (1998) 3. Hench L., Wilson J., An Introduction to Bioceramics. Advanced Series in Ceramics. Vol. 1. 1993, Singapore: World Scientific. 4. Wang M., Porter D., Bonfield W., Processing, characterization, and evaluation of hydroxyapatite reinforced polyethylene composites. British Ceramic Transactions, 93: 91-95 (1994) 5. Rho J., Kuhn-Spearing L., Zioupos P., Mechanical properties and hierarchical structure of bone. Medical Engineering and Physics, 20: 92-102 (1998) 6. Suchanek W., Yashimura M., Processing and properties of hydroxyapatite-based biomaterials for use as hard tissue replacement implants. Journal of Materials Research, 13 [1]: 94-117 (1998) 7. Doremus R.H., Review: Bioceramics. Journal of Materials Science, 27: 285-297 (1992) 8. Ratner B.D, Hoffman A.S., Schoen F.J., Lemons J.E., Biomaterials Science: An introduction to Materials in Medicine. 2nd ed. 2004, San Diego: Elsevier Academic Press. 9. Wang M., Developing bioactive composite materials for tissue replacement. Biomaterials, 24: 2133-2151 (2003) 10. Puajindanetr S., Best S.M., Bonfield W., Characterization and sintering of precipitated hydroxyapatite. British Ceramic Transactions, 93 [3]: 96-99 (1993) 11. Verbeeck R.M.H., deMaeyer E.A.P., Driessens F.C.M., Stoichiometry of potassium-containing and carbonate-containing apatites synthesized by solid-state reactions. Inorganic Chemistry, 34 [8]: 2084 (1995) 12. Carotenuto G., Spangnuolo G., Ambrosio L., Nicolais L., Macroporous hydroxyapatite as alloplastic materials for dental applications. Journal of Materials Science: Materials in Medicine, 10: 671-676 (1999) 170

PAGE 188

13. Ruys A.J., Wei M., Sorrell C.C., Dickson M.R., Brandwood A., Milthorpe B.K., Sintering effects on the strength of hydroxyapatite. Biomaterials, 16: 409-415 (1995) 14. Knowles J.C., Talal S., Santoffects in a glass reinforced hydroxyapatite. Biomaterials, 17: 1437-1442 (1996) ashimura M., Kakihana M., Yoshimura M., Processing and iskers. e ials, rsity wth 24. s e J.D., Sintering 15. Suchanek W., Y mechanical properties of hydroxyapatite reinforced with hydroxyapatite whBiomaterials, 17: 1715-1723 (1996) 16. Tanner K., Downes R., Bonfield W., Clinical applications of hydroxyapatitreinforced materials. British Ceramic Transactions: 93 [3]: 104-107 (1994) 17. Wang M., Bonfield W., Chemically coupled hydroxyapatite-polyethylene composite: structure and properties. Biomaterials, 22: 1311-1320 (2001) 18. Wang M., Joseph R., Bonfield W., Hydroxyapatite-polyethylene composites for bone substitution: effects of ceramic particle size and morphology. Biomater19: 2357-2366 (1998) 19. Kamat S.K., Ballarini R., Heuer A., Structural basis for the fracture toughness of the shell of the conch Strombus gigas. Nature, 405: 1036-1040 (2000) 20. Vincent J., Structural Biomaterials. Revised ed. 1990, Princeton, NJ: Princeton University Press. 21. Zhou B.L., Some progress in the biomimetic study of composite materials. Materials Chemistry and Physics, 45: 114-119 (1996) 22. Clupper D.C., Tape Cast Bioactive Metal-Ceramic Laminates For Structural Application, in Materials Science and Engineering. Ph.D. Dissertation, Univeof Florida (1999) 23. Hwu K.I., Derby B., Fracture of Metal/Ceramic Laminates-II. Crack GroResistance and Toughness. Acta Materialia, 47 [2]: 545-563 (1999) Chen Z. Mecholsky Jr. J.J., Toughening by Metallic Lamina in Nickel/Alumina Composites. Journal of the American Ceramic Society, 76 [5]: 1258-1264 (1993) 25. Chen Z., Mecholsky Jr. J.J., Control of strength and toughness of ceramic/metal laminates using interface design. Journal of Materials Research, 8 [9]: 2362-2369 (1993) 171

PAGE 189

26. Yun N.G., Won Y.G., Kim S.C., Toughening of carbon fiber/epoxy composite by inserting polysulfone film to form morphology spectrum. Polymer, 45: 6953-6958 (2004) 27. Marcolongo M., Ducheyne P., Garino J., Schepers E., Bioactive glass fiber/polymeric composites bond to bone tissue. Journal of Biomedical Materials Research, 39: 161-170 (1998) 28. Wijmans J.G., Baaij J.P.B., Smolders C.A., The mechanism of formation of microporous or skinned membranes produced by immersion precipitation. Journal of Membrane Science, 14: 263-274 (1983) 29. Arita I.H., Wilkinson D.S., Mondragon M.A., Castano V.M., Chemistry and sintering behavior of thin hydroxyapatite ceramics with controlled porosity. Biomaterials, 16: 403-408 (1995) 30. Hoepfner T.P., Case E.D., The influence of the microstructure on the hardness of sintered hydroxyapatite. Ceramics International, 29: 699-706 (2003) 31. Ramay H.R., Zhang M., Preparation of porous hydroxyapatite scaffolds by combination of the gel-casting and polymer sponge methods. Biomaterials, 24: 3293-3302 (2003) agh M., Graham E., Pantano C., Elastic Moduli of Silica Gels Prepared w American Ceram 32. Murtith Tetraethoxysilane. Journal of theic Society, 69 [11]: 775-779 (1986) d. 2000, a-es Science and Technology, 63: 1433-1440 (2003) of Brittle Solids. 2nd ed. Cambridge Solid State Science bridge: Cambridge University Press. 33. ASTM C1327, Standard Test Method for Vickers Indentation Hardness of Advanced Ceramics. ASTM C1327-99: 480-487 (1999) 34. Wachtman J.B., Capps W., Mandel J., Biaxial Flexure Tests of Ceramic Substrates. Journal of Materials, 7 [2]: 188-194 (1972) 35. Callister, W.D., Materials Science and Engineering An Introduction. 5th eNew York: John Wiley & Sons, Inc. 36. Karger-Kocsis J., Barany T., Moskala E.J., Plane stress fracture toughness of physically aged plasticized PETG as assessed by the essential work of fracture (EWF) method. Polymer, 44: 5691-5699 (2003) 37. Aksel C., Warren P.D., Work of fracture and fracture surface energy of magnesispinel composites. Composit 38. Lawn B., FractureSeries. 1993, Cam 172

PAGE 190

39. Chantikul P., Anstis G.R., Lawn B.R., Marshall D.B., A Critical Evaluation of Journal of the American Ceramic Society, 64 [9]: 539-543 (1981) 40. Anstis G.R., Chantikul P., Lawn B.R., Marshall D.B., A Critical EvaMeasurements. Journal of the American Ceramic Society, 64 [9]: 533-538 (141. Muralithran G., Ramesh S., The effects of sintering temperature on the propertie phosphates and the implications for bone tissue engineering. Materials Science 390 [1-2]: 246-254 (2005) 43. Thangamani N., Chinnakali K., Gn Indentation Techniques for Measuring Fracture Toughness: II, Strength Method. luation of Indentation Techniques for Measuring Fracture Toughness: I, Direct Crack 981) s of hydroxyapatite. Ceramics International, 26: 221-230 (2000) 42. Case E.D., Smith I.O., Baumann M.J., Microcracking and porosity in calcium and Engineering A-Structural Materials Properties Microstructure and Processing, anam F.D., The effect of powder processing on densification, microstructure and mechanical properties of hydroxyapatite. wood Cliffs, New Jersey: Prentice-Hall Inc. 45. Fan C.F., Hsu S.L., Application of the Molecular Simulation Technique to : 6244-6249 (1991) 46. Bennison S.J., Jagota A., Smith C.A., Fracture of Glass/Poly(vinyl butyral) Society, 82 [7]: 1761-1770 (1999) 47. Wu H., Yan X., Interlaminar stress modeling of composite laminates with finite 58 (2005) 48. Mallick P.K., Fiber-Reinforced Composites: Materials, Manufacturing, and 1959, London: McGraw-Hill, Inc. 50. Gruninger M., L.B., Farabaugh E., Wachtman Jr. J.,, Measurement of Residual American Ceramic Society, 70 [5]: 344-348 (1987) Ceramics International, 28: 355-362 (2002) 44. Allcock H.R, Lampe F.W., Contemporary Polymer Chemistry. 1990, Engle Generate the Structure of an Aromatic Polysulfone System. Macromolecules, 24 (Butacite(R)) Laminate is Biaxial Flexure. Journal of the American Ceramic element method. Journal of Reinforced Plastics and Composites, 24 [3]: 235-2 Design. 2nd ed. 1993, New York: Marcel Dekker, Inc. 49. Timoshenko S., Woinowsky-Krieger S., Theory of Plates and Shells. 2nd ed. Stress in Coatings on Brittle Substrates by Indentation Fracture. Journal of the 173

PAGE 191

51. Ivanova T.I., Frank-Kamenetskaya O.V., Kol'tsov A.B., Ugolkov V.L. "Crystal Decompo Structure of Calcium-Deficient Carbonated Hydroxyapatite Thermal sition." Journal of Solid State Chemistry, 160: 340-349 (2001) 2. biocompatibility of polyetherketone and polysulfone composites. Journal of failure in soda lime glass, Journal of the American Ceramic Society, 62 [11-12]: York: John Wiley & Sons, Inc. 5Wenz L.M., Merritt K., Brown S.A., Moet A., Steffee A.D., In vitro Biomedical Materials Research, 24 [2], 207-215 (1990) 53. Mecholsky J.J., Gonzales A.C., Frieman S.W., Fractographic analysis of delayed 577-580 (1979) 54. Sperling L.H., Introduction to Physical Polymer Science, 3rd ed. 2001, New 174

PAGE 192

BIOGRAPHICAL SKETCH Prior to entering graduate school at the University of Florida, Clifford A Wildid his undergraduate studies at the College of William and Mary in Williamsburg, VA, earning a bachelors degree in chemistry from the in 2000. While at William and Mary he was letter winner for the Tribe baseball team. After graduation, he decided to further his education and began graduate stuthe University of Florida. He joined the research group of Dr. John Echolike Jr. Cliff was inducted into Keramos, the ceramic honor society, son II dies at during the spring semester 2002. ff in In May of 2003, Cliff received a Master of Science degree in Materials Science and Engineering. Cliff was awarded the Robert C. Pittman Fellowship in August 2003. Clisuccessfully defended his doctoral dissertation on June 23, 2005, and received his doctorate during commencement ceremonies in August 2005. Cliff is married to Leslie Hoipkemeier Wilson, who also received her doctorate August 2005. 175


Permanent Link: http://ufdc.ufl.edu/UFE0011396/00001

Material Information

Title: Design, Fabrication, and Characterization of Laminated Hydroxyapatite-Polysulfone Composites
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011396:00001

Permanent Link: http://ufdc.ufl.edu/UFE0011396/00001

Material Information

Title: Design, Fabrication, and Characterization of Laminated Hydroxyapatite-Polysulfone Composites
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011396:00001


This item has the following downloads:


Full Text











DESIGN, FABRICATION, AND CHARACTERIZATION OF LAMINATED
HYDROXYAPATITE-POLYSULF ONE COMPO SITE S














By

CLIFFORD ADAMS WILSON II


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

By

Clifford Adams Wilson II


































To my wife, Leslie, and my parents, Cliff and Marilyn, for your love and encouragement
throughout this entire process
















ACKNOWLEDGMENTS


I would especially like to thank Dr. John Mecholsky Jr. for his guidance and for

always having his door open for students.

I would also like to thank the other members of my supervisory committee, Dr.

Anthony Brennan, Dr. Kenneth Anusavice, Dr. Wolfgang Sigmund, and Dr. Bhavani

Sankar, for their input and advice, which helped guide me through this proj ect.

I need to acknowledge the contributions to this work made by Leslie Wilson, of the

Department of Materials Science and Engineering, who prepared the initial PSu solutions

used for solvent casting, Dr. Sukj oo Choi of the Department of Mechanical and

Aerospace Engineering, who performed the finite element analysis presented in this

study, and Gil Brubaker of the Particle Engineering Research Center, who performed the

particle size analysis.

I want to thank Allyson Barrett and Ben Lee of the Department of Dental

Biomaterials for their assistance. Finally, I would like to thank the all other graduate

students who helped made coming to lab everyday a pleasure, but especially, Dr. Tom

Hill, and Dr. Alvaro Della Bona who helped to become acclimated when I first arrived at

the University of Florida.



















TABLE OF CONTENTS


Page

ACKNOWLEDGMENT S ................. .............._ iv..._.__....


LI ST OF T ABLE S ........._.. ..... ._._ .............._ viii..


LI ST OF FIGURE S .............. ...............x.....


AB S TRAC T ..... ._ ................. ..........._..._ xviii..


CHAPTERS


1 BACKGROUND AND RATIONALE ................. ............_........1..........


Requirements for an Ideal Hard Tissue Regeneration Device............... ..................
Material s Selection .............. .. ...............4....
Research Rationale and Objective ..................._...__ ......... ............1

2 HYDROXYAPATITE ............ ..... ._ ...............12....


Background on Processing of Hydroxyapatite ................. .............................12
Starting M material s........._.. ..... ._ ...............13....
Processing M ethods ........._.. ..... ._ ...............13.....
Tape Casting .............. ...............13....
Cold Pressing ........._..... .... ...._ ...............14.....
Organic Burnout and Densification ............_...... .__ ......__...........1
Sintering ............... ... .._ ...............17....
Specimen Refinement ............ ..... .._ ...............18...
HA Characterization Methods ............ .....___ ...............18..
Particle Size Analysis .............. ...............18....
X-Ray Diffraction ........._.._ _..... ._ ._ ...............18....
Density and Porosity ........._..... ........._ ._ ...............18....
Elastic Modulus and Poisson's Ratio............... ...............19.

Optical Microscopy and Lighting Effects .....__.___ ..... ... .__ .........._..._..20
Scanning Electron Microscopy ........._.._ ..... ._._ ....._._............2
M echanical Testing M ethods................. ..... ....... ....... ........2
Establishing Baseline Properties for Monolithic Hydroxyapatite ......................20
Indentation .............. ...............21....
Biaxial Flexure ................. ............ ...............22.......
Work of Fracture and Toughness ................. ...............22........... ...
Fracture Toughness ................. ...............23.................












Finite Element Analysis............... ...............25
Statistical Analysis............... ...............26
Results and Discussion .............. .... ...............26..
Starting Powder Characterization ............... .......... ...............2
Development of a Tape Casting Method for Hydroxylapatite. ................... ........28
Burnout and Sintering Process................. ............... ... .......3
Characterization and Mechanical Properties of Post-Burnout Hydroxyapatite..3 6
Problems with Specimens Fired to 10000C .....__ ................ .................3 8
Mechanical Properties of 10000C Specimens .......................__ ...............38
Constant Fracture Toughness............... ...............4
Firing Study .............. .......... ............. .......4
Sintering Temperature Effect on Hardness............... .... .............4
Sintering Temperature Effect on Biaxial Flexure Strength ................ ...............44
Sintering Temperature Effect on Density .............. ...... ...............44
Qualitative Determination of Hydroxyapatite Decomposition ................... ........46
Sintering Temperature Selection............... ...............4
Optimization of Sintering ...................... .. ...............4
Density and Elastic Modulus of 12000C Specimens .............. .....................4
Hardness of 12000C Specimens ......_......_.._.. .........._. ............5
Biaxial Flexure Strength of 12000C Specimens .............. ....................5
Work of Fracture and Toughness. ....._.._.. ..... .._._. ....._ .. ..........5
Fracture Toughness of 12000C Specimens ....._.._.. ........... ........_.......54
Monolithic Hydroxyapatite Specimens............... ...............5

3 LAMINATE FABRICATION............... ..............5


Design and Nomenclature............... ..............5
Laminate Design ................. .......... .... ...............59.....
Laminate Geometry and Nomenclature ................ .............. ......... .....59
Solvent Casting of Polysulfone ................. ...............60................
M material s .............. ...............61....
Methods .............. ... ........ ..... .........6
Solvent Casting of Polysulfone ................. ...............62................
Specimen Preparation .............. .. ..... .... ............6
Characterization of PSu Films and Laminates ................. ............... ...._...71
Results and Discussion ...................... ............... ........7
Problems With the Solvent Casting Methods ......... ................ ...............72
Thermal Analysis of Polysulfone Layers ................. ............... ......... ...79

4 LAMINATE THEORY .............. ...............91....


Back ground ................. ...............9.. 1..............
M ethods .............. ...............92....
Laminate Theory ................... ........... ...............92.......
Finite Element Analysis (FEA) ................. ...............95................
Material Modeling Parameters............... ...............9
Results and Discussion .............. ...............97....












Comparison of FEA and Laminate Theory Models ................. .....................97
Predicting Laminate Behavior............... ...............10

5 LAMINATE BEHAVIOR ........._.___..... .___ ...............111....


M ethods .........._.... .. .. ..._ ..... ..._._ .. .. .......... 1
Mechanical Testing and Characterization of Laminates ........._._... ................11 1
Laminate Testing Variables ......__....._.__._ ......._._. ............12
Results and Discussion ........._..... .... ....._ _._ ...............114...

Comparison of Laminates to Monoliths ....._.__._ ........__. ........_........1 14
Testing of Laminate Parameters ......__....._.__._ ......._._. ...........18
Determination of Failure Mechanisms .............. ...............135....
Fracture Analysis of 400-200-800 Laminates .............. ..... .... ...............139
Failure Mechanism of Laminates with Outer HA Layers < 400 pm ........._......155

6 CONCLUSIONS................ .............16


LIST OF REFERENCES ....__. ................. ...............170 .....


BIOGRAPHICAL SKETCH ................. ...............175......... ......


















LIST OF TABLES


Table Page

2.1 Mechanical data and fractography measurements for HA specimens fired
to 1000.C .............. ...............42....

2.2 Biaxial flexure strength data for specimens sintered at 12000C............................52

2.3 Work of fracture values for specimens sintered at 12000C............... .................5

2.4 Fracture toughness values for 12000C hydroxyapatite specimens ........................55

3.1 Hardness of laminates made through the matching halves method. ......................76

3.2 Weight percent solvent retained results for PSu layers fabricated at all
three drying temperatures. ............. ...............79.....

3.3 Comparison of laminates fabricated with polymer layers composed of both
PSu and sPSu .............. ...............90....

4.1 Initial comparison of laminate models to FEA for monolithic HA. ......................98

4.2 Comparison of corrected laminate models to FEA for monolithic HA. ................98

4.3 Comparison of maximum principal stress calculated through FEA and
Laminate Theory for various laminate geometries. ............. .....................9

4.4 Contact stress field radii for each laminate geometry that yields the same
results as the FEA ................ ...............100...............

4.5 Maximum stresses on the tensile surface of each layer for all laminate
geometries experimentally tested in this study. ................. ................10

5.1 Comparison of the average mechanical property values for the monolithic
HA versus the 400-200-800 laminates ................. ...............116........... ...

5.2 Hardness data for initial flaw size laminates. ................ ......... ................11 8

5.3 Mechanical property data versus indent loads ................. .......... ...............119

5.4 Hardness data for the laminates with different PSu layer thicknesses. .................1 22










5.5 Apparent toughness values versus increasing PSu layer thicknesses ................125

5.6 Hardness data for laminates with different outer HA layer thicknesses. .............127

5.7 Apparent fracture toughness comparison for different outer HA layers. .............131

5.8 Hardness data for the four laminate groups prepared with thicker middle
H A layers. ............. ...............13 1....

5.9 Comparison of flaw sizes for monolithic HA and the outer HA layer for
400-200-800 laminates............... ...............13

5.10 Middle HA layer failure stress calculations for the 400-200-800 laminates. ......144

5.11 Failure stress calculations for laminates with outer HA layers < 400 Cpm...........161


















LIST OF FIGURES


Figure Page

1.1 A plot of fracture toughness versus Young' s modulus of biomaterials
being developed for bone replacement adapted from Suchanek and
Y oshimura [ 6]. ............. ...............6.....

2.1 Weight % of each constituent in the tape casting slurry..........._.._.. ........._.._.. ..15

2.2 Green tape processing, shown are punched green discs (top left), the 25
mm (1") knife edge punch (middle), and scrap tape (top right). The section
of tape processed (bottom) is 30 cm (12") x 11 cm (4.5") and produced 36
green discs............... ...............15.

2.3 Organic burnout cycle performed on the consolidated discs produced by
cold pressing hydroxyapatite green tape. ........._.._.. ......._ ........__. .......16

2.4 Oblique allowed surface features to be seen (bottom) in greater detail than
with overhead lighting (top)............... ...............21.

2.5 Schematic representation of fracture markings that result from a brittle
fracture adapted from Mecholsky, et al [53]............... ...............24..

2.6 Particle size distribution for the starting hydroxylapatite powder ................... ......26

2.7 XRD spectrum for the as-received hydroxylapatite powder. .............. .... ........._...27

2.8 Effects of applied pressure (top) and hold time (bottom) on the reduction
in thickness of stacked green tapes during cold pressing. ............. ...................3 1

2.9 Optical micrographs of post-cold pressing edge delamination (top left),
delamination at the center (bottom left), full consolidation (top right), and
post-sintering center delamination. .............. ...............32....

2.10 The reduction in thickness of the stacked green tapes is plotted versus wt%
binder in the tape casting slurries (top), and binder:plasticizer ratio
(b ottom) ................. ...............34.................

2.11 TG/DTA data curves for a pre-burnout green tape sample (top) and a post-
burnout hydroxyapatite specimen. .............. ...............37....










2.12 XRD spectrum for HA specimens burned out at 10000C (top) and
specimens sintered at 12000C (bottom) .............. ...............39....

2.13 XRD spectrum for HA specimen fired to 12000C (top) and for naturally
occurring HA (bottom) [8]. Circles (*) indicate peaks found on both
spectra, the square (m) indicates the peak that is unique to the 12000C
spectrum .............. ...............40....

2.14 Optical micrographs of a Vickers indent for a 10000C specimen (left) and
crack propagation through a Vickers indent. ............. ...............42.....

2.15 Log-log plot of failure stress versus indent load for HA specimens fired to
10000C .............. ...............43....

2.16 A plot of hardness versus sintering temperature for the hydroxyapatite
specim ens. ............. ...............45.....

2.17 A plot of failure stress versus sintering temperature for the hydroxyapatite
specim ens. .............. ...............45....

2.18 Plot of density versus sintering temperature for hydroxyapatite specimens..........46

2.19 Color comparison of specimens sintered at five different temperatures.
Each row of specimens was fired at the temperature indicated to the right. .........47

2.20 Plot of failure stress versus hold times at 12000C ..........._... ......_. ............50

2.21 Plot of failure stress versus ramp rate sintering temperature of 12000C. .............50

2.22 SEM micrograph of the microstructure of the hydroxyapatite specimens
sintered at 12000C............... ...............51.

2.23 Comparison of stress-strain curves for indented and non-indented HA
specim ens. .............. ...............53....

2.24 All specimens tested in biaxial flexure fractured into either two (left) or
three (right) pieces. ............. ...............54.....

2.25 Plot of Kc versus Ksi for the hydroxyapatite specimens sintered at 12000C..._....56

2.26 SEM micrograph of the initial flaw caused by an indent (black bar) for a
12000C HA specimen. The outer boundary of the initial flaw (white
arrows) and a twist-hackle marking (black arrow) are shown ............... ...............56

2.27 Optical micrographs of an entire fracture surface (top) and initial flaw
(bottom) of a 12000C specimen indented with a 3.35 kg indent. The white
arrows indicate the outer edge of the critical flaw. ........._.._ ...... .._. ...........57

3.1 Schematic diagram of HA/PSu laminate ........... ..... .._ ........._.......60










3.2 Polysulfone .............. ...............61....

3.3 Sulfonated Polysulfone ................ ...............62........... ....

3.4 Step-by-step schematic of laminate fabrication method 1: The Matching
Halves M ethod ................. ...............68.......... ......

3.5 Step-by-step schematic of laminate fabrication method 2: The Bottom Up
M ethod ................. ...............69.......... ......

3.6 Step-by-step schematic of laminate fabrication method 3: The Pre-
Fabricated PSu Layer Method .............. ...............70....

3.7 Optical micrographs of (left) a large surface bubble formed during
laminate fabrication, and (right) a small bubble (black arrow) within a PSu
layer that caused a HA layer to fracture (white arrows). ............. ....................73

3.8 Left: Optical micrograph of a PSu layer that peeled from the HA disc
during drying. Right: Higher magnification image of the PSu layer (white
arrow) that peeled due to fracture of the HA layer (black arrows) which is
still bonded to the PSu layer. ............. ...............73.....

3.9 Optical micrographs of a large hole that formed during indentation (left),
and of a laminate with a defective PSu layer containing a large open cavity
(ri ght) .............. ...............76....

3.10 SEM image of bubbles (black arrows) which formed during laminate
fabrication using the bottom up technique ................. ...............77......_._. .

3.11 Schematic drawing of chipping failures that occur when fabricating thick
PSu layers through the bottom up method. .............. ...............78....

3.12 Complete TG curve for an as received PSu pellet. ................ ..................8

3.13 A comparison of TG curves PSu films dried at the three different
tem peratures. .............. ...............8 1....

3.14 TG curves for six different PSu films dried at 700C............... ..................8

3.15 TG curve for a PSu sample taken from a delaminated region of a laminate
fracture surface............... ...............82

3.16 DSC curves generated through three heating-cooling cycles for an as
received PSu pellet ................. ...............85........... ....

3.17 DSC curves generated through three heating-cooling cycles for a PSu film
dried at 700C. ............. ...............85.....

3.18 A plot of elastic modulus vs. drying temperature for PSu layers. .........................86










3.19 A plot of break stress vs. drying temperature for PSu layers. .........._... ..............86

3.20 A plot of elongation at break vs. drying temperature for the PSu layers............_...87

3.21 A plot of failure loads and stresses vs. laminate fabrication method and
drying temperature. Units of the y-axis are different for each series.. ........._........87

3.22 A plot of laminate failure loads and stresses vs. surface polishing medium.
The ordinate axis has units are unique to each series plotted. ............. ................89

4.1 A schematic representation of a laminate indicating the mathematical
variables required for laminate theory calculations. ............... ...................9

4.2 The Microsoft Excel spreadsheet designed to calculate the laminate stress
distribution for the 400-200-800 laminate ....._____ .... ... .__ ...........__.....96

4.3 Graphical representation of the stress field resulting from a 1 N applied
load on a 400-400-800 HA/PSu laminate. ............. ...............97.....

4.4 Stress distribution calculated using laminate theory resulting from a 1N
load applied to a 400-200-800 laminate. Laminate layers are drawn to
scale, with the 2.00 thickness representing the outer HA layer tensile
surface. ................. ................. 102........ ....

4.5 Maximum stresses for the 400-200-800 laminate as (a) a plot of individual
points and (b) a graphed smoothed curve. ............. ...............104....

4.6 Maximum stress curves for laminates as a function of varying polymer
layer thickness ................. ...............107................

4.7 Maximum stress curves for laminates as a function of varying outer HA
layer thickness ................. ...............107................

4.8 Maximum stress curves for laminates as a function of varying middle HA
layer thickness ................. ...............108................

4.9 Maximum stress curves for various laminate geometries having a total
laminate thickness of 2.0 mm. ............. ...............108....

4.10 A plot of maximum tensile stress values versus the ratio between the
thicknesses of the outer HA and PSu layers. ..........._ ..... .__ ................109

4.11 A plot of maximum tensile stress versus the ratio of thicknesses between
the PSu and middle HA layers. .............. ...............109....

4.12 A plot of maximum tensile stress versus the ratio of thickness between the
outer and middle HA layers. ...........__......___..... ............1










5.1 Comparison of load displacement curves for the 400-200-800 laminates
compared with monolithic HA. ...._.._.._ .... .._._. ....._.._...........16

5.2 Comparison of mechanical properties measured for monolithic HA as the
400-200-800 laminates. The units of the ordinate axis vary for each
property listed with units designated in the column headings. ........._.._...............117

5.3 Comparison of failure stress for 400-200-800 laminates versus both
indented and non-indented monolithic HA specimens. ........._._.... ......._.......117

5.4 Lateral cracking seen during indentation using a 9.35 kg load (left), and
the resulting chip-out that occurs during loading (right). ................ ................11 9

5.5 A plot of failure stress versus loading rate. ................ .............................121

5.6 A plot of absorbed energy at failure (toughness) versus loading rate for
400-200-800 laminates............... ...............12

5.7 A plot of failure loads versus PSu layer thickness. Monolithic HA failure
load is plotted as a PSu layer thickness of 0.................................... 123

5.8 A plot of failure stress vs. PSu layer thickness. Monolithic HA is plotted
as a PSu thickness of 0.................. ...............124..............

5.9 A plot of stress vs. outer HA/PSu layer thickness ratio for a comparison of
experimental values to laminate theory predictions ................. .........__ ......124

5.10 A plot of work of fracture vs. PSu layer thickness ...........__... ............ ........125

5.11 A plot of absorbed energy at fracture (toughness) vs. PSu layer thickness......... 126

5.12 A plot of failure loads versus outer HA layer thickness. The total
thickness of the specimens is shown in parenthesis ................. .....................128

5.13 A plot of failure stresses versus outer HA layer thickness. Monolithic HA
is plotted as a 2000 Cpm thick HA layer ................. ...............129.......... .

5.14 A plot of work of fracture versus outer HA layer thickness. .............. ... ............130

5.15 A plot of toughness versus outer HA thickness ................_ ................. ...._130

5.16 Failure loads for laminates with thicker middle HA layers. ............. .................134

5.17 Failure stresses for laminates with thicker middle HA layers .............................134

5.18 Experimental failure loads for all the laminate geometries having a total
laminate thickness of 2.0 mm ............. .............135.....










5.19 An optical micrograph of 400-200-800 laminate fracture surface. The
locations of the initial indent flaw (white arrow), middle layer flaw (black
arrow), and loading piston contact are all indicated. ............. .....................3

5.20 Optical micrographs of a laminate still intact after being loaded to failure
in biaxial flexure. Left: Failure occurred from the initial indent (indicated
by the arrows) in the outer HA surface, Right: Examination of the edge of
the specimen shows failure of the outer HA layer, delamination along the
interface, and failure of the middle HA layer. ..........___......._ ..............137

5.21 SEM image of punch through failure of a 400-400-800 laminate. The
remnants of the indented outer HA layer are indicted by the black arrows,
failure of the PSu layer were the loading piston breached all five layers is
indicated by the white arrow ................. ...............138........... ...

5.22 Optical micrograph of the initial flaw size (black arrows) produced by a
3.35 kg indent (white arrow) in a 400-50-800 laminate .............. ...................140

5.23 SEM image of the initial flaw produced in a 400-100-800 laminate
produced by a 3.35 kg indent load ................. ...............140........... .

5.24 Load-displacement curves for two non-indented laminates. Outer HA
layer failure loads are calculated either from the initial drop (top) or by
extrapolating the initial slope (dashed line) and estimating when the slope
change begins (bottom)............... ...............14

5.25 Failure stresses of the outer HA layer calculated for three different initial
flaw sizes............... ...............143.

5.26 Failure stresses of the outer HA a layer for laminates built with different
PSu layer thicknesses ................. ...............143................

5.27 Optical micrograph of the fracture oringin of a 400-200-800 laminate. The
center of the flaw is indicated by the white arrow ................. ................. ... 145

5.28 Optical micrograph of a middle layer fracture origin a 400-100-800
laminate. The specimen has been sputter coated to make the surface
marking stand out ................. ...............145................

5.29 SEM image of the middle HA layer fracture origin ................. ................. .. 146

5.30 SEM image of a 400-100-800 laminate showing contact damage (white
arrows) from the loading piston which occurred prior to propagation of
the primary fracture (black arrows) which was reinitiated near the HA/PSu
interface (dashed arrow). ............. ...............148....

5.31 Optical micrograph of secondary cracks (arrows) which occurred through
punch failure after primary failure from bending. ............. ....................14










5.32 Optical micrograph of a PSu layer of a 400-200-800 laminate fractured at
a crosshead displacement rate of 0.25 mm/min. The specimen was sputter
coated prior to analysis. ............. ...............149....

5.33 Optical micrograph of a PSu layer of a 400-200-800 laminate fractured at
a crosshead displacement rate of 25 mm/min. .................. ...............14

5.34 Optical micrographs of the PSu layer features through overhead (top) and
oblique lighting (bottom) ........._._. ...._..._ ...............150...

5.35 SEM images of brittle failure of the PSu layer, the pore (indicated by the
arrow in the left image) shows a twist hackle marking (arrow in the right
image) characteristic of brittle fracture ................. ...............150........... ...

5.36 SEM image of ductile deformation of the PSu layer prior to failure at the
center of the feature (black arrow) ................. ...............151..............

5.37 SEM image of two side-by-side features within the PSu layer that failed in
different manners, a twist hackle marking (solid arrow) indicates the left
feature failed in a brittle manner, while the pullout deformation (dashed
arrow) of the right feature indicates ductile failure. ................ ................. .. 15 1

5.38 Optical micrograph of a sPSu/HA laminate showing the depth of
penetration of the casting solution during solvent casting ........._..... ..............154

5.39 SEM images showing the penetration depth of PSu into the middle HA
layer. The fracture surface is shown top left, the box indicates the area
where the top right picture was taken, the bottom picture showing PSu
pullout was taken from the area indicated by the box in the top right image......156

5.40 SEM images of PSu bridging of small cracks in the HA layers. These
images were taken of the secondary interface crack shown in Figure 5.3 1.........157

5.41 SEM images showing PSu bridging of the HA/PSu interface. PSu fibrils
are indicated by the arrow. .........._ ....._ __ ...............157.

5.42 SEM image of PSu bridging of the HA/PSu interface. ............. ................158

5.43 Optical micrographs showing the pullout of the PSu layers during fracture
of the laminates. .........._ __..... ._ ...............158...

5.44 SEM images of brittle PSu layer originating from the HA/PSu interface.
The arrow indicates the location of the origin. ..........._.....__ ..............159

5.45 SEM images of a PSu layer fracture origin. The location of the origin is
indicated by the arrows on both images, failure occurred within the PSu
layer and not from the interface. .............. ...............159....










5.46 Optical micrograph of the fracture origin (solid white arrows) of a 100-
100-1600 laminate, the large damaged area is caused by the indent (dashed
arrow) ........._._._.. ...___.. ...............161....

6.1 Addition of the HA/PSu laminates designed in this study to the graph of
available biomaterials for bone replacement developed by Suchanek and
Yoshimura [6] .............. ...............163....
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DESIGN, FABRICATION, AND CHARACTERIZATION OF LAMINATED
HYDROXYAPATITE-POLYSULF ONE COMPO SITES
By

Clifford Adams Wilson II

August 2005

Chair: John J. Mecholsky Jr.
Major Department: Materials Science and Engineering

There exists a need to develop devices that can be used to replace hard tissues, such

as bone, in load-bearing areas of the body. An ideal hard tissue replacement device is

one that stimulates growth of natural tissues, and is slowly resorbed by the body. The

implant is also required to have elastic modulus, strength, and toughness values similar to

the tissues being replaced.

Hydroxyapatite (HA) is the primary mineral phase of bone and has the potential

for use in biomedical applications because it stimulates cell growth and is resorbable.

Unfortunately, HA is a relatively low strength, low toughness material, which limits its

application to only low load-bearing regions of the body. In order to apply HA to greater

load-bearing areas of the body, strength and toughness must be improved through the

formation of a composite structure.

The goal of this study to show that a composite structure formed from HA and a

biocompatible polymer can be fabricated with strength and toughness values that are


XV111









within the range necessary for load-bearing biomedical applications. Therefore,

Polysulfone-HA composites were developed and tested. Polysulfone (PSu) is a hard,

glassy polymer that has been shown to be biocompatible. Composites were fabricated

through a combination of tape casting, solvent casting, and lamination. Monolithic HA

and laminate specimens were tested in biaxial flexure. A unique laminate theory solution

was developed to characterize stress distributions for laminates. Failure loads, failure

stress, work of fracture, and apparent toughness were compared for the laminates against

monolithic HA specimens.

Initial testing results showed that laminates had a failure stress of 60 & 10, which is

a 170% improvement over the 22 + 2 MPa failure stress for monolithic HA. The work of

fracture was improved by 5500% from 11 & 2 for the monolithic HA to 612 & 240 for the

laminates. Work of fracture values gave the laminates an apparent fracture toughness of

7.2 MPa~ml/2 COmpared to 0.6 MPa~ml/2 for the monolithic HA.

Laminates with different geometries were built and tested in an attempt to optimize

the strength and toughness of the composites. Laminate behavior was characterized as a

function of initial flaw size, HA layer thickness, PSu layer thickness, and stressing rate.

The failure stress of the laminates was maximized at a value of 108 & 14 MPa, which is a

400% improvement over monolithic HA, and close to the 120 160 MPa range reported

for bone. The work of fracture of laminates was maximized at 724 & 206 J/m2, which is a

6400% improvement over monolithic HA, and yields an apparent fracture toughness

value of 7.5 MPa~ml/2. This apparent toughness value is within the 2-12 MPa~ml/2 range

for bone, and an 1 100% improvement over the fracture toughness of monolithic HA.















CHAPTER 1
BACKGROUND AND RATIONALE

Requirements for an Ideal Hard Tissue Regeneration Device

Tissue Response to Implanted Materials

Any material implanted in the body will illicit a response from the surrounding

tissues. Bioceramics, when implanted in the body will elicit one of four potential tissue

responses [8]. If the material is toxic, the surrounding tissues will die. If the material is

nontoxic and biologically inactive, the body will surround the material with a fibrous

capsule but not experience any negative consequences due to the presence of the

implanted materials. A material that elicits this type of tissue response is called

biocompatible.

A third tissue response occurs when an implanted material is nontoxic and

biologically active. These materials are called bioactive because they form an interfacial

bond with surrounding tissues. The final tissue response occurs when a nontoxic

implanted material dissolves in the body. These materials are called bioresorbable [8].

Definition of the Gold Standard

In order for successful use of a material or composite for hard tissue regeneration,

the material of composite must fulfill some basic requirements. These requirements are

not explicit rules that govern research and development; rather they are ideals that have

been set for the eventual gold standard. Through the study of hard tissues and hard tissue

replacements by Hench [1-3], Bonfield [4], and others [5-7], the gold standard has been

defined as a material, which will possess a few basic traits. The ideal hard tissue implant









must be biocompatible, bioactive, and mechanically identical to the tissue being replaced.

Additionally, the ideal hard tissue replacement would be resorbed and replaced by the

surrounding natural tissue.

Biocompatibility

Biocompatibility will be defined as the ability of a material to perform a desired

function in the body without causing any adverse effect upon natural tissue with which it

comes into contact [8]. A material is required that will not cause any type of

inflammatory or destructive response in the body. No matter how well a material

performs the required mechanical functions, if there is an adverse response to the material

within the body, the material is unacceptable for use. Biocompatibility is the greatest

limiting factor in biomaterial design. If the entire spectrum of known materials was

available for use, artificial bones and hard tissues replacement devices would probably

already exist. However, materials that are toxic or reactive in the natural environment

(for example, corrosion of metals over time when implanted in the body) cannot be used

due to adverse effects on surrounding tissues. Therefore, research is limited to materials

that are biocompatible.

Biocompatible materials currently used for biomedical applications include

metals, such as titanium, stainless steel and cobalt chromium, bioinert ceramics, such as

alumina, zirconia, calcium phosphates and inorganic glasses, and a polymers like

polymethyl methacrylate (PMMA), polyethylene (PE) polylactic acid (PLA), and

polysulfone (PSu) [8]. These are just a few examples of biocompatible materials used in

the body and should not be considered a comprehensive list since new materials are

continually being developed and approved for use in the body.









Bioactivity and Bioresorbtion

Bioactivity refers to the ability of a material to promote regeneration of the natural

tissues [9]. Regeneration is favored over replacement due to the long-term degradation of

replacement materials, which causes a need for repetitive procedures over the life of a

patient. Another reason for bioactivity is that is can be a useful tool in the implantation

of a hard tissue device. The conclusion on the biochemical reactions occurring at the

tissue/implant interface of research by Hench [1-3] and Suchanek [6] is that activation of

natural tissue generation and the subsequent growth in and around the implant will aid in

anchoring the implant into place. Furthermore, if bioactive materials are broken down by

the surrounding environment, the body may then resorb it. Bioresorbable materials are

designed to degrade gradually over a period of time, and be replaced by the natural host

tissue [1]. There is a very narrow range of materials that are both biocompatible,

bioactive, and bioresorbable and can be used for hard tissue replacement. This range of

materials is currently limited to bioceramics such as calcium phosphates and bioactive

glasses [2].

Matching Mechanical Properties

The third maj or requirement of these materials is that they have mechanical

properties that are similar to those of the natural tissue in which they are in contact.

Problems arise from stress shielding when there is a mismatch in properties from one

material to the next [8]. For example, replacing bone with a higher modulus material

leads to a lower strength healed bone around the implant due to the implant carrying

more of the load than the surrounding bone. Since bone remodeling is a continuous

process that the body uses it to adapt to changes in stresses. For example, if a normally,









inactive person were to start exercising on a regular basis, the body would remodel bones

to compensate for the greater stresses felt during exercise. In the case of a higher

modulus material being implanted to replace bone, the body would lay down weaker

bone in the areas surrounding a higher modulus implant because the implant would bear

most of the applied loads.

There is also a need for implanted materials to have toughness values greater than

those of the surrounding tissues, such that mechanical failures will not occur within the

implanted materials before they occur in the surrounding tissue. A failure of the implant

would lead to increased loading of the weakened, surrounding tissue and compound the

problems that are being resolved.



Materials Selection

Hydroxyapatite

Tissue regeneration is the most limiting factor in materials selection as there are

few materials capable of inducing bone growth. One of the better-studied materials is

hydroxyapatite (HA). Hydroxyapatite is the primary inorganic component of all calcified

tissues existing in the human body [8].

Synthetic HA powders can be prepared in a number of different ways. Processes

like sol-gel [6], precipitation [10], and solid-state reactions [11] have all been used to

produce to produce HA powder. HA powder fabrication is a common process such that

HA is readily available through a number of commercial supply companies. Numerous

studies have show that synthetic HA is bioactive and can be used as a bone replacement

[1, 6].









HA containing a Ca:P ratio of 1.67 is bioactive [2, 6] and theoretically can be used

for bone tissue replacement. Unfortunately, characterization of synthetically made HA

shows that it has mechanical properties that are inferior to the necessary properties to

sustain load-bearing applications in the body. Dense forms of HA have been fabricated,

and a wide range of mechanical data have been reported. Bending strengths from 38-250

MPa, tensile strengths from 38-300 MPa, and fracture toughness values in the range of

0.8-1.2 MPa~ml/2 have been reported [6]. Young's Modulus values have been reported in

the range of 10-30 GPa [5, 6]. Porous versions of HA have mechanical properties that

are much lower [12, 13]. Bending and tensile strengths for these materials have been

reported in range of 2-1 1 MPa and around 3 MPa, respectively [6]. Fracture toughness

values have been shown to decrease with increasing porosity [13]. These materials are

capable of matching the bending and tensile strengths of compact human bone, which

have been reported to be as great as 160 MPa and 124-170 MPa, respectively [6].

However, the fracture toughness of HA is well below that reported for compact bone,

which is in the range of 2-12 MPa~ml/2 [1, 13] HA, on its own, lacks adequate

mechanical properties to be applied to maj or load bearing applications.

HA Composites

Since monolithic HA lacks mechanical properties sufficient to withstand load-

bearing applications within the body, HA must be combined into composite structures

with other materials in order to meet the load-bearing requirements for bone replacement.

The goal of forming the composite is to increase both the strength and toughness of HA

to levels more consistent with that of natural bone.










HA was been formed into a composite with titanium, zirconia, and alumina [6], as

well as being reinforced through the addition of particles [14], whiskers [15], and fibers

[6]. All of these materials have significant limitations, which limit their usefulness. The

greatest problem is the modulus mismatch with bone. Figure 1.1, recreated from

literature [6], shows the relationship of the mechanical properties of these materials to the

mechanical properties of dense HA and to bone. Clearly these materials will need to be

further refined for use as a hard tissue replacement.


15





10
Bone



H Ap-based
a5 ceramic composites
HAp/bioglass

dense HAp

HAp/polymers

0 50 100 150
Young's Modulus (GPa)

Figure 1.1 A plot of fracture toughness versus Young' s modulus of biomaterials being
developed for bone replacement adapted from Suchanek and Yoshimura [6].



The only other materials incorporated with HA to form composite materials with

modulus and toughness values similar to bone are polymers. A few of the polymers that

have been incorporated successfully into the HA matrix include polyethylene, poly(L-









lactide), polysulfone as well as some biopolymers such as collagen [6, 8, 9]. Bonfield and

his group have reported extensive work on PE/HA composites [4, 16-18].

HAPEXTM is an HA/PE composite that is FDA approved for use in the inner ear.

HAPEXTM and other HA/polymer composites are produced by adding HA as the

reinforcement to the polymers. HA/PE composites have an elastic modulus in the range

of 1-8 GPa [6] and fracture toughness values that overlap that of bone, see Figure 1.1.

Since polymers cannot withstand the temperature required to strengthen HA through

sintering, HA/polymer composites are confined to forming through addition of HA

particles into the polymer matrix. The resulting composites have greatly improved

toughness values over monolithic HA but the high polymer content lowers the strength

below that required for load-bearing applications. Clearly, another composite structure

composed of HA and polymer is required.

Biological Structures

Fabrication of strong, tough materials has already been achieved many times over.

Nature, in many ways, has taken materials that on their own might seem useless and

combined them to create structures with extraordinary mechanical properties. Structures

like mollusk shells [19], arthropod cuticles [20], and bone [5] are fabricated starting at the

atomic level and built up to the macroscopic level. There are orientations within

orientations, and levels of organization within level of organization. While science

cannot yet make these materials on the same molecular size scale as nature, the

development of materials on the nanometer scale may lead to the development of

biomaterials that closely mimic mechanical structures found in nature.









Nature produces strong composites through the incorporation of polymers in the

form of proteins into composite structure with ordinarily brittle materials. Much research

is being done to develop materials, which mimic the structure and behavior of natural

materials [21]. Work has been done with conch shells [19], which are composed of

CaCO3, and proteins, have strength and toughness values that are orders of magnitude

greater than monolithic CaCO3. COnch shell achieves its superior strength, and toughness

through a hierarchical structure composed of a macroscopic three layer laminate

structure, with each layer having a second and third order hierarchy.

Bone itself is composed of a complex hierarchical structure [5] that could be

compared to a laminate with a thin cortical bone layer at the outer surface, and spongy

trabecular bone in the middle. Bone consists of collagen molecules mineralized with

hydroxyapatite crystals, which are grouped into fibrils, which are grouped into fibers,

which then form lamellar structures called osteons. Osteons align themselves parallel to

the long axis of bones. Looking at each osteon as an individual layer, bone can be

described as a laminate structure. While it is not currently plausible to build composites

with the same degree of sophistication as conch shell and bone, it is possible to

incorporate the premise of a laminate composite.

Laminate Composites

Lamination is an effective method of combining materials into composite

structures. Common ways of creating laminate structures are pressing at room or

elevated temperatures, using an adhesive to bond the lamina together or combining

materials that are mutually reactive to form chemical bonds. The resulting composites









will have laminas j oined either by a chemical bond or by a mechanical interlocking of the

lamina surfaces.

Past studies deal with combining ductile and brittle materials together through

lamination [22-24]. It has been shown that through the lamination of brittle ceramics

with ductile metals the toughness of the brittle ceramic can be increased [22, 25]. Some

systems that have been developed with laminate structures are Bioglass/Copper [22], and

Alumina/Nickel [24].

The mechanical properties of laminates are directly determined by the interface that

is formed between the laminated materials [25]. If there is a weak interface between the

constituents of the ductile-brittle laminate, then in the case of crack propagation, there

will be evidence of delamination along the interface between the two materials. The

result will be an unusable, low strength composite. A ductile-brittle laminate that has a

strong interface will have a significant increase in toughness resulting from crack

bridging or cracks arrest at the interface between the two materials.

Strength and Biocompatibility of Polysulfone

Polysulfone (PSu) is a tough, thermoplastic. PSu is used extensively for

engineering applications because it has good thermal stability with degradation

temperatures in excess of 4500C. PSu has greater strength and elastic modulus than PE,

which has facilitated research into replacing PE with PSu as the matrix material of

HA/polymer particulate composites [9].

PSu films have been used as reinforcements of carbon-fiber-reinforced epoxies [26],

as well as for filtration membranes. PSu films are made through solvent casting, or

through phase inversion techniques.









Polysulfone has been shown to be biocompatible as it is used as for blood filtration

membranes [8]. PSu has been combined with bioactive glasses for a bone Eixation study

[27] and with epoxy fibers implanted into rabbits [28]. Both studies demonstrated good

long-term stability of the PSu composites in vivo. In vitro studies of polysulfone have

show negligible cellular response to implanted polysulfone [52].

Research Rationale and Objective

The goal of this study is to develop a composite structure composed of

hydroxyapatite and polysulfone with modulus, strength, and toughness values similar to

those reported for bone. Reaching this goal will require the achievement of three

specific objectives.

The first objective is to design a fabrication technique for combining

hydroxyapatite and polysulfone into a laminate structure. Achieving this objective

requires first producing monolithic hydroxyapatite layers through a combination of tape

casting, burnout, and sintering of a starting hydroxyapatite powder, followed by

lamination with polysulfone through solvent casting, stacking, and lamination. Chapter 2

of this document covers development and optimization of a tape casting, burnout, and

sintering methodology for hydroxyapatite, as well as characterization and mechanical

testing of the monolithic HA. Chapter 3 covers procedures for fabrication of HA/PSu

laminates, as well as characterization and optimization of these laminate fabrication

procedures. Strength and toughness of the laminates will be characterized through

loading in biaxial flexure, which leads to the second obj ective.

The second obj ective of this study is to derive a mathematical model for describing

stresses resulting from applied loads during biaxial flexure of laminated circular discs.

Such a solution exists to describe the flexural behavior of laminated beams, but not for









circular discs. The derivation will be validated through a comparison with finite element

analysis of the HA/PSu laminates before being applied to calculation of laminate

strengths. Derivation of the laminate theory solution, along with a comparison to finite

element modeling is the subj ect of Chapter 4.

The final obj ective will be the characterization of the strength and toughness of the

laminates. The strength and toughness will be tested as a function of flaw sizes and

individual layer thicknesses. Fractography will be performed on fractured laminates to

determine the failure mechanisms. The strength and toughness of the laminates will be

compared with monolithic HA as well as to the reported values for bone to gauge the

success of the proj ect in terms of producing a composite with properties similar to bone.

Mechanical testing data and failure analysis of the laminates is presented in Chapter 5.















CHAPTER 2
HYDROXYAPATITE

Background on Processing of Hydroxyapatite

A number of important factors must be taken into consideration when working with

hydroxyapatite. The most important factor is the stoichiometric ratio of calcium to

phosphorus. The correct Ca:P ratio for bioactive hydroxyapatite is 1.67:1 [2, 6]. Ratios

less than 1.67 can lead to the formation of secondary calcium phosphate phases, while

ratios exceeding 1.67 can lead to CaO formation [13]. The presence of CaO and

Ca(OH)2 leads to cracking during cooling due to differences in the coefficients of thermal

contraction. There is also an accompanying decrease in mechanical strength.

Sintering temperatures for hydroxyapatite must be closely controlled because

hydroxyapatite is susceptible to undergoing a decomposition reaction at temperature,

which is only slightly higher than the sintering temperatures. Sintering temperatures for

HA range from between 1000-12500C [13]. Hydroxyapatite undergoes dehydroxylation

at ~8000C, which leads to a deficiency of OH- ions within the crystal structure, which is

remedied through rehydration during cooling. Hydroxyapatite is best sintered between

1000-12000C to achieve the highest densities, with density and porosity being highly

dependent upon starting particle sizes [13, 29, 30]. Decomposition of hydroxyapatite

occurs at a temperature between 1250-14500C and results predominantly in the formation

of tricalcium phosphate (TCP), tetracalcium phosphates (TTCP), or calcium oxide (CaO)

[13, 41]. A significant loss in strength accompanies the decomposition reaction.









Starting Materials

All hydroxyapatite specimens used in this study were fabricated from the same

commercially available hydroxylapatite (Alfa Aesar, Ward Hill, MA) starting powder.

The MSDS accompanying the hydroxylapatite gives it a molecular formula of

Calo(PO4)6(OH)2 which is the same as naturally occurring hydroxyapatite. This powder

was selected because the distributor advertises the powder as possessing the correct

1.67:1 ratio of Ca:P necessary for hydroxyapatite to be bioactive [3].



Processing Methods

Tape Casting

The starting hydroxylapatite powder was fabricated into green tape using an

Incetek tape casting machine (Integrated Ceramic Technologies, Inc., San Marcos,CA).

The composition of the tape casting slurries can be found in Figure 2. 1. Slurries were

combined into polypropylene bottles, beginning with the addition of the organic solvents.

Three organic solvents make up the 35 wt%/ of the slurry: methyl ethyl ketone (MEK),

toluene, and ethanol (EtOH). These are added in a ratio of 6.25:5.25:1. Plasticizer

(Santicizer@ S-160), dispersant (Blown Menhaden Fish Oil), organic solvents, and 10-15

alumina milling balls are combined all at once. The combination is swirled until the

dispersant is completely suspended. Hydroxylapatite powder and binder (Butvar@ B-98)

are combined apart from the solvent mixture and then gradually added. The bottle

containing all constituents was wrapped in paraffin wax to minimize solvent evaporation,

and ball milled for a minimum of 12 h.

After ball milling, the slurries are immediately tape cast. Slurries were not

filtered prior to casting because a skin forms rapidly once the slurries are removed from









the ball mill. Tape casting was performed at a speed of ~3.5 m/min, the best possible

speed of the tape caster, and at ambient conditions. The doctor blade was set at a height

of 200 Cpm above the moving tape. Cast tapes were allowed to dry inside the tape casting

machine for a period of 1-2 h to ensure complete drying, although after 15-20 min tapes

are no longer tacky, and can be handled. Dried tapes had a thickness of 100-150 pm.

After 1-2 h, the green tape is removed from the tape casting machine and processed in to

specimens for testing.

Cold Pressing

Green tapes are punched into discs using a 25mm diameter knife edge punch. 0.3

m (1 ft) of green tape yields 30-36 discs. Discs were stacked to a desired thickness and

cold pressed in a graphite mold and die at a pressure of 3500 psi (24 MPa) using a

hydraulic laboratory press (Model C, Fred S. Carver Inc., Menomonee Falls, Wis.).

Pressure was applied in increments of 1000, 1000, 1500 psi, allowing the system to come

to equilibrium after each pressure addition. Equilibrium was seen as being achieved once

the applied pressure remained constant for a minimum of 30 seconds without any

decrease in the pressure reading. Once the maximum pressure was applied, the system is

allowed to come to a final equilibrium and held constant for 1 minute to ensure that

equilibrium had been achieved.

Organic Burnout and Densification

Thermal analysis was performed on green tape samples using thermogravametric-

differental thermal analysis (TG/DTA), (Seiko Systems, Model 302) in order to identify

temperatures for removal of the organic constituents of tape casting slurries, Figure 2.3.

Organic burnout was carried out in a muffle furnace (Model FA 1730, Thermolyne,














40

35

30

25

20

15

10

5

0
HA Solvents Plasticizer Binder Dispersant

Figure 2.1 Weight % of each constituent in the tape casting slurry.


Figure 2.2 Green tape processing, shown are punched green discs (top left), the 25 mm
(1") knife edge punch (middle), and scrap tape (top right). The section of tape
processed (bottom) is 30 cm (12") x 11 cm (4.5") and produced 36 green discs.











1000



800-



600-



5 400-



200-




0 12 24 36 48
Time (Hours)

Figure 2.3 Organic burnout cycle performed on the consolidated discs produced by cold
pressing hydroxyapatite green tape.




Dubuque, IA) with a programmable digital controller. The pressed hydroxylapatite discs

were placed onto furnace plates that had been dusted with alumina powder to prevent any

possible adhesion of the hydroxylapatite specimens to the furnace plates.

The rate of heating of loC/min was constant for all heating ramps, while the rate of

furnace cooling was approximately 0.50C/min. Removal of the organic occurs during

the 3 h hold at 2000C and the 6 h hold at 4500C. The 1 h holds at 8000C and 10000C are


required for initial sintering of the hydroxyapatite in order to give the specimens enough

mechanical strength to be handled for further processing. Specimens cooled from 4500C,

and not fired through these upper two hold temperatures were too brittle to be handled.









Sintering

After organic burnout, specimens were removed from the burnout furnace and

placed in a second furnace. The second furnace was required since the burnout furnace

was not capable of reaching the desired sintering temperatures. Specimens were placed

in a drop down furnace (Del Tech Inc.). A short study was conducted to Eind the

sintering temperature that produced hydroxyapatite specimens with the best

combination of strength and processability. Seven specimen groups were sintered at

temperatures of 10000, 10500, 11000, 11500, 12000, 12500, and 13000C. The ramp rate

of 50C/min and hold time of 90 min were constant for each Biring temperature.

The mechanical properties (flexural strength, hardness, fracture toughness) of the

firing study specimens were characterized and compared for the seven sintering

temperatures. The sintering temperature that yielded the best combination of strength

and processability was 12000C. To optimize the sintering process, testing was done on

the effect of both the ramp rate and hold time of the sintering process. The effect of hold

time was tested by sintering two specimen groups at 12000C with different programmed

hold times of 1 and 10 h, and using a constant ramp rate of 50C/min. Three specimens

groups were tested for the effect of ramp rate on mechanical properties. The three ramp

rates tested were lo, 50, and 200C/min to 12000C with a constant hold time of 1 h.

The final sintering program consists of a ramp rate of 50C/min with a hold time of

90 min at 12000. This firing program was used to process all monolithic hydroxyapatite

specimens as well as the hydroxyapatite specimens later used to fabricated composites.

The thickness of specimens was measured after sintering. The total reduction in

thickness from the stacked green tapes was, on average, 55%.










Specimen Refinement

The post-sintered specimens were polished by hand to a desired testing thickness

using 45 and 15 Cpm diamond polishing wheels. In some cases specimens were indented,

these specimens were polished though the 15 Cpm finish. Specimens that were not

indented were further polished using 5, 3, and 1 Cpm alumina pastes to remove as many

large surface flaws as possible.



HA Characterization Methods

Particle Size Analysis

The particle size distribution and specific surface area of the starting

hydroxylapatite powder was characterized using a laser diffraction particle size analyzer

(Beckman Coulter, LSTM 13 320 series). Analysis was performed on the starting powder

as it was delivered by the manufacturer.

X-Ray Diffraction

X-ray diffraction (XRD) was performed using an APD 3720 automated powder

diffractometer (Phillips Electronic Instrument, Inc., Mahwah, NJ). XRD was performed

on the as delivered hydroxylapatite staring powder, post-organic burnout 10000C sample,

and post-sintering at 12000C sample. Measurements were taken for a range of 26 = 200-

700, with a step size of 0.05.

Density and Porosity

The density of the monolithic hydroxyapatite specimens was measured using the

Archimedes principle of volume displacement [3 1]. Specimens were dried at 1500C for

12 h and then weighed in air. Specimens were then placed in distilled water and placed

in a vacuum chamber for 30 min to ensure saturation. Specimens were then weighed










both saturated, msar and suspended, mszes. The density was then calculated by first

calculating the volume of water displaced, Vdis, by the suspended specimen:


Vdes sa u)(2.1)
PH20

where the density of water was assumed to be 1.0 g/cm3. The density of the HA

specimens was then calculated through:


pHA air (2.2)


The total porosity of the specimens was then calculated through:


%Porosity = porsit (2.3)
Total

where the pore volume is


V (ms, md' 2.4
Yporosity 24
PH20

and Vtotal= Vporosity + Vspecimen -

Elastic Modulus and Poisson's Ratio

The elastic modulus of the hydroxyapatite specimens was characterized using an

ultrasonic technique [32]. The elastic modulus, E, was calculated from the equation:


E = (2.5)


where p is the specimen density, V, is the longitudinal wave velocity, and Vs is the shear

wave velocity. Poisson's ratio, v, was then calculated from the equation [32]:


v = 1 (2.6)
2G;

where the shear modulus, G, is calculated from the equation [32]:










G = pVs (2.7)

Optical Microscopy and Lighting Effects

The maj ority of the fractography and other microscopic analysis were performed

using optical microscopy on an Olympus optical microscope. This microscopy proved

difficult at times due to the nature of the hydroxyapatite being analyzed. Hydroxyapatite

is a white, polycrystalline material that reflects and absorbs light. The color of this

material made it particularly difficult to see surface features necessary for fractographic

measurements. Two methods were used to overcome these difficulties. The first was to

sputter coat the specimens with Au-Pd alloy. This method was used primarily on the

most difficult specimens since it requires both time and equipment that can be expensive

to use. The second solution was to switch from using overhead lighting to oblique

lighting. The oblique lighting caused surface features to stand out (Figure 2.4), and made

analysis possible in a more cost effective manner.

Scanning Electron Microscopy

Scanning Electron Microscopy (SEM) was used to characterize the microstructure

of the hydroxyapatite, and to perform fractography on the fractured specimens. All SEM

work was performed on a Joel 6400 scanning electron microscope.



Mechanical Testing Methods

Establishing Baseline Properties for Monolithic Hydroxyapatite

Strength tests were performed to establish the baseline properties of monolithic HA

to compare with composite materials. Specimens were tested using a method of strength

indentation in which specimens were first indented for hardness characterization, and

then fractured in biaxial loading.







































Figure 2.4 Oblique allowed surface features to be seen (bottom) in greater detail than
with overhead lighting (top).


Indentation

Monolithic hydroxyapatite specimens were indented with Vickers indentations.

Indentation loads < 2 kg were applied using a Micromet3 microhardness tester (Buehler

LTD, Lake Bluff, IL), loads > 2 kg were applied using an older indentation tester (Zwick

Inc.) The hardness of the specimens was calculated using the equation:


H = 0.0018544d (2.8)


where P is the indent load in N, and d is the half diagonal length of the indent in mm [33].









Biaxial Flexure

The monolithic hydroxyapatite specimens were loaded to failure in biaxial flexure

using a piston on three-ball testing fixture. A loading piston with a diameter of2.2 mm

was selected for this study, and an effective ring radius produced by the three-ball

support varied depending on the specimens being tested. Hydroxyapatite specimens fired

to 10000C were tested with a support diameter of 23.2 mm. Specimens fired to 12000C

had a small diameter due to shrinkage compared with the 10000C samples and were

instead tested with a support diameter of 15 mm.

The failure stress, of, from bending in a piston and three-ball fixture can be

calculated using the equation developed by Wachtman [34]:


3P(1- v) a v b/ 2 2
cy = 2 1+ 21n -+I 1- 2a 2 (2.9)
4xit2 b 1+v 2a R2


Where P is the failure load, v is Poisson's ratio, t is the specimen thickness, a is the

support ring radius, b is the loading piston radius, and R is the specimen radius.

Monolithic specimens were tested using a tensile testing machine (Instron) at a

loading rate of 0.25 mm/min.

Work of Fracture and Toughness

The toughness, or absorbed energy at fracture, of the monolithic specimens was

calculated from the area under the stress-strain curves generated during loading in biaxial

flexure [35]:


T = crd (2.10)

the stress, o, was converted from the measured loads using equation 2.9, and the strain, 8,









was calculated from the equation for elastic modulus, E=0/E. The units for the toughness

are J/m3.

The work of fracture was calculated from the area under the load-displacement curves

using the equation [36, 37]:

IPd6
Twy = 2A(2.11)


where P is the fracture load, 6 is the measured crosshead displacement, and A is the

proj ected cross-sectional area of the created fracture surface. Work of fracture has units

of J/m2. Even though circular discs were fractured in biaxial flexure the cross-section of

the fracture surface was assumed to be rectangular. The area of the fracture surface was

dependent on the number of pieces created during the fracture process. Proj ected fracture

surface areas were calculated by measuring the lengths of fractures and multiplied by the

specimen thickness.

Fracture Toughness

The fracture toughness, Kc, for the monolithic specimens was calculated using

three different techniques. The first technique involved using the fracture mechanics

equation [38]:

Kc = Yo/Crcl2 (2.12)

where of is the failure stress calculated from equation 2.9, c is the critical flaw size, and Y

is a geometric constant equal to 1.65 for indented specimens and 1.24 for nonindented

specimens. The critical flaw size was measured using fractography and

c =(ab)1/ (2.13)

where a is the length of the semi-minor axis of the critical flaw, and 2b is the length of









the semi-maj or axis, see Figure 2.5. Figure 2.5 is a schematic representation of fracture

markings resulting from a brittle fracture process and is taken from Mecholsky, et al [53].

A detailed explanation of the different fracture markings can be found in this reference.

Since HA is a polycrystalline ceramic it is difficult to distinguish many of these features.

For this reason, only the size of the critical flaw was measured for fracture toughness

calculations. The fracture markings that were measured are designated acr and ber in

Figure 2.5.













Figure 2. ceai ersnaino rcur aknsta eutfo rtl
fracture~= adaptedfro ecosye l 5]

Thesecnd ethd fr alclatng racuretogheswstruhhetenh

indenttion echniqe [39. Ti ehiu acuae h rcue ogns sn h
equation

Ks; -4 o-Py 33 4 2.14









where E is the elastic modulus of the specimen, H is the hardness, of is the failure stress,

P is the failure load, and rl,Ris a geometrical constant equal to 0.59 +/- 0. 12.

Anstis [40] demonstrated that the fracture toughness values Kc and Ksi should yield

comparable results for brittle materials. The strength indentation technique was used to

verify calculations using the fracture mechanics equation, which relies on measuring the

critical flaw size, which in some cases proved difficult due to the fracture behavior of the

hydroxyapatite and complexity of the observed cracks.

The strength indentation technique was also used to ensure that the fracture

toughness was constant across a range of indent loads, and thus for different flaw sizes.

A log-log graph of failure stress versus indent load should yield a straight line with a

slope of -1/3. Deviation from the -1/3 slope would demonstrate that fracture toughness is

not constant for increasing flaw sizes and that phenomena such as microstructural effects

or R-curve behavior was occurring. However, as will be demonstrated later in this

chapter, fracture toughness values for HA were constant with increasing flaw size.

The third method used to calculate the fracture toughness from the work of

fracture and elastic modulus using the relation [38]:


K,,, =2E)',g(2.15)

Finite Element Analysis

Finite element analysis was also performed to analyze the monolithic HA

specimens. A detailed explanation of the analysis and the results are discussed in the

Chapter 4.








Statistical Analysis
All statistical analysis was performed using DOE Fusion Pro software. The
software is running ANOVA statistics. All statistical analysis was run at a = 0.05.


Results and Discussion
Starting Powder Characterization
The particle size distribution for the staring hydroxylapatite powder is shown in
Figure 2.6. The starting hydroxylapatite powder shows a bimodal distribution. The use
of a starting hydroxyapatite powder with a bimodal distribution has been reported

previously [51] with the bimodal distribution attributed to the presence of large
agglomerates. The starting powder was used in its as-received state and with no
refinement to the particle size distribution. Since the goal of this study was to create a



4.

3.5



12.5

1.

0.5
0 a-.e
0 5 10 15 20 25 30 35 40 45 50
Particle Diameter (microns)
Figure 2.6 Particle size distribution for the starting hydroxylapatite powder










porous material that could allow cell infiltration, the particle size was left large in order to

prevent packing during the processing steps. If the starting powder was refined to a

smaller, more uniform particle size distribution the net result would be denser

hydroxyapatite specimens. While the denser material would have increased flexural

strength and toughness, the ideal result for this study would demonstrate that

hydroxyapatite with a large amount of porosity can be fabricated to have mechanical

properties within the range of bone.

XRD spectrum for the starting powder is shown in Figure 2.7. The spectrum shows a

hydroxylapatite starting powder with low crystallinity, and is composed of mostly tri-

calcium and tetra-calcium phosphate secondary phases.


Figure 2.7 XRD spectrum for the as-received hydroxylapatite powder









Development of a Tape Casting Method for Hydroxylapatite

Due to the large particle size distribution of the starting powder, and the type of

tape casting machine being utilized developing a tape casting process for the

hydroxylapatite proved a long and tedious trial and error process. The Incetek Model 104

tape caster has one exhaust port through which evaporating solvents are evacuated to a

fume hood and a number of open seams around its doors and edges. These openings

make it difficult to control airflow through the tape caster and thus the drying rate of the

tape. Initial attempts were made to fabricate slurries with ethanol as the only organic

solvent. However, the lower boiling point of ethanol (~780C) leads to a rapid

evaporation rate and thus cracking of the green tape. To slow the evaporation rate,

toluene was substituted for the ethanol since it has a higher boiling point of 1100C. The

binder, B-98, is only partially soluble in toluene, which led to undissolved binder in the

tape casting slurries and large tape defects upon drying. To remedy this problem IVEK

was added to aid in dissolving the binder. It has a boiling temperature (800C) that is

slightly higher than EtOH, which would also slow the drying process. The final slurry

composition, which contains all three organic solvents, is a mixture which balances the

need to dissolve the binder completely with evaporation rates that allow for drying of the

green tape free of cracking due to overly rapid drying.

Once the drying problems were overcome, tapes were cast and consolidated

through cold pressing. The two factors, which affected the cold pressing process, were

applied pressure and hold time. A number of applied pressures (with a constant hold

time) and hold times (with a constant applied pressure) were tested for their effect on the

overall consolidation of the stacked green tapes during, Figure 2.8. Consolidation was










quantified by the measuring the percent reduction of total thickness of the stacked HA

discs was reduced during cold pressing.

The consolidation of the stacked green tapes occurred through plastic deformation

of the polymeric binder. This deformation determines the proximity of HA particles to

each other. The organic tape casting constituents are removed during burnout leaving

behind only the HA particles. With other factors such as phase structure, particle

composition, wetting, and heating programs being constant, the particle spacing

following cold pressing will control the Einal density of the HA specimens after sintering.

As shown in Figure 2.8, it was found that the reduction in thickness of the stacked green

tapes was not dependent upon either the applied pressure or the hold time. These results

indicated that deformation of the binder occurs almost instantaneously upon the applied

pressure reaching the threshold at which deformation occurs. The idea of the binder

undergoing instantaneous deformation was supported by the equilibration of the applied

pressure within a few seconds of the Einal pressure being applied and remaining constant

for the duration of the hold time. Since the total amount of force applied by the press

remains constant unless acted upon by an outside agent (i.e. additional ram strokes), a

drop in the pressure of the system would indicate an increase in the area through plastic

deformation. This deformation would result in an increase in the diameter of the discs.

However, since the pressure comes to equilibrium after only a few seconds and remains

constant all deformation must also occur during the initial seconds of applied pressure.

The fact that the reduction in thickness remains constant for the pressure tested indicates

that the threshold for deforming the polymer binder lies below the lowest pressure tested.









During the early stages of this study, a large amount of binder was added to the

tape casting slurries. Some of the first experimental slurries contained as much as 25-30

wt% binder. The strategy was to prevent cracking due to drying stresses through the

addition of large quantities of binder. This represented a "brute force" attempt for

overcoming the formation of cracks in the dried tapes by having more of the ductile

binder material in the system to absorb the drying stresses and effectively toughen the

green tapes. The large amount of binder was successful in preventing cracks; however, it

caused the manifestation of even larger defects during burnout and sintering. The larger

amount of binder increased the distance between HA particles in the dried tapes resulting

in large defects during burnout and sintering. Blowholes, delamination (Fig 2.9), and

large cracks resulted from particles not being in close enough proximity to bond during

the firing processes. The survival rate of these early trials was at best 40-45%. Some

experimental tapes had less than 10% of specimens survive the burnout process. Through

numerous trials with experimental slurries and burnout processes it was deduced that

poor survivability of specimens was due entirely to the HA particle separation being too

great due to the large amount of binder. The solution to this problem was to reduce the

HA particle separation thorough increased deformation of the binder during

consolidation. To do this the starting slurry composition was altered to include more

plasticizer, making the binder more ductile, and allowing for greater deformation of the

binder during cold pressing. It was also discovered that with the binder in a more

plasticized form, less of it was required to overcome drying stresses. The effect of both

the binder:plasticizer ratio and total amount of binder were tested to identify the

combination that provided the largest reduction in thickness, and thus the most












50




45
S5-

S2-







S5
a


O 4-
3000


3500


4000O 4;500
Pressure (pal)


5000


5500


40





25








Q 5-
r







0 t 12 3 4 5 6
Time (mlnutes)

Figure 2.8 Effects of applied pressure (top) and hold time (bottom) on the reduction in
thickness of stacked green tapes during cold pressing.

































Figure 2.9 Optical micrographs of post-cold pressing edge delamination (top left),
delamination at the center (bottom left), full consolidation (top right), and post-
sintering center delamination.


consolidation of the green tapes. The results of these tests, Figure 2. 10, led to a

decreased amount of binder within the tape casting slurries, and an increase in the amount

of plasticizer to a level equal to a binder:plasticizer ratio of 1:3 by weight. The results

were tapes with increased ductility, and greater deformation during cold pressing due to

the now highly plasticized nature of the binder, which decreased the separation between

HA particles to a level, which allowed for better densifieation of the specimens during

Bring. The net result was a survival rate of greater than 95% for the organic burnout

firing process.

Reduction in thicknesses were also measured before and after firing to see how

great an effect firing had on the consolidation of the stacked green tapes. The reduction









in thicknesses during both cold pressing and firing were then combined to quantify the

total reduction in thickness from the starting thickness of the stacked green tapes to the

post-fired thickness of the specimens. The total thickness reduction could then be used as

a design tool to estimate the necessary amount of stacked green tapes required to achieve

a desired post-Hiring specimen thickness.

Two other aspects should be noted for the tape casting method developed here. First, the

slurry composition in its Einal form, as described in the methods section of this chapter,

can be used for any desired slurry volume. The slurry composition shown in Figure 2. 1

can be used in small volumes for tape < 1 m, moderate volumes for 3-5 m of tape, which

is the limit of the research laboratory, or theoretically for large volumes necessary for

mass production.

Second, the addition of all slurry components prior to milling is somewhat

unconventional. Typically, slurries would be produced by first completely dissolving

binder in solvent, and then adding the plasticizers. Separately, the ceramic powder

would be suspended in solvent, dispersant would be added, and the powder-dispersant

system would be milled to allow the dispersant to disperse the ceramic particles. Finally,

the binder solution would be added, and the entire system milled to allow the binder to

inHiltrate between the dispersant-coated particles. The binder would be bonded to the

dispersant, and this is how the tape would be held together. This is the more

conventional stepwise approach to slurry formation. However, in all the attempts made

in the early portion of this study to utilize the conventional slurry formation process, a

usable tape was never processed. The lack of success with the conventional process

facilitated the development of the less conventional "all at once" approach applied here.











50

45







15



; 0






5 5 7 8 9 10 11 12 13
WVX~Islf Bind~e~r inCasng Slurry %)I



50

45







S25 + kage

g 20 C -m--PresngS~r~y
~i Is ~Flnnkagenr~
ce Shantagec- -- --






0,2 0.3 0 4 0,5 0.5 0.7 0.8 0;9
BInder:Plastici~:zer Rado

Figure 2. 10 The reduction in thickness of the stacked green tapes is plotted versus wt%
binder in the tape casting slurries (top), and binder:plasticizer ratio (bottom).









Burnout and Sintering Process

The burnout and sintering process performed on the consolidated green tapes was

determined after performing TG/DTA on samples of the green tape. TG/DTA was also

performed on specimens following the burnout process to verify that all organic, tape

casting materials were successfully removed. The resulting curve, Figure 2. 11, showed

that all the organic constituents burned out at temperatures at or below 4500C. The green

tape TG/DTA curve shows a large volume of material being lost between 200-2500C.

This represents the removal of binder and plasticizer as the binder has a glass transition

temperature in the range of 140-2000C and the plasticizer has a boiling point of 2400C

according to the material safety data sheets (MSDS) provided by the manufacturer. The

remaining dispersant was removed at a temperature of 2800C.

The burnout program derived from the TG/DTA data contained two holds. The

first 2 h hold at 2500C was added to slow the burnout process and allow degradation

products to diffuse out of the specimens. This first hold was essential to the survival of

specimens during firing. Without it, the degradation products would build up too rapidly

within the specimens building up internal pressure sufficient to cause blowholes to form,

thus destroying the specimens. The additional 6 h hold at 4500 ensures sufficient time for

organic removal from the HA specimens.

The TG/DTA curve for the post-burnout sample shows that the burnout process

successfully removes the organic tape casting components. The small decrease in weight

can be attributed to the loss of water bound within the HA crystal structure that was

released at elevated temperatures [13].









Characterization and Mechanical Properties of Post-Burnout Hydroxyapatite

Following the burnout and partial sintering process used to remove the organic

tape casting additives, samples were characterized using XRD to show the effect of the

firing program on the crystallinity and existing phases in comparison to the stating

powder, Figures 2.12-2.13. The XRD spectrum for the post-burnout HA shows that

firing the starting powder to 10000C increases the degree of crystallinity and causes the

transformation of the numerous secondary phases into a more homogenous material that

is very similar to XRD spectra for naturally occurring form of HA [8], Figure 2. 13. The

only difference between the 10000C HA spectrum and naturally occurring HA spectra is

a small peak at a 26 value between 370 and 380. This peak is not found in the XRD

spectrum for the starting powder and is therefore formed during the firing processes.

This peak represents the formation of a CaO during the firing process, and this is

supported by a similar identification of the peak by another group [41].

Additional XRD was performed on specimens sintered at 12000C. The XRD

spectrums for these specimens sintered at 12000C closely resemble the spectrums for

thel10000C specimens with the only difference being the intensity of a few of the peeks.

The result shows that the second firing of the specimens does not cause any phase

transformations that would result in the existence of secondary calcium phosphate phases.

The result of the firing processes as explained through the XRD characterization is the

conversion of a starting powder that contains secondary calcium phosphates into a more

homogenous hydroxyapatite phase that closely resembles naturally occurring

hydroxyapatite and is stable at the sintering temperatures used in this study.

















E1



70 -



50
0 0 0 00 40 50 60 0 0 0

Teprtr C


100rr 200 300 400 500 600 700 800 000
Temperature (C


Figure 2. 11 TG/DTA data curves for a pre-burnout green tape sample (top) and a post-
burnout hydroxyapatite specimen.


E




e~8-









Problems with Specimens Fired to 10000C

Initially it was proposed that the hydroxyapatite specimens would undergo only

one firing process effectively minimizing the total energy required to produce testable

specimens. A single firing at a relatively low temperature of 10000C would ensure a

large amount of porosity within the system, which would be necessary for future use in

biomedical applications. However, it became clear very quickly that a second sintering

process was a necessity for specimen fabrication. Less than 10% of the burned out

specimens survived the necessary polishing and refinement in order to be used for

mechanical testing. Of the 24 samples that could be burned out at one time, only 1-2

samples usually survived to reach mechanical testing, the rest fractured during polishing,

most often at the outset. Thus it was determined that in order to conserve starting

materials, and produce the number of specimens necessary for completion of this study

that the one firing route would have to be abandoned for a two firing system that would

yield specimens with greater mechanical stability. Therefore, a short firing study was

performed to establish the best temperature that would yield the most usable specimens.

"Usable" was defined as having the greatest percentage of specimens survive the

polishing process.

Mechanical Properties of 10000C Specimens

A total of 29 specimens were tested after burnout at 10000C. These specimens

were polished to a thickness of2 mm, and indented prior to loading in biaxial flexure.

Indentation controlled flaw sizes were produced in order to lessen the number of

specimens required for a statistical analysis. The six indent loads tested and the resulting

mechanical properties are reported in Table 2.1. The diagonals of the Vickers












4900
Counts]
3600


2588






988-










13 20 38 408 58 68 E' 201 738
HCoteeeC .3o unteeeC.Snt






[counts 3-
3600


2500


1600


900










13 28 38 48 58 60 ['201 76
HA120BC .RD HCI2ZBBC,8M

Figure 2.12 XRD spectrum for HA specimens burned out at 10000C (top) and specimens
sintered at 12000C (bottom)


























































I



r


['80


I


I I ~ I I ______ ~


2 Theta (degrees)

Figure 2.13 XRD spectrum for HA specimen fired to 12000C (top) and for naturally
occurring HA (bottom) [8]. Circles (*) indicate peaks found on both spectra, the
square (m) indicates the peak that is unique to the 12000C spectrum.


t8-

[corunts]
3600


2500
1B-


98-




488


**


Safrr ?
16
HRIti2BC .RD






20000


29 39 4858


- -


*i


."*


=3
;ZIp

o
Ea


r
a,
r


15000




1000




5000


10


30


40










impressions (Figure 2. 14) were measured for hardness calculations prior to biaxial

loading, with the hardness calculated using Equation 2.8. A statistical analysis (a = 0.05)

showed that there was no significant difference in the hardness values for the six tested

indent loads.

Stress at failure was calculated from the experimentally measured failure loads and

Equation 2.9. A statistical analysis (a = 0.05) showed that failure stresses were

statistically different for the range of indent loads used, i.e. a decrease in failure stress

with increasing indent load, as expected. Since the fracture toughness is constant for all

specimens, a larger indent load results in a larger initial flaw and consequently lower

failure stress.

Following failure in biaxial loading, fractures were examined to ensure that cracks

originated from the indent flaw Figure 2.14, and these flaw sizes were measured using

fractography. Fracture toughness was calculated using both the strength indentation

method, Equation 2.14, and direct flaw size measurement, Equation 2.12. These two

values should coincide with each other, but do not for the 10000C samples. The reason

for the difference shown in Table 2. 1 is flaw size measurements for the 10000C

specimens were difficult due to the very low toughness of the material. Localized

crushing around the indents made flaw size measurements potentially inaccurate. An

80% correlation will be shown between Kc and Ksi for the 12000C specimens later in this

chapter.

Constant Fracture Toughness

Figure 2.15 is a log-log plot of the failure stress, of, versus indent load P. As

discussed in the materials and methods section, if the slope of such a plot is equal to -0.33

then the fracture toughness is constant for the range of indent loads tested. This is










important because deviation from this slope indicates that other factors must be taken into

consideration when calculating the fracture toughness. A best-fit line through the six data

points, factoring in standard deviation not shown on Figure 2. 15, had a slope of -0.35.



Table 2. 1 Mechanical data and fractography measurements for HA specimens fired to
1000oC
Initial
Indent Failure Failure
Hardness Flaw Kc Ksi
Load n Load Stress 1a-'2 La- '2)
(GPa) Size (Mam ) ( a- )
(kgf) (N) (MPa)

0.16 31 10.2 194 0.23 0.27
0.50 5
S0.03 & 6 & 0.9 & 46 & 0.04 & 0.02
0.13 27 8.2 234 0.20 0.28
1.00 5
S0.02 & 3 & 0.9 & 40 + 0.03 & 0.02
0.18 24 7.5 284 0.02 0.31
2.00 6
S0.03 & 3 A 37 & 0.01 & 0.02
0.15 21 6.6 379 0.21 0.32
3.35 5
S0.02 & 2 & 0.7 & 145 & 0.04 & 0.02
0.15 21 6.7 398 0.22 0.34
4.35 6
S0.02 & 2 & 0.7 & 43 & 0.03 & 0.03
0.16 23 7.1 366 0.22 0.41
7.35 2
S0.01 + 0.2 & 42 & 0.01 & 0.01


Figure 2. 14 Optical micrographs of a Vickers indent for a 10000C specimen (left) and
crack propagation through a Vickers indent.










100












S10




0.1 1 10
Indent Load (kg)

Figure 2. 15 Log-log plot of failure stress versus indent load for HA specimens fired to
10000C




Firing Study

In order to improve the mechanical strength of the monolithic hydroxyapatite the

specimens burned out at 10000C were then sintered at a higher temperature. A range of

elevated sintering temperatures was tested to see their effects of the mechanical strength,

density, and hardness. The selected sintering temperatures included 1050, 1100, 1150,

1200, 1250, 13000C. As with the 10000C treatment, all specimens were indented prior to

loading in biaxial flexure, and for comparison purposes all specimens were indented with

a 4.35 kg load. The 4.35 kg indent load was selected to ensure that fracture occurred

from the indentation flaw. Six specimens were tested at each temperature except for

13000C. Only three of the specimens sintered at 13000C survived the fabrication

processes. Most specimens fractured under the loads applied during polishing.









Sintering Temperature Effect on Hardness

The diagonals of Vickers indent impression were measured and the hardness

calculated from Equation 2.8. Hardness values increased with increasing sintering

temperature, Figure 2.16. The increase in hardness values associated with increases in

sintering temperature has been shown to be the result of larger grains sizes [42].

Sintering Temperature Effect on Biaxial Flexure Strength

The indented specimens were loaded in biaxial flexure and the failure load was

recorded and used to calculate the failure stress of the specimens using Equation 2.9. The

failure stress increased over the sintering temperature range from 10000C up to 12000C,

as shown in Figure 2. 17. The failure stress then remains statistically constant for a

sintering temperature of 12500C. The 13000C specimens had an increased failure stress,

however since only three specimens were tested it is difficult to identify the increase as a

real occurrence or just as an anomaly resulting from the three strongest specimens

surviving the fabrication process.

Sintering Temperature Effect on Density

The density of specimens fired at increasing sintering temperatures was measured

using the Archimedes method described in the materials and methods section. The

density of the hydroxyapatite increased with increasing sintering temperature up to a peak

at 12000C, Figure 2.18, and then it decreased for the elevated firing temperatures. The

decrease in density is a direct result of decomposition reactions, which lead to the

formation of less dense calcium phosphates, calcium oxides, or oxyapatites and water as

byproducts [13, 41].


















































35


30-


25-


20-





10-


5-


0
950 1000 1050 1100 1150 1200 1250 1300 1350
Firing Temperature (oC)

Figure 2.17 A plot of failure stress versus sintering temperature for the
hydroxyapatite specimens.


5i

4-
















950 1000 1050 1100 1150 1200 1250 1300 1350

Temperature (oC)

Figure 2. 16 A plot of hardness versus sintering temperature for the
hydroxyapatite specimens.











3.00



2.75-








2.25-



2.00
1000 1050 1100 1150 1200
Temperature (C)


1250 1300 1350


Figure 2. 18 Plot of density versus sintering temperature for hydroxyapatite specimens.




Qualitative Determination of Hydroxyapatite Decomposition

Hydroxyapatite has been shown to decompose into anhydrous calcium phosphates

[13] at sintering temperatures between 1200-14500C depending on the characteristics of

the starting HA powder. XRD analysis of specimens sintered at 12000C showed that

decomposition had not occurred up to this temperature. However, a color change, most

likely the result of some decomposition, starts to be seen at sintering temperatures at or

above 12000C. A side-by-side comparison, Figure 2.19, of specimens showed that at

12000C specimens begin to take on a greenish hue, which darkens at 12500C. Specimens

fired at 13000C, not pictured, take on a purple hue, indicating a further decomposition of

hydroxyapatite into calcium phosphates.



























Figure 2. 19 Color comparison of specimens sintered at five different temperatures. Each
row of specimens was fired at the temperature indicated to the right.


Sintering Temperature Selection

After analyzing the mechanical properties of hydroxyapatite specimens, it was

determined, that the hydroxyapatite specimen used for composite formation would be

sintered at a temperature of 12000C. There were three reasons for this selection. The

most important was that XRD performed at 12000C showed that hydroxyapatite, in a

form closely resembling natural HA, had not undergone any decomposition significant

enough to be show up in the spectrum. The XRD analysis showed that the

hydroxyapatite was stable through both the organic burnout process and the additional

sintering firing at 12000C.

The second reason for selecting 12000C as the sintering temperature was that it

provided greatest reproducible failure stress. While potentially the 13000C firing

temperature would produce stronger specimens, the inability to fabricate intact specimens

made using this temperature unfeasible. It would be possible to identify the reason for









the brittle nature of the 13000C specimens, but it was determined that such an analysis

would be costly in terms of time and materials.

The third reason for selecting the 12000C is that it represents the lowest

temperature that could produce specimens in bulk quantities. Being the lowest

temperature, sintering would require less energy and consequently less heating and

cooling time to produce the number of specimens required for this study.

The only draw back of the 12000C specimens is their density being a relative

maximum for the sintered specimens. The total porosity estimated from the Archimedes

method calculations was 3 5-40% of the total volume. While the proj ect rationale for this

proj ect called for producing HA specimens with the greatest possible pore volume, as this

is desirable for biomedical application, using more dense specimens sintered at 12000C

was necessary for increased processing ability.

Optimization of Sintering

Once 12000C was established as the sintering temperature for this study, the

sintering process was optimized to reduce processing time. The mechanical properties

were tested as a function of ramping rates, and hold times. Specimens were fabricated,

indented, and loaded in biaxial flexure. The failure stress was then compared for two

hold times, Figure 2.20, and three ramping rates, Figure 2.21. The results showed that the

failure stress was constant for holds times of 1 and 20 h. This was confirmed by a

statistical analysis (a = 0.05) that showed that there was not benefit to increased hold

times. It was assumed that testing of additional specimens at hold times within in the

range between 1 h and 20 h would produce strength values consistent with the 1 h, and 20

h hold strengths. Therefore, no further testing was performed.









Failure stress was shown to decrease with increasing ramping rates. This is

mostly likely the result of an inability of the Deltech furnace to keep up with the higher

heating rates and consequently the temperature reached was lower than the desired

12000C. It was determined that 50C/min was the fastest ramping rate that the Deltech

furnace could handle and still ensure reaching the desired sintering temperature at the

programmed time.

Density and Elastic Modulus of 12000C Specimens

The density of the 12000C specimens calculated from the Archimedes method was found

to be 2.85 + 0.06 g/cm3. The theoretical density of hydroxyapatite is between 3.14 3.20

g/cm3. The calculated density is 10% less than the theoretical density value indicated the

existence of significant porosity. The porosity of the 12000C HA was calculated to be

30-35%. The strength values for the specimens sintered at 12000C are consistent with

hydroxyapatite specimens verified to have 30% porosity through helium pycnometry by

another study [13]. It is assumed therefore that the HA specimens produced for this study

possess 30% porosity, and that the accuracy of the density measurement is limited due to

the low precision of the Archimedes method. SEM imaging of the HA microstructure,

Figure 2.22, show that the 30% porosity measurement is plausible. Mercury porosimetry

was attempted in order to better quantify the density and porosity of HA specimens,

however at the time of writing no data was available.

The elastic modulus of 10 hydroxyapatite specimens sintered at 12000C was

measured using an ultrasonic technique. The specimens were randomly selected from

different sintering batches to gain a better statistical average for later us in laminate

composites calculations. The elastic modulus would be assumed constant for all of






50









25







~15











0 2 4 6 8 10 12
Hold Time (hours)

Figure 2.20 Plot of failure stress versus hold times at 12000C


30


25-






S15


S10







0 5 10 15 20 25

Ramp Rate (OC)1min)

Figure 2.21 Plot of failure stress versus ramp rate sintering temperature of 12000C.










the HA used in composite fabrication. The elastic modulus was measured at 56 & 3 GPa,

with the lowest measured value being 52 GPa. The 52 GPa value will be used for all

composite calculations as it represents the minimum possible modulus for the HA

specimens developed.
























Figure 2.22 SEM micrograph of the microstructure of the hydroxyapatite specimens
sintered at 12000C.



Hardness of 12000C Specimens

The hardness of the 12000C specimens was measured for the firing study, Figure

2. 16. The hardness was measured at 1.4 & 0.7 GPa for the size of the Vickers indent

impression. This value represents an approximately 1000% increase from the hardness

of the 10000C specimens. This increase in hardness as a result of increased sintering

temperature has been well established for HA [30, 41, 43].










Biaxial Flexure Strength of 12000C Specimens

The mechanical properties of specimens sintered at 12000C were characterized in

the same manner as the 10000C specimens. Groups of six specimens were indented and

loaded in biaxial flexure to fracture. Four groups of 12000C specimens were tested,

three of these groups were indented with 3.35 kg loads, and one group was tested without

indents. The three indented groups were polished to three different thicknesses; these

were 1.5, 2.0, and 2.2 mm. Three thicknesses were tested for a failure load comparison

with composites having a similar range of thicknesses. The non-indented sample group

tested the inherent strength of the hydroxyapatite. Failure stresses of the 12000C

specimens, Table 2.2, showed an increase of approximately 200% over the values

measured for 10000C specimens with the same indent load (Table 2.1i).



Table 2.2 Biaxial flexure strength data for specimens sintered at 12000C.
Indent Load Thickness Failure Load Failure Stress
Grou n (g) mm) (N) (MPa)
1 6 3.35 1.5 31 &3 19 &2
2 6 3.35 2.0 58 & 6 22 & 2
3 6 3.35 2.2 70 & 11 24 & 4
4 6 N-I 2.0 89 & 19 39 & 8



Figure 2.23 is a comparison of stress-strain curves for an indented specimen

versus a non-indented specimen. The comparison shows approximately constant elastic

modulus as the curves have identical initial slopes, with the elastic modulus being 64

GPa. This is greater than the calculated value from ultrasound, but well within the

acceptable range for hydroxyapatite.




































60


50


40 Non-Indented
a. -3.35kg Indent
30-


20




10


0 0.0005 0.001 0.0015 0.002
Strain (mm/mm)

Figure 2.23 Comparison of stress-strain curves for indented and non-indented HA
specimens.


Work of Fracture and Toughness

Work of fracture values were calculated using Equation 2. 11, and are reported in

Table 2.3. Specimens fractured into either two or three pieces as shown in Figure 2.24.

The length of all cracks was measured from fracture surface. All work of fracture

integration and toughness calculations were performed in Microsoft Excel spreadsheets.

Work of fracture and toughness values were not calculated for Group 1 because most of

the loading data was lost with only the greatest measured load and loading curves being

saved on paper. Calculating the area under load-displacement curves by hand would

have resulted in inaccurate data being included in the study.

























Table 2.3 Work of fracture values for specimens sintered at 12000C
Work of Absorbed
Indent Load
Group n Fracture Energy at Failure
(kg) (kJ/m2) (kJ/m3)
2 6 3.35 0.009 & 0.002 5 & 2
3 6 3.35 0.011 & 0.002 5 A 1
4 6 N-I 0.020 + 0.006 17 & 8



Fracture Toughness of 12000C Specimens

Fracture toughness values were calculated using the three methods described in the

Methods section of this chapter. The facture toughness values measured using all three

methods are reported in Table 2.4.

The measured values for the fracture toughness calculated from the size of the critical

flaw, Kc, and from strength indentation, KsI, are in reasonable agreement. It has been

demonstrated that these two values should be equal, and a plot of these values against one

another should yield a line with a slope of 1.0. A plot of the fracture toughness values

measured in this study, Figure 2.25, produces a best-fit line with a slope of 0.8. The

discrepancy most likely arises from the direct measurement of the flaw sizes

and Equation 2.12 which assumes a particular flaw geometry resulting from indentation


Figure 2.24 All specimens tested in biaxial flexure fractured into either two (left) or three
(right) pieces.









Table 2.4 Fracture toughness values for 12000C hydroxyapatite specimens
Initial
Indent Loads Flaw Size Kc Ksi Kwof

(kg) (Cpm) (MPa-ml 2) (MPa-ml 2) (MPa-ml 2)
3.35 313 &39 0.59 0.66 1.0
Non-Indented 98 & 43 0.63 N/A 1.5



yielding a constant value of 1.65 for the geometric constant, Y. Fracture surfaces

generated for the HA were often complex due to the porous nature of the HA

microstructure and loading in biaxial flexure, and thus there is some experimental error

involved in the fracture toughness calculation. This error accounts for the discrepancy

between the two values for fracture toughness. Critical flaws for two of the 12000C

specimens are shown in Figures 2.26 2.27.

The fracture toughness value calculated from the work of fracture, KH'oF, iS

statistically greater (a = 0.05) than the other two values due to an underestimation of the

fracture surface area generated during the fractures process. The proj ected area of the

fracture surface was used for calculation of KH'oF, however an area of the fracture surface

calculated on smaller length scales is required for an exact calculation of fracture

toughness. Because of the area estimation, KH'oF Should be viewed as an apparent

toughness value and not a direct measurement of fracture toughness.

Monolithic Hydroxyapatite Specimens

The focus of this study is on the ability of a composite structure to enhance

mechanical properties. To this end, the role of the monolithic testing was to establish

baseline mechanical properties for comparison to composite properties. The mechanical

properties of the monolithic hydroxyapatite produced here were considered acceptable

regardless of how they compared to previous work done with hydroxyapatite. The firing






56




0.9o







0.70 -1 I y=0.8x+0.2





S0.50





0.40
0.40 0.50 0.60 0.70 0.80 0.90
Fractrue Toughness, Kce (MlPs-m )'

Figure 2.25 Plot of Kc versus Ksi for the hydroxyapatite specimens sintered at 12000C.


Figure 2.26 SEM micrograph of the initial flaw caused by an indent (black bar) for a
12000C HA specimen. The outer boundary of the initial flaw (white arrows) and a
twist-hackle marking (black arrow) are shown.
















































Figure 2.27 Optical micrographs of an entire fracture surface (top) and initial flaw
(bottom) of a 12000C specimen indented with a 3.35 kg indent. The white arrows
indicate the outer edge of the critical flaw.


aOOym

r~iJ~










study and analysis performed to ensure that hydroxyapatite specimens could be produced

in sufficient quantities for composite formation and that fabrication would be both

efficient and reproducible. There are many possible ways to further refine the

hydroxyapatite and increase the mechanical properties. The optimization of the

hydroxyapatite prior to composite formation is a tangential aspect of this work that was

left to future researchers.
















CHAPTER 3
LAMINATE FABRICATION

Design and Nomenclature

Laminate Design

Symmetric laminates were designed with Hyve total layers. The Hyve layers

consisted of two thin outer ceramic layers, each bonded to a polymer layer, and then a

thick middle ceramic layer. This design is based on the principle that the stress at which

brittle failure occurs in a ceramics is controlled by the fracture toughness of the ceramic

and size of the initial flaw as shown by fracture mechanics [38]. Based on this principle

it was hypothesized that thinner ceramic layers on the laminate surfaces would limit the

potential size of the initial flaw and increase the failure strength of the laminate. The

thicker middle ceramic provide the mechanical strength of the laminate. Polymer layers

are positioned to increase laminate toughness through crack arrest and reinitiation.

Similar toughening mechanisms have been shown to occur in ceramic/metal laminates

[24].

Laminate Geometry and Nomenclature

A laminate design with a Hyve-layer structure of alternating HA and Polysulfone

(PSu) layers in the order HA/PSu/HA/PSu/HA was selected for fabrication. Laminates

with various geometries were fabricated for studying how different variables, such as

individual layer thicknesses, influence mechanical behavior. The initial laminate design

was 2.0 mm in total thickness, with a symmetric geometry such that individual layer










thicknesses were 400 Cpm/200 Cpm/800 Cpm/200Cpm/400Cpm. These laminates will be

designated as 400-200-800. This nomenclature represents the individual layer

thicknesses in microns in the order of outer HA layer/PSu layer/middle HA layer. Even

though not explicitly designated through this naming scheme, all laminates for this study

have Hyve total layers. In later chapters, laminates with different geometries will be

discussed using this same designation scheme.



Outer HA Layer
PSu Layer
Middle HA Layer
PSu Layer
Outer HA Layer
Figure 3.1 Schematic diagram of HA/PSu laminate



Solvent Casting of Polysulfone

Solvent casting, a.k.a. fi1m casting or solution casting, as described by Allcock

[44] is a simple technique requiring only a polymer and a solvent, which readily dissolves

the selected polymer. Once the polymer has been dissolved, solutions can be poured onto

any surface on which a polymer film or coating is desired. Drying occurs through

evaporation of the solvents out of the polymer. The result is a polymer film coating on

the desired surface.

Casting of polysulfone by immersion precipitation was shown to produce porous

membranes used for filtration [28]. This technique is analogous to the solvent casting

technique, and indicates a high probability that solvent casting of PSu will produce fi1ms

with some degree of porosity.









Materials

The polymer selected for this study was commercially available polysulfone

(UdellTM 1700, Solvay Advanced Polymers Inc.). Polysulfone has the repeat unit shown

in Figure 3.2. Data sheets provided by the distributor for the polysulfone listed it as

having a molecular weight of 35000 g/mol, and a glass transition temperature of 1800C.

Molecular simulations and experimental values have demonstrated that polysulfone is an

amorphous polymer with little long range order [45]. PSu is distributed in the form of

small pellets. The term "PSu pellet" seen at times in this chapter refers to

characterization performed on PSu in as received form.



CH,

O OI
i CHB O
Figure 3.2 Polysulfone



A small number of composites were also fabricated out of a sulfonated version of

the polysulfone (sPSu), Figure 3.3. sPSu was prepared through a sulfonation reaction

performed by the research group of Dr. Anthony Brennan of the Materials Science and

Engineering Department at the University of Florida. sPSu was incorporated into this

study because it can be used in association with coupling agents and therefore can be

chemically bonded to the HA to increase the mechanical integrity of the PSu-HA

interface. However, studying the effect coupling agents have on the mechanical behavior

of the laminates has been left to future work. The only results of any significance gained

from the sPSu will be shown in discussion on interfaces in Chapter 5.




















Figure 3.3 Sulfonated Polysulfone



Methods

Solvent Casting of Polysulfone

Casting Solution

As received PSu pellets were dried at 1000C for 24 h before solution preparation.

Solvent casting of polysulfone was performed by first preparing 10 wt% solutions of

polysulfone (PSu) in trichloroethane (TCE). Solutions were prepared by dissolving 10 g

of PSu pellets per 100 mL of TCE. Solutions were typically prepared in volumetric

flasks and had a total volume of 250 mL. Once PSu pellets were added to the required

volume of TCE the volumetric flasks were placed on a hot plate with magnetic stirring

capability and the solutions were stirred on host plates for a period of 24 h, or longer as

needed to ensure all PSu was dissolved.

The 10 wt% solution was used for all polysulfone layer fabrication in this study.

The original casting solutions used were furnished by the research group of Dr. Anthony

Brennan of the Materials Science and Engineering Department at the University of

Florida. Successful early attempts at casting the 10 wt% PSu solution on glass slides,

coupled with the fact that higher weight percent solutions are difficult to produce due to









solubility limits between the PSu and TCE, were the reasons that only one wt% casting

solution was used in this study.

Solvent casting of the sPSu was performed using the same techniques as the PSu,

but instead of TCE, the sPSu was dissolved in dimethyl formamide. (DIVF)

Casting of Polysulfone

Casting of PSu films was performed by pipeting fixed volumes of the 10 wt% PSu

solution onto surfaces. The volume of solution deposited on the surface was controlled to

within 0.05 mL. Through early tests, it was established that 0.0012 mL of casting

solution was required per mm2 Of surface area being coated with the PSu film. This ratio

of casting solution volume to surface area was sufficient to produce a 100 Cpm thick PSu

film on a surface. For laminate formation, the casting process was repeated as many

times as necessary to build up the desired PSu layer thickness.

After each PSu solution addition, the solution was dried at 700C for a period of 4-6

h, or until the surface of the PSu was no longer tacky to the touch. Great care was taken

to ensure that specimens were dried on a level surface to avoid any variation in the

thickness of the PSu layers being cast.

Drying of PSu Films

Drying of the PSu films was studied at temperatures of 700C, 1500C, and 1900C.

700C was the drying temperature recommended by the research group that provided the

initial solutions, and 1500C and 1900C are temperature above and below the glass

transition temperature of the PSu. Solvent removal through drying at these temperatures

was characterized using TG/DTA and DSC. The elastic moduli, break stresses, and

elongations at break for PSu films dried at each temperature were characterized through









tensile tests conducted according to ASTM D63 8-02. The 700C drying temperature was

used for the maj ority of laminates fabricated, with one group of laminates being

fabricated using a 1900C drying temperature. 700C was selected because characterization

of the dried films showed no statistical variation between the elastic modulus and failure

stress for the three drying temperatures. In addition, the group of laminates fabricated

through the 1900C drying temperatures was lower strength that laminates dried at 700C,

as will be shown in the results section of this chapter.

The 700C drying is also more time efficient because the higher drying

temperatures require drying to be done in drying oven with controlled ramping rates to

avoid rapid solvent diffusion out of the solutions. Theoretically drying at a lower

temperature should take more time, but when the added time necessary to set up drying

ovens and slowly ramp to a higher drying temperature is factored in, the lower drying

temperature becomes more efficient.

Specimen Preparation

Tensile Testing Specimens

PSu films for tensile tests were prepared by casting 1.0 mm thick films onto glass

slides. From each glass, slide two dogbones were punched using a steel die with

dimensions conforming to ASTM D1822-68.

Laminate Specimen Preparation

HA-PSu laminates were fabricated using three methods. These methods will be

designated: the matching halves method, the bottom up method, and the prefabricated

polymer layer method. Each of these methods was used to fabricate laminate specimens

for this study. The matching halves method was the original method used to fabricate

laminates. Drawbacks associated with this method lead to the development of the bottom










up method, which was used to fabricate laminates designed with PSu layers < 200 pm.

Again, this method has some drawbacks, which facilitated the development of the

prefabricated polymer layer method. This third method was used to fabricate laminates

designed with PSU layers > 200 pm.

All HA specimens used for laminate fabrication had been processed according to the

procedure described in the Methods section of Chapter 2 and polished to a thickness of

0.80 mm.

Laminate Fabrication Method 1: Matching Halves Method

Figure 3.4 shows a schematic of this laminate fabrication technique. This

technique begins by building PSu layers of equal thickness onto HA layers. In the case

of the 400-200-800 laminates, a 100 Cpm PSu layer was built onto two HA specimens.

Layer thicknesses were controlled by measuring the change in thickness that resulted

after each casting. Once the difference in thickness from the starting HA disc thickness

to the new total thickness was measured at 100 Cpm the matching halves are ready for

fabrication. PSu layers were dried in order to remove as much of the TCE solvent as

possible. Once dried, the PSu surfaces were re-wet with a small volume (< 0.1 mL) of

the casting solution. The two wetted surfaces were then placed together and a 300 g

weight was placed on top of the three-layer sandwich structure. The sandwich structure

was then dried at 700C for 4-6 h. The next step was to build PSu layers on a third HA

specimens and on one of the outer surfaces of the sandwich structure. Again, PSu layers

were built to 100 Csm and dried at 700C. The dried layers were wetted, placed together,

and the weight re-applied to the now five-layer structure. The specimen was then dried









for a final time at 700C. Once dried the specimen was ready for polishing and

mechanical testing.

Laminate Fabrication Method 2: Bottom Up Method

As shown in the schematic for the bottom up method, Figure 3.5, processing of

laminates begins by building a PSu layer on a single HA disc, which forms one of the two

outer layers of the laminate. Once the PSu is built to the designed thickness, it is wet

with casting solution (< 0.1 ml), and the second HA layers in placed on the wetted

surface, a weight is applied and the sandwich structure is dried at 700C. After drying, the

second PSu layer is built onto what will become the middle HA layer of the finished

laminate and the four-layer structure is dried at 700C. Following drying, the second PSu

layer is wetted with casting solution and a third HA discs in placed onto the wetted

surface. Weight is applied to the five-layer structure, and the entire laminate is dried at

700C. Once dried the laminate is ready for polishing and mechanical testing.

Laminate Formation Method 3: Pre-fabricated PSu Layer Method

As the name implies, the third method involves fabrication of PSu layers with the

designed thickness prior to building laminates. A schematic of this method is shown in

Figure 3.6. A casting solution in the required volume is pipetted into a 1" deep aluminum

pan with a diameter greater than that of the HA discs. The increased volumes of casting

solutions necessary to produce thicker PSu layers requires extended drying times to

ensure that as much solvent as possible was removed from the PSu films. Subsequently,

PSu films were dried for 48 h at 700C. The dried films were then trimmed to be the same

diameter as the HA discs.









Thin PSu layers (<50 Cpm) were cast on two HA discs, and dried at 700C.

Following drying, the two PSu layers were wetted with casting solution, a thick pre-

fabricated PSu layer was placed between the two wetted surfaces, and a weight was

applied to the sandwich structure. The structure was then dried at 700C. Again, thin

PSu layers (< 50 Cpm) were cast on a third HA disc, and on one of the HA layers of the

sandwich structure. These layers were dried, wetted, and a second thick PSu layer was

placed between them. A weight was again applied, and the five-layer structure was dried

at 700C. Once dried, the finished laminate was ready for polishing and mechanical

testing.

Salvaging of HA Discs

As with any processing techniques, there exists the potential for defect formation

during the solvent casting process. The ease of the casting process means that any PSu

layers with defects could be discarded and a new defect free layer cast. However, the HA

discs used for laminate formation require many days of processing and refinement. On

the rare occurrence that solvent casting was performed poorly, the valuable HA discs

were left coated with a defective PSu layer and could not be used in such condition for

continued laminate processing. The remedy for this situation was to reclaim the HA disc

by dissolving away the PSu layer. Defective specimens were placed in chloroform until

all visible PSu was dissolved away, then the HA discs were heated to 6000C to burnout

any residual PSu. HA discs were inspected for any macroscopic flaws, and were then

reincorporated into laminate processing.











1


12 1

Step 2


Begin With 2 Polished HA Discs


1 |


Casting Solution Pipetted Onto
HA Surfaces and Dried


PSu Layers Wetted, Matching
Halves Stacked with PSu Layers
Touching, Weight Applied, and
Sandwich Dried


2 |)
1


Casting Solution Pipetted Onto HA
Surfaces of Sandwich and a Third
Disc


PSu Layers Wetted, Matching
Halves Stacked with PSu Layers
Touching, Weight Applied, and
Sandwich Dried


Outer HA Surfaces Polished to
Designed Thickness



Figure 3.4 Step-by-step schematic of laminate fabrication method 1: The Matching
Halves Method


Step 3


+
3

,,,,1


Step 5









Begin With 1 Polished HA Disc



Casting Solution Pipetted Onto HA Surface and
Dried, Step Repeated Until PSu Layer Reaches
Designed Thickness


PSu Layers Wetted, Second HA Disc Stacked on
Top, Weight Applied, and Sandwich Dried


Casting Solution Pipetted Onto HA Surfaces and
Dried, Step Repeated Until PSu Layer Reaches
Designed Thickness


Second PSu Layer Wetted, Third HA Disc Stacked
on Top, Weight Applied, and Laminate Dried


Outer HA Surfaces Polished to
Designed Thickness


Step 5


Figure 3.5 Step-by-step schematic of laminate fabrication method 2: The Bottom Up
Method


SStep 1


SStep 2


SStep 3


Step, 4










Begin With 2 Polished HA
Discs, and 2 Pre-Fabricated
PSU Layers


Casting Solution Pipetted Onto
HA Surfaces and Dried


1


2


Casting Solution Pipetted Onto HA
Surfaces of Sandwich and a Third
HA Disc, and Dried


5


Figure 3.6 Step-by-step schematic of laminate fabrication method 3: The Pre-Fabricated
PSu Layer Method


Step 1~


Step 2~
PSu Layer Wetted, Pre-Fab. PSu Layer Stacked On Top
and Wetted, Second HA Disc PSu Layer Side Down-
Stacked on Top, Weight Applied, and Sandwich Dried


Step 3~


Step 4~

Second PSu Layer Wetted, Second Pre-Fab. PSu
Layer Stacked On Top and Wetted, Third HA Disc -
PSu Layer Side Down- Stacked on Top, Weight
Applied, and Laminate Dried


Step 5~


---~Outer HASurfacesPolished to
Designed Thickness









Polishing of Laminates

At the start of fabrication, all HA discs have been polished to a thickness of 800

microns using a 600 grit diamond polishing wheel. After the Einal drying, the laminates

possess three HA layers of equal thickness. The thickness of the entire laminate is

measured, and then each outer HA layer is polished to the designed thickness. In the case

of the 400-200-800 laminates, half the thickness of the outer HA layers is polished away.

Polishing to the Einal thickness was done through 600 (45 Cpm) and 1200 (15 Cpm) grit

diamond polishing wheels if they were to be indented with non-indented laminates

polished through a 5 Cpm alumina paste.

To test the effect of surface finish on bonding, a set of laminates was prepared for

which the surface of the middle HA discs was polished through the 5 Cpm alumina paste.

Failure loads and stresses of these laminates where compared to the same properties for

middle HA discs polished with the 600 grit Einish. Failure loads and stresses were

determined using methodology that will be described in detail in the following two

chapters .

Characterization of PSu Films and Laminates

Thermal Analysis of PSu Films

TG/DTA was performed using the same equipment described in Chapter 2. As

delivered pellets and solvent cast PSu films were tested at a heating rate of 40C/min to a

temperature of 8000C.

Differential Scanning Calorimetry (DSC) was performed using a differential

scanning calorimeter (Seiko Systems). Heating profiles were established through a

temperature programs that included three heating and cooling cycles. The cycled










temperature range was from 00C to 3500C and heating was constant at a rate of

100C/min.

Mechanical Testing of Laminates

Failure loads were measured for laminates fabricated through each of the three

methods described using the biaxial flexure test described in Chapter 2. The process for

the converting these loads to stresses is described in detail in Chapters 3 and 4.

Results and Discussion

Problems With the Solvent Casting Methods

General Casting Problems

The nature of the solvent casting technique leads to bubble formation within the

dried PSu layers. In some cases while diffusing out of the drying layers, the solvent is

trapped in pockets. When this occurs on the surface, these pockets can expand to form

large bubbles, see Figure 3.7. Solvent pockets forming within a PSu layer between two

HA discs form locations of stress concentration, which occasionally caused fracture of an

HA layer, see Figure 3.7.

The other problem associated with solvent casting was drying stresses arising from

shrinkage of the PSu layers due to solvent loss. The drying stresses were of sufficient

magnitude to cause fracture of the HA discs. The result was either a completely fractured

HA disc or peeling of the PSu layer off the HA discs (Figure 3.8). In one case of

complete fracture of the HA disc, a critical flaw was identified through fractographic

analysis of the fracture surface. The measured size of the flaw showed that a failure

stress comparable to the biaxial flexure strength of the HA was reached. The maj or

problem with this particular failure mode was that detection of the failed HA discs proved

difficult as the PSu layers would hold the fractured pieces together causing the crack to
























rlgure 3. 1 uptical micrograpns or (iert) a large surrace ouoole lormeo aunng laminate
fabrication, and (right) a small bubble (black arrow) within a PSu layer that caused
a HA layer to fracture (white arrows).


Figure 3.8 Left: Optical micrograph of a PSu layer that peeled from the HA disc during
drying. Right: Higher magnification image of the PSu layer (white arrow) that
peeled due to fracture of the HA layer (black arrows) which is still bonded to the
PSu layer.









be undetectable through a macroscopic inspection. Once this failure mode was identified,

all HA discs were inspected microscopically at each step of the laminate formation

process to ensure that no fractured specimens were incorporated into laminates.

The second failure mode, peeling of the PSu layer from the HA discs, was the more

prevalent of the two types of failure cause by solvent casting. These failures also

occurred by fracturing the HA. However, in this case the fracture tended to propagate in

the direction of the PSu/HA interface. Figure 3.8 shows that this failure was not of the

interface itself, but of the HA as evidenced by the thin HA layer still bonded to the PSu

post failure. Although no direct evidence was found, it is most likely that these failures

occurred along within the HA along the interface of what was formerly two green tape

layers that were not fully bonded during sintering. Most of the post-firing delamination

was seen closer to the center of the HA discs rather than near the outer surfaces. Through

polishing of the specimens these delamination flaws were moved from the center to the

surface of the HA discs, marking these defects the most plausible explanation for the

peeling failure seen during solvent casting. These failures were prevented through

casting of multiple thin PSu layers instead of one thick layer. The drying stress

associated the thin layers was of lesser magnitude and consequently prevented failures of

this type from occurring.

Problems with the Matching Halves Method

The matching halves method for fabricating laminates was the first method

developed for laminate fabrication. This method is the most time efficient of the three

methods, but turned out to have the largest drawback. This method requires half the

processing time of the other methods because each PSu layers can be built









simultaneously on matching HA specimens and then combined to form the designed PSu

layer. This is a simple two-step process to producing each polymer layer. However,

problems arose from the inability of gases, trapped when combining the two halves, to

escape through the already dried layers. The result was the formation of a large open

cavity at the center of the PSu layers. The cavity formed at the center because the wetted

PSu layers dry the quickest at the outer edge where diffusion of the solvent into the

ambient environment occurs almost immediately resulting in a less permeable skin

around the edge. As the rest of the PSu layer dries, evaporating solvent and gases

migrate to the center of PSu layer as this is where drying occurs the slowest and the PSu

is still the most permeable. The resulting pressure forces separation of two halves of the

PSu layer. The two layers dry separately and can trap a large pocket of evaporated

solvent and air.

This processing problem was discovered when specimens were indented prior to

mechanical testing. Instead of indenting the outer HA surface, the indenter proceeded to

penetrate through the entire outer HA layer resulting in a large hole, Figure 3.9. Shortly

after indentation of these specimens, the distinct odor of TCE could be detected in the

environment surrounding the indenter, indicating the release of the trapped solvent.

Specimens were fractured despite these huge flaws, and analysis of the fracture surfaces

revealed the large cavities at the center of the specimens, Figure 3.9. Thirty laminates

prepared through this method were indented with 3.35kg Vickers indents. Hardness

measurements, shown in Table 3.1, indicated that cavity formation occurred in 43% of

the laminates. This percentage represented an unacceptable number of defective





















17 1.20 + 0.05 57

13 0.05 & 0.03 43


specimens produced through this method and facilitated the development a better

laminate formation method.


Table 3.1 Hardness of laminates made through the matching halves method.


( Ntotal = 30)


Hardness

(GPa)


(n/Ntotal)


Figure 3.9 Uptical micrographs or a large hole that tormed during Intentation (lett), and
of a laminate with a defective PSu layer containing a large open cavity (right)


Problems with the Bottom Up Method

The bottom up method was developed to prevent the formation of a large cavity in

the PSu layers. Laying down the complete PSu layer onto one HA disc provides a greater

surface area for solvent diffusion. Drying of laminates fabricated through this method

was always done with the newest added layer positioned at the top to allow for solvent









evaporation through the new layer prior to drying. Hardness for 19 laminates made

through this method indented with 3.35 kg Vickers indents was measured at 1.33 + 0.20

GPa. This value includes 100% of the laminates fabricated demonstrating that there was

no longer the potential for large cavity formation during processing. SEM, Figure 3.10,

taken after mechanical testing revealed that instead of one large open cavity at the middle

the solvent formed many small cavities or bubbles throughout the layer.

The drawback to this method was that it could only be used to fabricate laminates

with PSu layers with thicknesses I 200 pm. The reason for this is that when the casting

solution is pipetted onto the HA disc surface it forms a meniscus. As the PSu layer dries





















Figure 3.10 SEM image of bubbles (black arrows) which formed during laminate
fabrication using the bottom up technique.



it retains the rounded edge formed by the meniscus. Subsequent PSu castings to build up

the thickness of the PSu layer result in an increase in the distance between the outer edge

of the HA disc and the outer edge of the PSu surface layer. When the next HA disc was





added, the round edge of the PSu layer led to an open area between the surfaces of the

HA discs. When the outer HA layers are polished these open areas act a pivot points and

concentrated stresses which leads to chipping of the outer HA layer (Figure 3.11i). For

PSu layers < 200 pm. The chipping occurs outside the diameter of the biaxial flexure


Figure 3.11 Schematic drawing of chipping failures that occur when fabricating thick
PSu layers through the bottom up method.



support ring and thus does effect mechanical testing. However, with thicker polymer

layers chipping can occur inside the diameter or the three-ball support. Chipping inside

the support diameter was deemed unacceptable was this could result in lower values for

failure loads and thus j eopardize the statistical integrity of the study.

Problems with the Pre-fabricated PSu Layer method

The method of pre-fabricating PSu layers was developed because it prevented the

chipping problem associated with the bottom up method. The pre-fabricated PSu layers

were trimmed to the exact diameter of the HA discs and thus there are no gaps formed at

the edges. There are two drawbacks to this method. The first is the 48 h required to dry

the PSu layers is equivalent to the time required to completely fabricate laminates

through the other methods. Therefore this method is more time consuming that the other

methods.










The second drawback to this method is alignment of the stacked five layers during

drying of the laminate. Due to both sides of the thick PSu layer being on contact with

wetted surface the layers were susceptible to sliding if stacking and drying were not done

on a completely level surface. The slightest angle would cause laminate layers to slide

off-center and as such are unusable. When this occurred the defective laminates were

placed in chloroform, the PSu layers dissolved, and the HA discs reused in other

laminates.

Thermal Analysis of Polysulfone Layers

TG/DTA of Polysulfone

A full TG/DTA curves generated for the as received PSu is shown in Figure 3.12.

TG data was plotted as a percentage of the starting weight versus temperature. The full

curve for the as received PSu shows the onset of degradation occurring at approximately

4500C, with all remnants of the PSu completely burned away by 6250C.

TG curves for the solvent cast layers dried at the three different temperatures area shown

versus the as received PSu in Figure 3.13. The TG curves were used to measure the

amount of solvent retained by the PSu films after drying at different temperatures. The

amount of solvent is quantified in Table 3.2. The retained solvent data shows that a



Table 3.2 Weight percent solvent retained results for PSu layers fabricated at all three
dyng. temperatures.
Drying Condition Wt% Solvent Retained

PSu, As Delivered < 1%
700 C 20%

1500C 10%

1900C 4%










significant amount of solvent is trapped within the PSu films until the films are heated

past the T, of the PSu at which point softening occurs and allows for diffusion of the

remaining solvent out of the film. Tensile testing of this films which will be shown later

in this chapter, proved that there was no benefit to increasing the drying temperature

above 700C. Therefore, it is reasonable to assume that the PSu layers of the laminates

contain approximately 20 wt% solvent.

Consistency of Solvent Casting

Once 700C was selected as the drying temperature of the solvent cast PSu films,

TG/DTA was run on six sample films cast at different times to see, qualitatively, how

much variation there is in this solvent casting process. The six curves are shown in

Figure 3.14. The solvent casting process shows good consistency with there being only a

4% difference between extremes. The TG data suggests that the solvent casting

technique has sufficient reliability to assume that trends which arise during mechanical

testing are a function of laminate behavior an not due to variations in the solvent casting

technique.

TG/DTA Analysis of a Fractured PSu Layer

One of the 400-200-800 laminates tested according to the parameters that will be

discussed in the subsequent chapters has a fracture surface which demonstrated a

considerable amount of delamination. Enough of a PSu layer was exposed to use for a

TG/DTA run. Figure 3.15 is the curve that resulted from this run. The curve seems to

confirm the idea that as the PSu layers dry, solvent segregated into small cavities or

bubbles. The numerous drops in weight seen in the curve could possibly coincide with






































I;P


100


00


S60


S40


20


0 100 200 300 400 500 600 700
Temp (C)
Figure 3.12 Complete TG curve for an as received PSu pellet.


40


)Su Pellet


1900C



1500C



700C
0


0 50 100 150 200 2
Temperature (OC)


50 300 350


Figure 3.13


A comparison of TG curves PSu films dried at the three different
temperatures.