<%BANNER%>

Three-Dimensional Visualization Display Device

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 E20110114_AAAACQ INGEST_TIME 2011-01-14T14:26:57Z PACKAGE UFE0007064_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 216795 DFID F20110114_AABPTA ORIGIN DEPOSITOR PATH lewis_d_Page_006.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
14524b65096559556f777d7950776ac3
SHA-1
19a1e8f19865b5364cd858e41dff405069ba520e
1051971 F20110114_AABPSM lewis_d_Page_078.jp2
f60c3c9c8f1faef0084a16fff828d0bb
36187ca495f36837051df91acd3cdbb649f945e3
43105 F20110114_AABPRX lewis_d_Page_083thm.jpg
89081570db79bd3e36094824329d47ed
605df150fe7a8ff94dcbd3dcdaa979a55d12db96
23332 F20110114_AABOPK lewis_d_Page_119.pro
85524fbae62bea6c8b4ec8fbaf76c66b
7baecda41cff2ad2ad3c2af794f06b689c9828f1
119513 F20110114_AABOOV lewis_d_Page_005.pro
0e0ac6ce3d7c68559f71cb18c88b1de6
a89cf1e7a12376e6db22eec1c8a4c892faa683c7
215678 F20110114_AABPTB lewis_d_Page_007.jpg
220f5772864b8593ccbc0e2d60606e43
041b93e4a7c5475675911bf69c7c989c5951284d
25271604 F20110114_AABPRY lewis_d_Page_128.tif
506b98c1118c9ec192eedb882ec8b60c
2c1a4d8824190dcc8686f79c158718dff223876f
769618 F20110114_AABOOW lewis_d_Page_055.jp2
d5978744a8f4f5142506436a92c57db2
5ace2b48f5e02f5a3838c63b2dab99b537092969
257618 F20110114_AABPTC lewis_d_Page_008.jpg
cc42be199decca15945305a8f865338d
9660b0bedf0bd8b72c8c2fd9a44ece56906e1bf7
63355 F20110114_AABPSN lewis_d_Page_146.QC.jpg
8ab4c1d72f10d7a873a913f4f8097a19
8bd6029449b9c6b764a5fac5b00b13fa40f1a21d
1542 F20110114_AABPRZ lewis_d_Page_140.txt
8109cdf0c1213dbaf8d4f57d093cea96
b58eec810df2838e3733e2329b63ab51ba5e41f2
F20110114_AABOOX lewis_d_Page_020.tif
b539a8dd4e2b6d1617d73d3ce93afac3
d02850c22eb18e4f2fb0ba17c043a4457a37cb78
33626 F20110114_AABOQA lewis_d_Page_020.pro
75c248b086a54747fdbf806524bc22e5
819514a06ac21414847b4fc427cd8eb9161845f8
40347 F20110114_AABOPL lewis_d_Page_052thm.jpg
72d4d28cf36ae8ea6f82a6f394c2a80c
d11e4303f5b85824b70153f9d5baf68c42f884b8
183304 F20110114_AABPTD lewis_d_Page_011.jpg
4c29d870d6cdbe1dac11ffe3f2b417f8
7fbc1db6b09edc895b35ae70e15f7814722cd84e
44939 F20110114_AABPSO lewis_d_Page_016thm.jpg
b48abad77dc58f2b8e629421250c4b37
31459b591ef812712e64b1955c731db2cc16425e
37399 F20110114_AABOOY lewis_d_Page_100.pro
2cd8814766cd3d7e676088fa07ffd741
d8a8fdf5719674022c4852cda256b547885421fb
93770 F20110114_AABOQB lewis_d_Page_116.jpg
b43f83060cd65896e08165392e93402f
3642bf68282689c17d2a002c6b3684334f390d42
1530 F20110114_AABOPM lewis_d_Page_132.txt
d355f641ef91f558e293a351af0ae8dd
b213a11cb50052ebeda672477da78c01fbffc249
155835 F20110114_AABPTE lewis_d_Page_026.jpg
1b9a5302589b207b32213b2d8effe715
386e8d3f55d3618273cf2707ed663858e6ce9fcb
47138 F20110114_AABPSP lewis_d_Page_085.QC.jpg
042d4b27f53bf5582e7c5a0a52ed3bcf
1706f793fbdee485b74dc1cf1aca5d1ac89df8f4
264845 F20110114_AABOOZ lewis_d_Page_085.jp2
8ea44f702b61f037f16ce74a550eab91
0502a826b517c8b1cb630a98ba2a95a856bae3a5
189272 F20110114_AABOQC lewis_d_Page_122.jpg
b487e049127b4dd9f9c3a2a54e72dcb1
343f7750e2b81d486702bab80bca7d8a42ddc633
71393 F20110114_AABOPN lewis_d_Page_087.QC.jpg
bc5eccb7f734ebcfb674c81acabe53c1
e7870b0df791a4720f1e06be1746fd7eb922bd2b
148932 F20110114_AABPTF lewis_d_Page_039.jpg
1f76b984405e4e20cde78d39272ed2b8
45096fcb65dfacf556a0b2a8efce3b46432afa7a
16317 F20110114_AABPSQ lewis_d_Page_001.QC.jpg
d6241ebae32d2887b9851c61aad10037
70d77c862011df8713f2f4b31c4037ecce6dea0e
30996 F20110114_AABOQD lewis_d_Page_147.jp2
77549c04da493d60d7c81e12385f33cf
e4ad81ed43804c612754eb4253826541459f1195
1051982 F20110114_AABOPO lewis_d_Page_018.jp2
e8e6b54d0ccb6818ed5c008aff40e5b6
a354c6d7d7649023f98c0dd97b106f939744049d
166375 F20110114_AABPTG lewis_d_Page_040.jpg
fe9c90aa0c1999a8956788c8021d4bbe
88e0d74c2ca6d01bc4d3798e2f0e8babf3b10fda
49031 F20110114_AABPSR lewis_d_Page_008thm.jpg
81905440790a59684cb60a07f1f1d54e
68fe7d00bfce4e749f79e4c5800b9e4a18c37409
221124 F20110114_AABOQE lewis_d_Page_078.jpg
b3232555f281ced7a11cacd9917665bc
5e3fb9d71663edbd89b1773d65794c4d54aa66de
1026920 F20110114_AABOPP lewis_d_Page_102.jp2
1b72026ef2fbef534793cf8ae9ba2198
d74fa4088facb229cab06f3af2769851d28d7cea
153215 F20110114_AABPTH lewis_d_Page_044.jpg
863db1286aed2d9248b7ac61419e858d
47f69a7b60e2999c3d69898c0e204f9d25273b61
43024 F20110114_AABPSS lewis_d_Page_047thm.jpg
33fad8cf98d3292760f725059a7935c1
3b1d503f0d4c0a180d52ff78d80c54a1ba7ade10
209237 F20110114_AABOQF lewis_d_Page_018.jpg
724f81cbbf78b4b14fd47c29d4bd2f5c
b7c275d9ca98919667d293ff3b1ce8de1c18bdac
1051943 F20110114_AABOPQ lewis_d_Page_152.jp2
b1cc831273bd175c24b5a73607115bad
45255ac0e1cdec837c74f4bd644df10d148ec92e
170533 F20110114_AABPTI lewis_d_Page_056.jpg
832e63f766d42442ba05d5b012c76a51
f5c57234934693e1a3d813da1a18a49779877e4d
F20110114_AABPST lewis_d_Page_010.tif
db6e44e1368fa471f41e4c6ded180440
cd056bb774a54bed37261211483bbb2330b7014a
34346 F20110114_AABOQG lewis_d_Page_060thm.jpg
70f1c6162c1e4b144eb7200956df0bbf
8afd8523281d50d7ee5844208d3309b2cda0bd17
F20110114_AABOPR lewis_d_Page_058.tif
16c12d300abd4b60c13f7a5722df76de
1a6a3b2a87c14480f6b511e6730fe8e31b32b348
137386 F20110114_AABPTJ lewis_d_Page_064.jpg
6de79daf0ced174ff7ea0b240faee0f1
d3584359708a355efdd74ba47428fdda9d6740b6
86872 F20110114_AABPSU lewis_d_Page_085.jpg
20727ebcdcb2f14341e74e0f70ab3fa2
2bf5c465e9dc8e2c9092e3147bf4e30f96ee77ed
42230 F20110114_AABOQH lewis_d_Page_027thm.jpg
7dd9c5c6e99c8ba792d1d3b485b54de2
2da3f5f7ffd9641ad0e37f6a5fcdf506d14c7413
1283 F20110114_AABOPS lewis_d_Page_102.txt
879495dafde05a1a13233d6260a247e5
f7327f157f178df03f156bd75f9334876b7000af
89093 F20110114_AABPTK lewis_d_Page_065.jpg
0ce0bf1a7fbaeeaed9d9a3dd2f1cdf3b
cb83c7d0e96b686b048e9f425287dc9f97dc2306
229229 F20110114_AABPSV UFE0007064_00001.xml FULL
c55af7c7f0706c5151b91c897d76a6db
a0e70d13a4eb32c55bdb61a88287cc94b886560d
46198 F20110114_AABOQI lewis_d_Page_042thm.jpg
66d34a88bdfc194e1310f45105a82366
3679609d15c02c5f722a39060e3edf5ae39d1474
2364 F20110114_AABOPT lewis_d_Page_075.txt
2430307803f64e6b7ad76a084520e77f
88e7d1da2a9d8b942425d278fc2cb997f2c262c6
130111 F20110114_AABPTL lewis_d_Page_067.jpg
ccfa2cb7591f699c9fcb7b28a626d68d
ea57ccbf807742099db4f1e1cb445eb31c477156
64731 F20110114_AABOQJ lewis_d_Page_069.QC.jpg
29b1e8bedd018ac7e6c8db46a36009d1
3c5776d3036109f1569d517082a6ef9bba0b6ebb
F20110114_AABOPU lewis_d_Page_097.tif
a523745a0345d7358afea8032cc1a7a9
3f5cc6a09d459e87c9a3d2dc410fbaf8a67e8265
654972 F20110114_AABPUA lewis_d_Page_039.jp2
88f3d8865a6091cccbfa597e5c65c5b8
dfd5f53d70ef32f39247b247b5a6f4094d7bd520
143866 F20110114_AABPTM lewis_d_Page_077.jpg
8ffc7472769bafb90ed09e3bfd01c2cc
b556cbc594f6c621f46f0e51e0acc5e4bcae7c94
F20110114_AABOQK lewis_d_Page_007.tif
706279b31d55a3f741150d3ff6c11a20
d3d825390415aa2248a6a6f8bcf36a07574857ed
79529 F20110114_AABOPV lewis_d_Page_051.QC.jpg
0168428d29fa28020e6131a95cabec39
c45771102f086e30d7cf33af53bd793e90882c54
750869 F20110114_AABPUB lewis_d_Page_047.jp2
86aa624ece3202ce8af7d314b11395cb
f5971d3a98459928de24f79276b232d0fec49017
131276 F20110114_AABPTN lewis_d_Page_084.jpg
f5707a6ae99c551ea4d4193006c6134e
0e938faf1b4d26a00e7ee9b9b792a063989f21db
240440 F20110114_AABPSY lewis_d_Page_004.jpg
cfd47b998de1403be386116e47c71f6d
d7ef0b3702b68a5005f05c413423a7af11944656
175673 F20110114_AABOQL lewis_d_Page_098.jpg
f8c5d6c2609ce1f9f40b6517f731790b
51b7cec5e01ed2bea526988b77e72c16f1409c64
156394 F20110114_AABOPW lewis_d_Page_140.jpg
4e45e0a8fa8b3b9f0ff16c80edfc40df
e7ca0929f9a7ac924907e0dca680c202e159f227
84162 F20110114_AABPUC lewis_d_Page_056.jp2
f9ca6da6620e336324837344a67d1430
67dc082e456c949ff7e5d27a6db7e5da1c7b2dd8
303810 F20110114_AABPSZ lewis_d_Page_005.jpg
d2d6479c29a93682ea8eff8e2113a109
8266094b273a1fc74a1c71177d681cbb0e846934
153338 F20110114_AABORA lewis_d_Page_121.jpg
d9c8b799d3033e08f574b8b5df92f770
cabe1afb174d054795ea26363040ce3521072e20
24216 F20110114_AABOPX lewis_d_Page_088thm.jpg
ec62efd60439a74d931a7e2b07a68dd0
46c9738584ef20632c6c2a9cbd8d5c7d2636618a
629814 F20110114_AABPUD lewis_d_Page_064.jp2
845ecaee5fbd6ba680831915bfadeefe
894e0289ae5966263b268d05d002af79006b5859
170814 F20110114_AABPTO lewis_d_Page_094.jpg
51ae18df975e671ca213e0d8b792a93a
7f18b97b6bb489f489999003ef3093bbc866afb9
F20110114_AABORB lewis_d_Page_109.tif
a887c0541868373b887e3e578c5d9bda
ca3a0a738f9cad6f87edf02c96b4511bef36eff6
23050 F20110114_AABOQM lewis_d_Page_038.pro
8680db608a0f115a86362e0abcbaaff8
305ab063770dfc65b0ef715820d1261ab204d524
16516 F20110114_AABOPY lewis_d_Page_148thm.jpg
1723b56985bb147f85e16446bf75c12a
10a47617e59e886976df3803117f6186e8fff42a
596943 F20110114_AABPUE lewis_d_Page_068.jp2
ba657bd714e4b6555e5f0164652a6cb9
d3776dca818bee4ab92c29a29be8dc185747fc75
74459 F20110114_AABPTP lewis_d_Page_095.jpg
8df9add25f339aa52f0fb3fc3011bdea
03944b5563d498f4eab01592828d53f3d705332a
20169 F20110114_AABORC lewis_d_Page_147.pro
943d00ab99ccba56b675c4ab7c8e1746
2bf871748ad9427d0b2fff30c6f07fa677b8c96f
8006 F20110114_AABOQN lewis_d_Page_003thm.jpg
acd9f24e66165f47dacc69dcb05a3552
70e60bbfa506dcbc957d6075b34da2261292acb9
21288 F20110114_AABOPZ lewis_d_Page_151.pro
cf5986a2cd6f4dfa33ab19d7ad692a6b
bc479b4069743ec2f350b4e88f74d92ba1e78a85
623405 F20110114_AABPUF lewis_d_Page_069.jp2
d16ed21bdc2c660292678b5d36eb5124
8a1fa73ebb1d8779b33fa37661530b1ac69faddb
193413 F20110114_AABPTQ lewis_d_Page_111.jpg
6059d8ea7ca99cea7fe4d464bcc236a7
750911633cd86598ebd11a6fac908622432dbc10
88750 F20110114_AABORD lewis_d_Page_120.QC.jpg
f4d3ba44a96e4c3cc219624c3ff6d179
98c68056c15c03834f8a0828a96e55407a80991c
50493 F20110114_AABOQO lewis_d_Page_009thm.jpg
fcb31a9c4770c891c657ecf982be4afe
8aeb652e189c8371c0986c313ddf3ef16f127e1f
82859 F20110114_AABPUG lewis_d_Page_100.jp2
323efed16ad2ff6221df988b36547d2a
ad6040fbc104c25fd197db8d2a750de28b5e5ff7
216973 F20110114_AABPTR lewis_d_Page_129.jpg
a36ce0febeebda5d1d032c06d97f61db
27ea35c52a9bb694a256f7a218613a978cbfb895
762921 F20110114_AABORE lewis_d_Page_066.jp2
e6b1a53231f1d5620fa81979de60901e
62f27c4f55bf26dc57d0e00ea7ec7acb9388e73f
794 F20110114_AABOQP lewis_d_Page_064.txt
dec643dcfe03ab8dc55154fe42e4eb91
49b2cf4e70e17b9920bda0ad1619107f2e683c2a
1051979 F20110114_AABPUH lewis_d_Page_103.jp2
6a0701e36349db8eec3a4820af0b6806
1f45682ef47cb8fb79ab154f597a8c7671e0a87f
195239 F20110114_AABPTS lewis_d_Page_135.jpg
6f7a04e62bad579366e4b6d0f38f5a68
e492cbfeb9bbdedaf16c6cb973067ab0357fa161
696688 F20110114_AABORF lewis_d_Page_044.jp2
3e257c51f56fa3b72444008cbad6560c
53ade689c7108f13b2b76a67a23d0d37a0e41dee
14156 F20110114_AABOQQ lewis_d_Page_062.pro
2dca00e5236c84869789372888b2252f
daf6cd76f7fa537e71a82d8ceeb652eb2e832f48
152133 F20110114_AABPUI lewis_d_Page_105.jp2
0803af5810d05363711badda1503ad35
306146da7c341126a8a6b496474b85da3744ab53
1051849 F20110114_AABPTT lewis_d_Page_015.jp2
6fcb984f5abe371b4945b1b44f6991c8
017074593e2c01c56e31924378ba737fb3220ed9
42084 F20110114_AABORG lewis_d_Page_013.pro
bc105c465cc2bc58b1dc44ce566712c3
395898ad0424936551c04d54f071168cc9e009ba
39310 F20110114_AABOQR lewis_d_Page_133.pro
7ea0ade94dee0e77545612cca7b3f595
1ceb1313d15cee4de3447070ac56ed774dce1ccd
76911 F20110114_AABPUJ lewis_d_Page_115.jp2
b59411fa6fd60fff997e1b7979794cfb
f0815912d6b16675e8be7d1dd04d3a635302dc32
951338 F20110114_AABPTU lewis_d_Page_016.jp2
2a0308a6215ad4b83fd99d9eed684df5
93aac6203d85bd88193de429ee94f2b59ecb3767
75084 F20110114_AABORH lewis_d_Page_022.QC.jpg
8b493bc4bec1fe9cfacab80c7fa018c5
9cad56f877384fe6b3488d96e8714d53085f3c00
92379 F20110114_AABOQS lewis_d_Page_035.QC.jpg
55a5ef3d371e11a3944b9ac89d1545dc
9707df6a8406e54f51f99391e6e6a4590459e1e8
709185 F20110114_AABPUK lewis_d_Page_119.jp2
98527111739ccb593adc6b92d1d4c9b8
1e6ab5c4fb0043aa83dc9afa95f27a5c6641a496
851914 F20110114_AABPTV lewis_d_Page_019.jp2
8f238a7bda001df49de2f5da6060de46
2bdee34d2e2220d8061c816e1a93667d877231f9
932895 F20110114_AABORI lewis_d_Page_108.jp2
6f29d7260446bb40c723aa4571d3279b
5cf756c19942c21911348c9945bcadf321ffd04c
18235 F20110114_AABOQT lewis_d_Page_150thm.jpg
12bc6349989791f9709d731abcda2532
7b06d449ef3f62bea2a38b40130035a4f46577bb
892395 F20110114_AABPUL lewis_d_Page_122.jp2
fd50e47acc0ef99536889b68cf9d1797
69fb1e276b19af3d6daa87e17f16e78c0e58b235
946375 F20110114_AABPTW lewis_d_Page_020.jp2
1630e9b6aeb122c25a7b9c192e776e2e
764dfee9436b42a48da8fbdb517436f2546f1d07
184175 F20110114_AABORJ lewis_d_Page_113.jpg
2ad503cd403d91e2bb67edd601a03397
60d4638946777e7cd552b71716699ca290f0f0a6
1053954 F20110114_AABOQU lewis_d_Page_151.tif
f8df37cf36c7af7aa19a5519b3f5e722
86080c897c271b3db6da580cf873ece7ea916602
F20110114_AABPVA lewis_d_Page_081.tif
b00ea494dc62f486579281ee6f8b612c
69d0da939bdd2d3bc2153d7fdcc9726a2e4fafc7
446336 F20110114_AABPUM lewis_d_Page_128.jp2
5c5470fc6e79713622b9c2a9455d4832
5c7802098c307acd980803e4f9f87813ccdf3904
798143 F20110114_AABPTX lewis_d_Page_022.jp2
c3443be6c0faf5ea1a0dd541cb2468ff
b7f9c4a4b17bc1c3c5a226839be86a3e174ecae3
1733 F20110114_AABORK lewis_d_Page_092.txt
e2742f5bc61c86da7a5c5ddd87749557
0830deef3843a19a57e04b33226f649f7f319c27
1381 F20110114_AABOQV lewis_d_Page_114.txt
446ecbd1f86c12fc028dcbd1d66abf55
58fdd02dddc7ae7749cec7fc1f05e5d59d3744ba
F20110114_AABPVB lewis_d_Page_098.tif
393c1e5371893f5671f2c6d9ca9df43a
3c7d4aa1f64563066ed1c2cd214f1503ff55c47d
105707 F20110114_AABPUN lewis_d_Page_149.jp2
4651b3de4cfbf546fe85ad86d2c670d7
fc7a21c3087b7de6b3ee1ab27206cb5e69fa1c1a
80513 F20110114_AABPTY lewis_d_Page_023.jp2
e589c537e75b9b86065feab58c3a53cf
fa2ea7ccaa47c92a1d1d4d8a7d76688e6618e69a
142582 F20110114_AABORL lewis_d_Page_134.jpg
8ad1a1bee67a9babbb01dcc8e8af9e78
708cecb1ca0f81363af6589debbf0fda380319ba
103855 F20110114_AABOQW lewis_d_Page_087.jp2
56ee5e25bc3e7ff04bd79016d849c1b7
8ff2d9fda4e0a4172b293ac304a3aafb44d1aecc
F20110114_AABPVC lewis_d_Page_110.tif
5ccf598173df611fcf93a588927439f6
85f2aea84ffdd68e6463fc5f3451dec69fa2bcbb
F20110114_AABPUO lewis_d_Page_001.tif
7053231935b438c16cac04ffe93d7688
a2bd9151658132536cfc1b696d5799f09e600436
513368 F20110114_AABPTZ lewis_d_Page_036.jp2
767e16cf2c903a67cdbdcb0f27bd840c
13f633b19cbca7518d56afc3519c74f5ae75b9ca
185943 F20110114_AABORM lewis_d_Page_132.jpg
46b09992b0c11250fafda9e5207f0245
52e8099d7f6d6d1bce1fb355917525fa2178ce58
1724 F20110114_AABOQX lewis_d_Page_096.txt
9c9d8f1358617851b38a7ff4776bab97
6f6cd0916f9a5835cc58979c1d5976a4aa6f033a
123860 F20110114_AABOSA lewis_d_Page_024.jp2
a5e47f93efc9c80f59a2c438f2c73bdf
006c56c52fad8e8a0af8a1cbc49491895dfe3087
F20110114_AABPVD lewis_d_Page_117.tif
eacfcb4a9a73976b51149033cc4e4893
b4fb890769478000aabf2106500935da70062ed5
180122 F20110114_AABOQY lewis_d_Page_097.jpg
99806266167a8a2f1a39006368d06a23
d7ba1a139ab414b3715ae7494d2d7f0420fadd34
F20110114_AABOSB lewis_d_Page_056.tif
8daa73596077dfdc65e013afa593856b
7117fea56c9055af25c3eae33d0576124242543d
F20110114_AABPVE lewis_d_Page_121.tif
dd251a916eaefb20099e911034755bef
a8b2e6eb0edc5b5ee6fe6918d1771a5b74f4693e
F20110114_AABPUP lewis_d_Page_003.tif
a06ec0ae1b08673adbf50548fb4420d6
a79b10ad63b213bb4d81ddee8bcbf2b7983c7cbb
64012 F20110114_AABORN lewis_d_Page_014.pro
27fedfebfdc57658f23a0462b71c4728
7c33d774463078e1f9a4c1f70622840e88f73346
1162 F20110114_AABOQZ lewis_d_Page_049.txt
d7e0c575dd6388046449decc8ba120a5
7dcbc338b28b1c9da668ddccc197c12b3e213f01
44721 F20110114_AABOSC lewis_d_Page_059.QC.jpg
9bf6f70dc3f5d2dda48b7ef3a4c8f735
52d330ef07322ad8dabdc0486c537030edcbe499
F20110114_AABPVF lewis_d_Page_143.tif
2c37c36b9f1c3aea8c3d669d8e783314
2221a2819850947ad1290faac26a55b027026292
F20110114_AABPUQ lewis_d_Page_013.tif
4633ef11284793489a2c28e7ae6ae3f1
177e2e7d06576d65fd21e20346240afcdac56f6d
30349 F20110114_AABORO lewis_d_Page_018.pro
bf0170d7cdb378750911c283b2634f6d
7e9d768a46bb65aaa1cb5fefc1a69cee5031d336
45224 F20110114_AABOSD lewis_d_Page_091thm.jpg
8ef16c350013c4406ba3cb60d29db9bf
9e6ac0dabd4dc34c63728955b1faa2479a29ceb2
62005 F20110114_AABPVG lewis_d_Page_007.pro
3ed4a5525da883bcd15cb34e692cad0c
f010234f3798d99c8995fe3ce73ff338d92f2be4
F20110114_AABPUR lewis_d_Page_027.tif
6a9adc43b6c3d439cb5b3b051c6446b0
3d1d3162e74794c4bb333adecc97c6229ad83069
1157 F20110114_AABORP lewis_d_Page_083.txt
fc6c0eb443fc054e01086c01ef79fa26
43532091f5908f979451fbac6a32431b4f846db3
54553 F20110114_AABOSE lewis_d_Page_071.QC.jpg
7936825b07ac3e7a6fc2ab37fd2bbfee
116b6306e521ac04e177bc98eb646681783c50b8
75452 F20110114_AABPVH lewis_d_Page_008.pro
607becdc96053082eb1cca432b209ba7
55fbcbfa7a442b7a66146ab22e933c6e5b6749bd
F20110114_AABPUS lewis_d_Page_037.tif
7575eddcec5cca98f23222e8378b8192
f62689373b86c9a3f3085b3377ea23e50da6d5db
1050369 F20110114_AABORQ lewis_d_Page_017.jp2
1b75412df3be6def3159bc1edc19f4b9
010ef79f9044afc523bd3b9f5e2f57f33fd2bf17
F20110114_AABOSF lewis_d_Page_014.tif
b56e4a76cbdb251ce74c16ab06b036f8
459638158b5a851a102c9e1972d05482506eeb31
44286 F20110114_AABPVI lewis_d_Page_011.pro
530275a5f7ad64050e51cf094c32d5d0
8171dd1976cecbcffba4bc0ec04603a3f6b396d3
F20110114_AABPUT lewis_d_Page_043.tif
2cf028aaf8c9dd524e5c3ef7af26282c
7e30d0ade6b2f526326403094767b0aba598be06
66915 F20110114_AABORR lewis_d_Page_041.QC.jpg
29f173ee9b0d45794d51b6528ad86174
57bd83596ec438bd497c60a24ddb94f03d036be3
164365 F20110114_AABOSG lewis_d_Page_141.jpg
7778f4efa7922faafd20c333333eb8a8
91ab6115c6554e163fb47bb044923211ed3ef397
35097 F20110114_AABPVJ lewis_d_Page_019.pro
ed816993ad8295ab0818561d3012d100
074f60981f832a43473fea8ade7dd5f8912a42ba
F20110114_AABPUU lewis_d_Page_047.tif
ebaec54696ff3ee77590a5d1e71a7523
a1ecfc52c590827026ecf8352804b7d6755eb091
1266 F20110114_AABORS lewis_d_Page_032.txt
8827980f0cb8c1a0b6af05c9f9572fd1
3aac0d956cd7605f6cdec0b86695b8834dfcb37a
1491 F20110114_AABOSH lewis_d_Page_121.txt
210346804d724850abe6d835cde2a95d
da3cf62de7df8975333bac1a81b4fe22724c2763
14540 F20110114_AABPVK lewis_d_Page_021.pro
728f1566ffdedf9f34c443c024c995f4
e146da3d107626115bf932b619ecc4f995b378a0
F20110114_AABPUV lewis_d_Page_051.tif
c20a33e8831f6c11305e4ce98c5adcfa
d17a19bef61f7272eff10a7c0b58712f51298537
2992 F20110114_AABORT lewis_d_Page_009.txt
0c62cdd4adb4971f8d14f0aa3f14fd6b
c8f97999b23f29df7e4e77df337e18c7234ce879
201583 F20110114_AABOSI lewis_d_Page_054.jpg
8349b7bd351e5be565d27994851972d6
80b8de47ed753ead4eefd5348f75841ebf6141fd
24324 F20110114_AABPVL lewis_d_Page_044.pro
20119073641d9031ac61f8dbf970f011
08f6459b45e262d76d111da82f0bea31204eb7f5
F20110114_AABPUW lewis_d_Page_055.tif
1bb9babb7a544726e3415016a46b4f3e
00a41e0b5448f6fbe765281e8433506f15282328
1051955 F20110114_AABORU lewis_d_Page_129.jp2
d9b2a15257f64bed8c8ba88f066fe64d
b0f49ace5e597cbfcaf09b04795c43d3afeee27d
41792 F20110114_AABOSJ lewis_d_Page_025.pro
02b059fed44e916cbbe90738ea84a7e1
197e5ba6f708bf725df5136b59983ba8987fdfb8
39958 F20110114_AABPWA lewis_d_Page_118.pro
873c93ed63fbcced8fa6da28748a414e
59e58252d88bab294dc3092b3d17f7467b3af2c6
26014 F20110114_AABPVM lewis_d_Page_049.pro
a90a47ebbb02fbb78c80c38a9d5846a7
f634eac5e0a5c75ffee1c99e01521f049e0a9e34
F20110114_AABPUX lewis_d_Page_067.tif
501b03c72b4f88f5bb5a55157760b8bf
e4857b95b81fcd548f4870e38bb10cee87bf19a0
F20110114_AABORV lewis_d_Page_045.tif
d5989f2460a28ea4258ebc4c86add5cd
4f9878a333e86e08e3765f31bc0d6a9e46f678aa
20928 F20110114_AABOSK lewis_d_Page_001.jp2
87bb663483e6c571098ad9eaaa42403a
24f670d4d70db74190b74dfeeada740739e26fc7
16304 F20110114_AABPWB lewis_d_Page_124.pro
762243d1c9f70d2bdea5e6990d4bddd5
e8f275c83ececd11749a7d48ae67c71f8e51c5f0
8596 F20110114_AABPVN lewis_d_Page_052.pro
1d9728ff262d060d88c088ac3194927f
552b0d46ccf1e1be925ae6e759dd3f5d8d225fef
F20110114_AABPUY lewis_d_Page_070.tif
3175acbb847ea8ad28295bef0ec92f08
930d684c911b1534158133269d15f934f3fb71b6
75443 F20110114_AABORW lewis_d_Page_126.jp2
1de0842d1cb797d26a4fa9aa7491a96d
190ec8e6d1e6e69af3584d50e65ef82abb10a60b
93290 F20110114_AABOSL lewis_d_Page_108.QC.jpg
c12ba4ee516a80065e057ae07bdf3b6f
fa130ecee9d104e9caec04e69bc2c0c43524dff8
49131 F20110114_AABPWC lewis_d_Page_130.pro
bf6c5b6078b4afbf6ae5dba412f6e75b
8f46acaed8c49ad7b0e3290109fdaab7a5fd0a32
11307 F20110114_AABPVO lewis_d_Page_061.pro
9f3aea7c3b53a6f05e976fc5616bb49e
337735d810d948cbe1301d50c41c730844d447b8
F20110114_AABPUZ lewis_d_Page_076.tif
7c3f431ab9380b430823d2a750c8581b
cbe281b855943426d1eb1420d82def5e92a18d9c
35019 F20110114_AABORX lewis_d_Page_071thm.jpg
5eb12725ca2d2a0271143afee58cd341
cc5787e38770c26ae39774154e583df2a9e70d08
84056 F20110114_AABOTA lewis_d_Page_054.QC.jpg
f79922defd311c01dcecc29948e06697
12f9271e9a96f8ab622b3d43bd902b52907b0fd7
1029 F20110114_AABOSM lewis_d_Page_081.txt
a94f93741fe58d7e07edadb35f1cf82a
a697d36c3c1f4dc67339b396ceaadb781ec0066a
40056 F20110114_AABPWD lewis_d_Page_135.pro
e724f77a731873a832fd74ffd513ec57
a8812b74df22d7df1f37a2b39adbd3968a13196d
1560 F20110114_AABPVP lewis_d_Page_071.pro
3646dbd48fa37eb466fd378610c27996
a6a99869f360ef8bf154cb8f1040369794f60d1d
686933 F20110114_AABORY lewis_d_Page_058.jp2
0f72c6ede288f2862de6311db2bdec25
1f4a4aba2448c32102ee4dc42732fade3c3be8fe
1360 F20110114_AABOTB lewis_d_Page_033.txt
e61f63c8cc7d937b615859a53a665168
8a06cd8d93e4d549a23f1cdb7aec24cd4e633eaa
144151 F20110114_AABOSN lewis_d_Page_081.jpg
a08d0d46fcc76ea175a0ec003c3c840d
a87403bc5c7f3f6d41786158e77da8dc00b27aa1
17804 F20110114_AABPWE lewis_d_Page_142.pro
9338ce89a41372a582da457284c4dcd2
2d39e0cbd19ae8d21d96b702c809c20e16598bcb
15349 F20110114_AABORZ lewis_d_Page_068.pro
3fc556f2d6947deb44020d806bb62787
daa3904c7f7957aaada2e2cece514d1726b67790
694064 F20110114_AABOTC lewis_d_Page_141.jp2
f336d81d97d698e36cbbce4923e2f4e7
c0e053995e0b2e884f7adc580c140daa0030d79d
1673 F20110114_AABPWF lewis_d_Page_012.txt
20b263c9f9ebfa0767d360146dfac614
0bdd411e99f9fc9ad15c36b73f872fa25c0abcba
33192 F20110114_AABPVQ lewis_d_Page_073.pro
45da4455dff5f07f4cbf7d90af901022
3c6299a959e378b8e22a9ac806721f37a3bb9da3
17178 F20110114_AABOTD lewis_d_Page_045.pro
a8f5f41204a6c3e425754af2fa92bf4f
7b01e933bae5bf7b2cc76641b710702c84f27f7f
84899 F20110114_AABOSO lewis_d_Page_132.QC.jpg
e1050b3c581934ea35cef4d14dbcdbb7
0bd89489847eaa73a5b09179e0c5404048e4114a
1612 F20110114_AABPWG lewis_d_Page_023.txt
b790733c6adb17a8be0254580bfb0f61
5944ee36ef1f984916aba9df633ac4db6026faa8
42373 F20110114_AABPVR lewis_d_Page_082.pro
ac7c1483087ce1140ed1fd6d645a6ff2
e598e8bc074e188be681032fda94b09d90563ab2
199306 F20110114_AABOTE lewis_d_Page_033.jpg
053d65f45765c3b3e9b2f0fc0b619fea
7aa66f9a6d5e8ccf74b90eb3baeec4a75f9e56c1
180420 F20110114_AABOSP lewis_d_Page_082.jpg
3658b65ef43d18a4e732ffcf787d84b5
b2c92c3b4f77a873b3eae14192e2cda7d93cc214
942 F20110114_AABPWH lewis_d_Page_027.txt
87170a9bbee40c92387a2ca0573c2da5
cadc6d36fd43e3806626d1fe988a32cc3ded3e09
17119 F20110114_AABPVS lewis_d_Page_084.pro
9970171015301605c6afbd4dfe3308fb
a19e42f3eb4abff6b1e11140f35a5acd21f7f8fc
169940 F20110114_AABOTF lewis_d_Page_025.jpg
9d41510e41336d8ff8b8104b930cf191
066e31fcfbc5a0aea594049b1ce22132f02bfb68
178950 F20110114_AABOSQ lewis_d_Page_150.jpg
77ffab2c046b64cdb296bf5ae6184c48
f7158e4c39b2b5b451f1acb75de432a8fe830112
1002 F20110114_AABPWI lewis_d_Page_038.txt
fdc4458dc8d11f927e35cc7cb412d111
4be930e85d28d1fca045d88ce6b706b30861cb84
37588 F20110114_AABPVT lewis_d_Page_090.pro
4ef1dafdaf646b915ef113e5ffb648b2
894ca404f4757da40a76b784120b6fdf2ba1ebb5
1759 F20110114_AABOTG lewis_d_Page_031.txt
b3f8d38bf193c077fbfc8dfb28e27d6b
1d76f0f244c940ea74a31876d2f5356a4f895ac2
131420 F20110114_AABOSR lewis_d_Page_014.jp2
8dd166f4bf55fe57e143635035ba9530
40a15184391dc15cdb9146d79a52abf32d2ca6d3
1365 F20110114_AABPWJ lewis_d_Page_041.txt
415eec2fd0f0777e00397208983dd6ab
7d6a077e0d41b275fb879e5fbf5866434058e12e
22968 F20110114_AABPVU lewis_d_Page_098.pro
d3bc1ce16f5d12f5c0ec37c3ded7b718
3f97c9076949109f672af46d4c2ace925e80a49e
76414 F20110114_AABOTH lewis_d_Page_009.pro
2e6134186257ef59094b035a7d7828b1
82d7c6db99b27f045db6f7ff44816af647006b1d
40861 F20110114_AABOSS lewis_d_Page_022thm.jpg
e326b4bc17c7a9b152887a96dfdb0d89
c1ab5f49cbc4f0dde6b0f29728b0ad622d17e1a0
1758 F20110114_AABPWK lewis_d_Page_048.txt
68c44b35881ed24456a0e018cd754478
486bac266a4b6dd473333bcd96f906b1f4e7cc99
30493 F20110114_AABPVV lewis_d_Page_102.pro
94f5cac71af1e30cdbc5bbde74671e3f
45368dcbdd5828101544f3c06fb5c6288a1b08c4
45994 F20110114_AABOTI lewis_d_Page_057thm.jpg
5df16adda6e5514b1597e2165f741d09
c8f5809fb1da30f47325dcf74cecc387758caa13
1051986 F20110114_AABOST lewis_d_Page_099.jp2
bb69142be03434f968124aaa34a69666
1aff60d6d1a7404e14563dd2b1cc4a3fe3343b70
649 F20110114_AABPWL lewis_d_Page_051.txt
6d6c99f1f2b46d746db2a37691ec45e4
6cfcda84367c6d7d0002fb5a50704d6498f5b924
23673 F20110114_AABPVW lewis_d_Page_103.pro
baf215e4cec8ada2f6213dccc9c6f195
4a8fd11cca754f852af6d4de3c837bbfd6282b04
9113 F20110114_AABOTJ lewis_d_Page_053.pro
25c9d6576a22b58bfd6a88e981908c80
e75d177c88089ce63e70cd2022c12efc14aafab1
41423 F20110114_AABOSU lewis_d_Page_055thm.jpg
a0938f5029bb5fe003e464283a491549
284fd0b2c7db5a941b9a35e792b3f046fb8a48a8
1918 F20110114_AABPXA lewis_d_Page_152.txt
72f0fd61635a45450db6b73f58be620c
30df2cdc291e99fc7f07f3cfe3080369f774159a
434 F20110114_AABPWM lewis_d_Page_053.txt
db2fea284ae90ec54b19df4aee9433dd
ab8685829b1c19d55d0011b7f751904b77a19e5d
30880 F20110114_AABPVX lewis_d_Page_108.pro
05d4737c559b00c1db7daeb8328aecc8
43ee68e96c32b6bd9d819094361fa6b5e9566bf7
18567 F20110114_AABOTK lewis_d_Page_138.pro
a679e184d4a51c16302e544922defefa
cfce0eae9bf755440919a723828d816fa3fb25d8
85411 F20110114_AABOSV lewis_d_Page_104.QC.jpg
9ac1106e034175c3fe830e183cd64ded
3b0f46bf91f9c9b2e127a71217a126a3c0a75c19
213 F20110114_AABPXB lewis_d_Page_153.txt
6a87b19e4b1f780265883d016ab4eb34
55bb655d1b85f573b8d0b666a21bf424bf5cde88
80 F20110114_AABPWN lewis_d_Page_076.txt
71e7a28f153887fc9b9183444c1517f8
ee9a94be0655c75a0062cb7e42b20374381d9b0a
54126 F20110114_AABPVY lewis_d_Page_110.pro
263c7e5cf510d3845c3fa69f4a7ca55b
2b98e3bc45c7e0b7b067f3e2cf89f2ee0f7a1673
71809 F20110114_AABOTL lewis_d_Page_049.QC.jpg
03111936b3d37b9ba8606b5b20e26a57
ad58bfad438301f942fb42ae19501ebbe52e7336
39967 F20110114_AABOSW lewis_d_Page_074thm.jpg
201ced7cda03b20063dc074cf85dd5ee
96006bb88f0932eb3ba41ba909a2e3baad98733c
6059 F20110114_AABPXC lewis_d_Page_001thm.jpg
80be9138323cb51e68636ceb776c2cab
0912fccb641600f6b8d236b31b7a8c576af6b3eb
2361 F20110114_AABPWO lewis_d_Page_088.txt
c3d069f2af51fc167aa58f4b15341fed
b7099986d9e51d38ce8ac046d20fbca76db9f84e
20513 F20110114_AABPVZ lewis_d_Page_116.pro
2338b46f1bbadea69c0323290cf728f4
f5998edc4d2eb6482e9918c4e20819e79b09c137
F20110114_AABOUA lewis_d_Page_114.tif
effd4366252b6203a9f3693ae4cb69e2
c7d65c6151a5f21db9d0d5a70bfadae6e98c4c87
F20110114_AABOTM lewis_d_Page_092.tif
8b97edbb19ddcc31e1eb10cd52cc7dba
47c14307678e384801b50a1be8986bde5dd44cb0
F20110114_AABOSX lewis_d_Page_018.tif
ad146dfdf08425390f168ab1dd39e7f2
51083aedc10dc27ed053c3ce35d227d730fac638
120569 F20110114_AABPXD lewis_d_Page_010.QC.jpg
2ac4e54c61e661cef2b9d54bfceca7c8
f8ff662980e562f3dfa6f0053e2884a8de23eb47
1061 F20110114_AABPWP lewis_d_Page_099.txt
5012ad004f51089506164b75b00988bf
9b4be8c709732e2eb51014ceb5a79ae81b58855e
F20110114_AABOUB lewis_d_Page_154.tif
16ef4903e144c06976b30afdea51c0f6
218fdc5f3e6909ca6f525d29426018668b3dc7b2
F20110114_AABOTN lewis_d_Page_120.tif
a88321c4306355f28fd729f46887109a
21e0bf79286f51f9f0691ca42b964226babc280f
F20110114_AABOSY lewis_d_Page_059.tif
81e92fc22ca68fb8208d8465bab0817a
6a1a471715039594dae07bd6027fbbaf0ceae026
21483 F20110114_AABPXE lewis_d_Page_013thm.jpg
f55a01068b68a13b8157dd5cbcf2bdf7
c041cffcfcdac8d718a6273fd4ade649c7617ac9
961 F20110114_AABPWQ lewis_d_Page_103.txt
db570fd1e71e0c1f8869cbc978868038
99afc0a603f5b9281a7cdfbb0256bff8ada90eec
80309 F20110114_AABOUC lewis_d_Page_144.jp2
b344aa002e6f6c5a490f32c898f89473
a424ceed8ed195fe2508993ce3767111bfc735ea
2956 F20110114_AABOTO lewis_d_Page_008.txt
1a083272c2925db650b73a3c57136654
d08096fd454a1efee1f8e4e36c85a60ee78c44dc
146586 F20110114_AABOSZ lewis_d_Page_115.jpg
908c289ce333cd110ed7862fe3e67af9
4e6e5b37f8d8ceb0b4897489ffaf7065e23d093a
86194 F20110114_AABPXF lewis_d_Page_019.QC.jpg
f4e97336545357ac70fecdfa72daee13
d6a4987479d4baa9d2eebe66fc80207165f11453
47179 F20110114_AABOUD lewis_d_Page_116.jp2
24a5ab47166147b449512609e51679df
5e1bde12960348770dcec1f78e7faf7703ef94fe
45531 F20110114_AABPXG lewis_d_Page_020thm.jpg
96f4b524f9094acfdb084791010a6948
77f7eba771e3bc86602432fcc68f6a74bc5491ed
1355 F20110114_AABPWR lewis_d_Page_108.txt
cd08a679a8e24d0da28cb0efc8db40c1
b4971a083a06288c8ee37f0f891d9ac064d4e1e1
225100 F20110114_AABOUE lewis_d_Page_145.jpg
2a5cb7c044f5fa4f71ee76256406c632
a841ad476554cbfb2111467c0a26bccd5c8dd19f
F20110114_AABOTP lewis_d_Page_126.tif
62602864cf137e6cd8c98ddd48615321
886272d5f09262b22dfd7bce7216ce567efbef7d
25060 F20110114_AABPXH lewis_d_Page_024thm.jpg
f0e173ebc1e77cf0675a4b25a82c5b5a
a56b09d72cb31a9b03ff56548b94bba16605d811
2310 F20110114_AABPWS lewis_d_Page_110.txt
c92d372172ee638731f8b9031e39149b
9efe58d51d19f992f069a4e6b17c9f41ec80db0e
300794 F20110114_AABOUF lewis_d_Page_105.jpg
c41ace392510b734ac317166e96fb66d
394f5595fbdf0495493a7ea570db926f32e21e9f
98085 F20110114_AABOTQ lewis_d_Page_101.QC.jpg
5007e833f50a207fdf9ec196f0128e2f
1b999fc9c1f79f6623c9e65872b058ff23de4629
71808 F20110114_AABPXI lewis_d_Page_026.QC.jpg
f1665e10e8e34e84762d7a2d8c712f5c
70b307c519125e33ef073453a63cbe0c5e3b3aab
1972 F20110114_AABPWT lewis_d_Page_122.txt
cd956d2e29ab7691c173961446faddfa
3e295cd01237019fb791a42eab1a3e8265822b8c
81876 F20110114_AABOUG lewis_d_Page_010.pro
e08b7723c6eaca5e0dcf08c1ab286f4e
318000be363d7add6ba53069cc5a42acda498ac2
975 F20110114_AABOTR lewis_d_Page_030.txt
7dffffaffa18e2f2eb68cbd398fa3993
2823e186ea0f72c890d545a55dd6e0dfa89d3bca
28438 F20110114_AABPAA lewis_d_Page_097.pro
b8fe937046c890aada846f244d9ccfed
ec42edc5289f96c470368682f3f6e233a69709c9
73660 F20110114_AABPXJ lewis_d_Page_027.QC.jpg
e0490b5ec5b847bc5f0bed6948c7ed66
a587098b47a562124d1235040bad361b581e661f
851 F20110114_AABPWU lewis_d_Page_125.txt
4160bdddcb4013a7440a957ced29f555
0a2a090cd7e7d54ec1e20a1ff546aeac1777eb30
485 F20110114_AABOUH lewis_d_Page_003.txt
6a006d3a8a2f055d31a31e99d023337a
5f9d79655b65b02147742656c8ee880b673c9399
1311 F20110114_AABOTS lewis_d_Page_089.txt
d03d2d72677408ff4df223c0b4d76489
61f7738dc7ba5cf9b9737bade260afa22a9bffe1
181602 F20110114_AABPAB lewis_d_Page_038.jpg
16860d59da18c3604e890366cd3812a1
3f792e84516543b87e9b536e99282d21eff18f36
65756 F20110114_AABPXK lewis_d_Page_031.QC.jpg
726474509a6b638a9a5ef3aa5e087dcb
49fc226e468a8bf62d5cf9503b29067a57e055b3
1621 F20110114_AABPWV lewis_d_Page_127.txt
72b78ccbe716a21b4b0e20a953a8ba38
3b15eab7e26cba9d20e8f448b6c264efc68f5859
927881 F20110114_AABOUI lewis_d_Page_035.jp2
f9379c6960a0ec22a811dd360daa2376
75d5b52fed8c4440016a5830ed6b78f558e61aa1
107886 F20110114_AABOTT lewis_d_Page_071.jpg
17018d5ae28d20e2782a64e224b7038e
974819008c714831f4674862ecd3469bc2ba8cb0
94306 F20110114_AABPAC lewis_d_Page_105.QC.jpg
889dfabd62ff3b612143589d22ceedc0
08b5a5ec8e1c31a0fc9c721534005612c9f2c698
79541 F20110114_AABPXL lewis_d_Page_032.QC.jpg
16a380b41f0fa03816dc12e8e634d1ab
3bf0ca70b800432980937007fcec91cd1bfd7c44
1086 F20110114_AABPWW lewis_d_Page_134.txt
fe699dab27b5d0bfa9ab2a1cf1d2597a
49a9b747760318e1770f037b84b09c9799b9f06f
49440 F20110114_AABOTU lewis_d_Page_144.pro
d45912df2de3d44d1297332c0d68315e
41078f3352aa340d93fa6f6249d080d6c982c627
49381 F20110114_AABPAD lewis_d_Page_089thm.jpg
b5c334e9f129f775b9cc1231472e6f0d
091b93d04d32accec9ef868689fc87ed34863f3c
186724 F20110114_AABOUJ lewis_d_Page_048.jpg
7377145f8c615de6c0d283809979623f
81d6751cde9099285ccf9733a70679cba9f6b7d4
47480 F20110114_AABPYA lewis_d_Page_102thm.jpg
3822f45f621d33c6ef80d5516b3dd5dd
3e415ad183fdc4bb1ad90244b33ec63bd1941fd5
89194 F20110114_AABPXM lewis_d_Page_033.QC.jpg
524d1a7c5b4f633a0c29de07fc63c741
c9e46dbcd962ed29ddd7435f0dc996a79139c5c3
1488 F20110114_AABPWX lewis_d_Page_136.txt
600b454b40dfba59c9cd63f384fa6993
1576559914343c61a3435bdc6cf134382ab513fa
F20110114_AABOTV lewis_d_Page_052.tif
9236f4690659ee8700199c1b321f7fc3
59c5aa29ff3a1f3340cb70f87b671828eacf8230
F20110114_AABPAE lewis_d_Page_116.tif
23934dbdda4ecb4dadfc1b15123bf8c9
ecd60ba0751b85c4200823471d072d060d9e9ce9
42293 F20110114_AABOUK lewis_d_Page_032thm.jpg
3dbba7bf970195c4b09669c19fabaaa5
be03abf38083a1f97ba14e6c8b0962b0ff62607f
20622 F20110114_AABPYB lewis_d_Page_107thm.jpg
e24572ea618d96fa3fe3d687ac4c0242
bdc3515b574ae82dd4c8036563a18bfbac95d0df
64575 F20110114_AABPXN lewis_d_Page_036.QC.jpg
fddfea8cc0a68173884b505a71d1ca14
1b71ab9b4c5c2d560b8c845be922b9544e712a8e
1140 F20110114_AABPWY lewis_d_Page_141.txt
fe57c6f38149e0ea5c49cb8722bf901e
62d82b9d83c16bdaddf6bfac6b65bda10f62ee06
71124 F20110114_AABOTW lewis_d_Page_113.QC.jpg
7f72ae1f709626235397d1e0ccfedb5e
f23e5a28e8b48cad6dc2f6c5494715dc9f88e3a0
41982 F20110114_AABPAF lewis_d_Page_042.pro
cf4b92a56105490a1394dc517b38dc7c
dc20bb14c8e9b440f2a40e002cfb91b110ae25b6
877 F20110114_AABOUL lewis_d_Page_151.txt
bb1c50aa0021f473d3ffb69a8ac61daf
6806ae3fee2e804fbf80b97c2b2a86cb348cc93b
46519 F20110114_AABPYC lewis_d_Page_108thm.jpg
3da978c169085c318574fbc988c67b0a
4bcde4e111a090185811e2f91266ee560a8d6ed1
39602 F20110114_AABPXO lewis_d_Page_037thm.jpg
c155ed59b44d06decf3c9a7dbfdef0ad
0f74e20f6a732c21ec5dd9d28b7aa4a8f96eb2da
4099 F20110114_AABPWZ lewis_d_Page_146.txt
fe4fa0e92dffc202e8f7a6227681b1ff
1c9e36b0e54edb5eca0c61eb15ffe81155e0ed18
76249 F20110114_AABOTX lewis_d_Page_040.QC.jpg
b1cd33faf6d2c5b268355bcfabbe981d
986220e4b790c6aaea22695a17af6925a4599a8a
F20110114_AABPAG lewis_d_Page_024.tif
6176b2eb03900b424705208a45f4c097
67c06e898693ed9e8f96466e66a1ec3b8d03693c
39454 F20110114_AABOVA lewis_d_Page_041thm.jpg
3a190e93b205f425b9421969d99f4932
d9b8747f5dc3794f2486366f15f03d7b06c1f455
1303 F20110114_AABOUM lewis_d_Page_055.txt
8b242bc2b8b162db9bc99e9ceb3b50c0
589400f17bf26334f9dfae089d505d391e7e9650
44018 F20110114_AABPYD lewis_d_Page_109thm.jpg
8463a6a0d65dc10d13d7b6de86796c97
b4f4a07d0bfff213bec5969bf7761468187a1123
74463 F20110114_AABPXP lewis_d_Page_052.QC.jpg
436dc9f12d122f642b4aa8536fa7ac3e
550cc2bd346e1a4f145b1649aa6d10df9cfed9a2
23128 F20110114_AABOTY lewis_d_Page_041.pro
6ed917aa6228e36682d9f4132af38859
aa282adb9eb73f92775a014a95947cb3ececd662
62626 F20110114_AABPAH lewis_d_Page_133.QC.jpg
9761c5b71f2dbd69511e81d24de97f6c
195dbeecc567a95bd1fc43247fc2ac37451a8ff3
413 F20110114_AABOVB lewis_d_Page_154.txt
6a5c40879b3c6b25eb7a0903de0a603a
cda60077361fb7961e906e72f407f35b56119438
19209 F20110114_AABOUN lewis_d_Page_125.pro
0c16c7eafcc3be085984f9d7383eb125
072aa2e785a54711936f7eea31e5fa44d2c1757c
84732 F20110114_AABPYE lewis_d_Page_109.QC.jpg
c93d65c3f62d6f17380ce39f20be9143
0f2e6a4a801353669fac20577b7d924ad9be28ed
65724 F20110114_AABPXQ lewis_d_Page_053.QC.jpg
503d62da0fa4b8255cad9d2a903e2032
9cb7a35ee3cc31e277944a7c38dfa9cd1734459f
65980 F20110114_AABOTZ lewis_d_Page_145.pro
2dcc49141886082d99a198c52e1c67ee
34a1f489007c82bd4939f480baf847f9c556ca2d
1979 F20110114_AABPAI lewis_d_Page_079.txt
5473f301b3cd1aa84d4f8be330f24f8b
8e39596235b316d4fbfe1dce08df2830e60edeaa
177 F20110114_AABOVC lewis_d_Page_095.txt
0c38186f520dbcb861a2a22c4072ff24
5252bf39a7d6620e82eef78156b31ce23e7b63cd
46525 F20110114_AABOUO lewis_d_Page_111.pro
e6fd3cfb212f18f11a53f5ef13708ffc
5696a411250de0bbd330c5e2e53903efcfe67aaf
48121 F20110114_AABPYF lewis_d_Page_112thm.jpg
b97fe6af5a8d3465b698a15cc6cf4a05
f326cfb7dea8db8fcf6aeab7aed0b947772f75e9
54926 F20110114_AABPXR lewis_d_Page_063.QC.jpg
2c3493c0e1f9beb7f9f6290f0ce9d168
f254bf11ce2a556486e53fdc731579e4d967d77d
575830 F20110114_AABPAJ lewis_d_Page_134.jp2
6be497f5c25373c4e1bc537d7a88ce1a
955871dbc2820501909b9cd7931b0ddd84758b45
54165 F20110114_AABOVD lewis_d_Page_086.pro
d53899dace6cd33a037d18e67d8b0775
2abeef1dbe0d53d99333c4c0a0a5469f600f83bf
45528 F20110114_AABOUP lewis_d_Page_090thm.jpg
81f2a89b3759ba920dbbf3e0d4c30dec
e19463b398dd05220989c3e409db1c008914e862
35085 F20110114_AABPYG lewis_d_Page_116.QC.jpg
85761bfbad52a2d37647167abc108315
b67a23d8fbfed7de9edb1f39d4cd48dbc8de3a11
86781 F20110114_AABPAK lewis_d_Page_094.jp2
156634d4fc534d377d1921751ee92c5f
37e5ceff90d33ec2c9a963252ec6af9ddf410714
4850 F20110114_AABOVE lewis_d_Page_005.txt
0449e4e0f758f6cf9cdf306d6d44f8dd
4a0ff710d40b5d6398395890c0bfab3b707e59da
21325 F20110114_AABPYH lewis_d_Page_118thm.jpg
779ca6d57ac70bccda65569b25f2786d
be6b5303599ea3c97faf57e127db169c14611939
44223 F20110114_AABPXS lewis_d_Page_066thm.jpg
0c78eacd7bd282218405292ee77eadc5
e9d56b7850b004bb95b1ac8f736d9c3d9421fba8
1752 F20110114_AABPAL lewis_d_Page_011.txt
68feda7bd512800c8057c52e3ffe56d5
b7a1ba04b973926b7e0015b0d36ce920a3c68828
43308 F20110114_AABOVF lewis_d_Page_150.pro
3d3df465fc35c4ca9c008a493a78d258
5335a146755fdba647e2d8f955b210b7c984e2bb
F20110114_AABOUQ lewis_d_Page_102.tif
6cbdadd080cedb683de68844e648a327
3fc2ad79fb8a9e1f1add9614f8841b08283cb53a
20900 F20110114_AABPYI lewis_d_Page_121thm.jpg
8562afb3da043f469adde76f2144ed9f
57f7df18ea436959744073c5e5ff5471bae6422b
37874 F20110114_AABPXT lewis_d_Page_067thm.jpg
e577e962045cdab0d5635813baeb9699
d17b24bd26482b7cb50eaa4638dfcdcb6dc70e29
F20110114_AABPBA lewis_d_Page_085.tif
2a4dc92feb685d0087822efbcebdad82
54ad8dc9e90c0f470f214542e2b5d48d81582745
23914 F20110114_AABPAM lewis_d_Page_081.pro
23357e5b9dd7b06deaf07cb95117db23
6463d868d51e183559d9c2fcc849f2e6c8d13032
102 F20110114_AABOVG lewis_d_Page_071.txt
07d697e64add2b3847d4dff50e5e2f50
81a516e0e44f5673b70c5aa980788d89c5b3f945
38145 F20110114_AABOUR lewis_d_Page_127.pro
7ddace65046f38c05790b65ca75b2650
28125bd166dd8ee98d2904f89e4d9ef5db2b68ba
24318 F20110114_AABPYJ lewis_d_Page_123thm.jpg
3773d6365447ebd8c654dd45c5e18835
933f71307f7173fcccac9e6741266e5b28776bcf
62111 F20110114_AABPXU lewis_d_Page_068.QC.jpg
732856e024dea04c1c81c97aa1fc9337
2a167ef9a8edd8fb9d471f0b306da033cfa9b562
2273 F20110114_AABPBB lewis_d_Page_135.txt
c1f7e46fbeeb0530a21603ac1a330301
642a4958eb5105b35c3b349b779833f2ca498f62
56116 F20110114_AABPAN lewis_d_Page_093.pro
d7de0b69c9762277dd28ccb868be8ba3
8fb8f438c25160adbd3afb355096927a8ded02bc
40914 F20110114_AABOVH lewis_d_Page_064thm.jpg
fd9f15941cd2cdb419bb2c8de51a0f9c
749d46093524c202b179af56795914ea6689e03d
1443 F20110114_AABOUS lewis_d_Page_019.txt
7f2480453365a2f063b2cb1a9b2d0dee
029e09374441530299af8822ce4f2beea6915f53
95387 F20110114_AABPYK lewis_d_Page_129.QC.jpg
21aeb96df29d1ffeb810b1f3116dadaf
54fce86cdf307d549185bab551be0746396d19ab
21600 F20110114_AABPXV lewis_d_Page_072thm.jpg
0b73f2477c8ce857356f5227d6f520a4
2528aec681f346e7793b7bf623c27f878567bf2a
81285 F20110114_AABPBC lewis_d_Page_046.QC.jpg
6e02af424a0dad9c3d04483d19c1f068
846e961d68b16d3df3591f4bd783e5e3c2e8fa98
48888 F20110114_AABPAO lewis_d_Page_154.jpg
26cdbc5b61b86faaa6528f5f5181dad8
ddd930e02b0dcfca4e3110aa0ba1b810120a9b98
F20110114_AABOVI lewis_d_Page_021.tif
d155333904e006f313522d9b7e6356f2
117c99e20722cee5855554186497ef2de452ce33
46256 F20110114_AABOUT lewis_d_Page_029thm.jpg
548c6a7592c0ca50cba03b3f8e9b49fb
00fb9cc72137e3e86f8a6ca9ec05b8fdebe54ace
44693 F20110114_AABPYL lewis_d_Page_132thm.jpg
cc9498be40bc6f42c295971aa8a138a8
60370d68c962a3b0622d39b5fac54a9212bb531d
35619 F20110114_AABPXW lewis_d_Page_076thm.jpg
dd3246f48d76d9800464027d435eb3f5
577cb702037173917eeab826880ebdd19b15e703
44055 F20110114_AABPBD lewis_d_Page_113.pro
5ad6613e0cce6f9c9b2f873200e1d99e
c836db7a475e652ccc1875a43c43271d4e1ec72f
398 F20110114_AABPAP lewis_d_Page_052.txt
ef58df781d7d7d5bb90e16859b6d1fc2
a6066e2ec7815cdfdfdeac64c94e010625a2e6e1
87445 F20110114_AABOVJ lewis_d_Page_122.QC.jpg
5ccf927e3830e4308cf0cf7f53ff581d
5c88ff17d84700d7bfc6a4c989258e3cf1f6269d
34795 F20110114_AABOUU lewis_d_Page_115.pro
1774b9f2d6bef71a579054b51e1fc6e6
e0a688327e2fbcccd1b605732a3da68c4afcf91a
71361 F20110114_AABPYM lewis_d_Page_134.QC.jpg
1320de53048951e2937e249b8ac62eb8
4b8d4755d25ad0d2d1bb6b5438cccc45e84ba48b
66084 F20110114_AABPXX lewis_d_Page_082.QC.jpg
a5ed7cfc04857098a229e97226940a53
c4b1c82b1a2d142e1de60f751feee3c0f9ec4f69
52315 F20110114_AABPBE lewis_d_Page_115.QC.jpg
991a1cff7fde80773a293194472a821f
eb22e325df6bb87bd794744b493611269066ed01
791299 F20110114_AABPAQ lewis_d_Page_091.jp2
da49a0267e8322f12e6c99d085f1fa92
7063e8930e04e1f2073ab03f8b0eb5d611c4c6a0
45584 F20110114_AABOVK lewis_d_Page_127thm.jpg
09b338b345fe71bfe9aa25d4eeb5fa82
252919ec4d6b491a513758636a7cc9893dd462ba
19813 F20110114_AABOUV lewis_d_Page_094thm.jpg
d95fe2a153abbfae33975a1f435ac278
c744003b352dc8ac07943301508aa6d22b52ff92
37316 F20110114_AABPYN lewis_d_Page_137thm.jpg
14218f2ff1ee6fbfcedc98e413182798
79155c6c0d855f26fba665851a317e06f97d27d0
38068 F20110114_AABPXY lewis_d_Page_084thm.jpg
c96983a436527c836752509e6f22136b
db124b90ed5c59d25f337a7a10da7e5199996bd1
24331 F20110114_AABPBF lewis_d_Page_030.pro
50528792f9b474203d4fb8b6cf848dc3
9f0b14c2099d93f29a54bf00931493fc6752615a
257171 F20110114_AABPAR lewis_d_Page_014.jpg
ef096c7604ee19d15481538f5f714123
fa72a20822922422d9d70561eaba71f40a88741f
60902 F20110114_AABOVL lewis_d_Page_100.QC.jpg
ae5e018cd0480f0bde68ed1dc783f19b
5d01176be4a5a577c34971755fe0f8cfbc68f9f4
231098 F20110114_AABOUW lewis_d_Page_152.jpg
00623bd288d111392dd32ffc0ab6728d
c450d6651d9aa2a8f09baeac252b5d1c9e233591
59171 F20110114_AABPYO lewis_d_Page_137.QC.jpg
16447d5103a3503bf46ed28202d667b4
5f090657267509363471d9ca8de4ce3076adf524
87068 F20110114_AABPXZ lewis_d_Page_090.QC.jpg
cc3e28529ed6d1e203ff4a800da32620
1d3418fc6515bb8f66b394651fc3460ca3fe9cdc
228664 F20110114_AABOWA lewis_d_Page_093.jpg
19d8ee928c6a851d33f63f5c0e0aa767
9bd704c1ec03cf930ee6b8d3ea6c411017c65f96
36689 F20110114_AABPAS lewis_d_Page_023.pro
3533dd77365baed5b6176e5b65d31968
60ad8ff700e70934c9c2c353482f376d8947b8ab
674829 F20110114_AABOVM lewis_d_Page_062.jp2
dd873c7435f2bf775d152b78e4c5c87a
71bed139d4d86b09497ac8e79e9fea07bda19365
F20110114_AABOUX lewis_d_Page_030.tif
4593e0025d64733aef59b19ef16836c9
eb235e91b5ab3e35551e75a6376e89ef6a7e3d46
F20110114_AABPBG lewis_d_Page_084.tif
1d6886c75128e14757ca316256559e46
04040440c086f70893750f1b19ed68ae860ccbe9
43668 F20110114_AABPYP lewis_d_Page_139thm.jpg
9c19ee0162e052dab610c930d7cbdcd4
60cdd42244dea71b870d2871e4a73daf3cf4e4bd
27830 F20110114_AABOWB lewis_d_Page_059.pro
795fde9e1abd1198b1fdecdae248d4bf
9f0d41e3ea3bef99b22cac87e7d786eaaa064d8b
195387 F20110114_AABPAT lewis_d_Page_015.jpg
7eb8757131de7b2811fa34fda111c9b2
17df50f12dd8b229f884c5f4560cfc76415de670
122326 F20110114_AABOVN lewis_d_Page_059.jpg
e6be9cbb739442e480699f450569af37
faa334ecfefd38d88c37fefdab94ae8a67cadef5
F20110114_AABOUY lewis_d_Page_119.tif
96723cc6e25eba258c93f5270147e8e4
3e1aa6ba113a45ec24425cc8f4f232daa4cd5c1c
F20110114_AABPBH lewis_d_Page_132.tif
650f836a9c8df3fac5390eb59b5b3a4a
45d6ae1bb15db021602ca46d54b2d6eae0658286
17713 F20110114_AABPYQ lewis_d_Page_140thm.jpg
bb6a6064a42b5b80ca69a7f6b16f4540
e0b59a549376760cdb8e4138afb9f3864009fe09
86339 F20110114_AABOWC lewis_d_Page_135.QC.jpg
1d205139747745af0872925633789a94
058cde352395ab665cf2fb0417647e8bed41f783
1629 F20110114_AABPAU lewis_d_Page_118.txt
024448ee9f6455be3037c2576e466d86
8753e1b8bdbd913d74d55cb385c83a46caaef5ac
64354 F20110114_AABOVO lewis_d_Page_118.QC.jpg
d8ac1ba3bb32c7961f631bb8e9b6203b
f0594991c52ce7c85ad969728caf77a3a2b1e319
22056 F20110114_AABOUZ lewis_d_Page_003.QC.jpg
0487d02706a323d0a691a23393ffcf3f
a1593b6bbcbe2a481910680fad2f11fc3fb001d3
1051984 F20110114_AABPBI lewis_d_Page_011.jp2
9999278e0a8e85c6f7e88eb2a3017bad
c5bcb1f0b7403d31061d3c365834aba8879abdce
10550 F20110114_AABPYR lewis_d_Page_151thm.jpg
2ecc4b35873dd0334840e7db99371a8e
8700bd2a3912558110a68d547592c67614564282
1455 F20110114_AABOWD lewis_d_Page_115.txt
a873009abf5b38a709cc6b03924e4bbf
8320927b5b96fbaf2bf15d3b74b1841772ac15ff
46489 F20110114_AABPAV lewis_d_Page_001.jpg
81cb07d5156bed34a84340da5f366d89
a758c567205cece24a1d6703bf8c4048bb185799
73208 F20110114_AABOVP lewis_d_Page_021.QC.jpg
38740351d47b407a1b3fd007eaa44e5c
98608ded955230ec5fa05d1694df5c555848eed8
723 F20110114_AABPBJ lewis_d_Page_138.txt
0a8dc7196f0ed2a588d18d3da3f1b853
e0af64c04f6d27ac617040c6ca619404ae4ec4c5
176783 F20110114_AABPYS UFE0007064_00001.mets
06d920fd905a70a65d25e9789222fc15
b840ef04e1281ce315d2bf448aff9869e9b34223
44667 F20110114_AABOWE lewis_d_Page_015thm.jpg
635a513e805015b2a82c17b92b15bb82
4680b1ec0ebc34e4a928562d6bbe45fc7cf9347a
F20110114_AABPAW lewis_d_Page_109.txt
1ff5b60a25c6594fea9d184748968e77
cfc611067262c073c55c0b7c1be616d5c4aa42de
21773 F20110114_AABOVQ lewis_d_Page_110thm.jpg
0d41f9472911d3ab2920b4ac8e776c2c
916fe0b6429611bebcf2b8920366cf33c20aa7ec
21879 F20110114_AABPBK lewis_d_Page_087thm.jpg
6039a52f188604a9ac29248b93a5775b
d00792278596b585b31daf680b15476ced769eee
198099 F20110114_AABOWF lewis_d_Page_042.jpg
e1515663128dbea82586d7b62b88a8ba
cc9d73363683bcd7af101b1f27408a235db59cd7
339520 F20110114_AABPAX lewis_d_Page_065.jp2
f63000a3773aa6ef25ca126a2820b28d
62ef72ed1032ffc84d04b85d2b87dfec3118967a
702752 F20110114_AABPBL lewis_d_Page_032.jp2
2deb336069ef60df55d0297a4ab951d8
8269993e51d2b9130c5e54261f95f3d80c3ab8ff
943 F20110114_AABOWG lewis_d_Page_142.txt
5d4f03acdf132071f6db109d60af34ad
e3a53a21b885d4cdc8344a346e8a457504c02321
36898 F20110114_AABPAY lewis_d_Page_063thm.jpg
e00c18b62d67b011bdeb8d5206ad0873
62f39b96eec61cb8e317e53c42c028b973d0b1b5
944 F20110114_AABOVR lewis_d_Page_036.txt
514e288baee926e577088a81c33b35f3
c646de6d66b39df8ee0486a7dac6c56a592197d8
98057 F20110114_AABPCA lewis_d_Page_103.QC.jpg
cb8f1f95450a51690ac8fba19551e531
363f710f9435e7eb3890586aa153b83ccccfa0a5
97572 F20110114_AABPBM lewis_d_Page_063.jpg
81b8c3ad9df8701a9686fdc0f1201c8e
97d7e94de49aee27fdf8c87661d33c04bcf874e5
158931 F20110114_AABOWH lewis_d_Page_136.jpg
6f6a684e366aa6e05ca4b97edfc2f3d6
4559850390a20e3c6ca6b827acc881a29d2a6a6c
42177 F20110114_AABPAZ lewis_d_Page_040thm.jpg
755ea9c7452286a6ab72dd8eb005dfb2
7ba9feae73c925d3071f5ed39a73d59345a3b43e
976068 F20110114_AABOVS lewis_d_Page_070.jp2
df7ff32633f3facf7ff40ea861fd3dfe
dd9791a3b14370e67fb325e77c34342f0971c42f
639482 F20110114_AABPCB lewis_d_Page_125.jp2
1f0a8b5f436ed129f73f39e018ce5472
678b2608b9036729b0a7756ae20a12d69b40f44f
68924 F20110114_AABPBN lewis_d_Page_077.QC.jpg
11966680cd14d9b729196b319be6232e
8b12998cabfd843041dfa3965f5262c7a4419aa3
44322 F20110114_AABOWI lewis_d_Page_141thm.jpg
2f68ecef27f03fba55dce2791a01e783
4404890420da4d13f0cab039e57b0ff59c52ba74
20691 F20110114_AABOVT lewis_d_Page_031thm.jpg
43d5b9b0d4264248759dc3faf4422440
4dace052f53b1ab493fef736d19bf9adf363657c
64646 F20110114_AABPCC lewis_d_Page_149.QC.jpg
1dec56c9556a1d9a72f38c58fa1494e6
6c4a7cfc1cf0301dec6555e75be29c1d311471ac
9186 F20110114_AABPBO lewis_d_Page_154.pro
88873c8a3d188234a477945fd02e6454
339f5895e4ba629169873a917aee499daead2d37
1124 F20110114_AABOWJ lewis_d_Page_147.txt
7d679ee38a97fedc2e68c41b64f3a6f9
cd0cc9eef1e86c5d83b911eaf0692823d832abfa
F20110114_AABOVU lewis_d_Page_093.tif
8c94d6b1b2589b945f2919115f6616f9
325328446305cfd51ffb386a647d881c40cc909a
F20110114_AABPCD lewis_d_Page_061.tif
295c00a821a4e56fc8ef5e649b6c927c
c201b59086f9012a1755aad7f7eba5f89095ecd8
79714 F20110114_AABPBP lewis_d_Page_141.QC.jpg
aee9ba163dfa012d9d5df193b842a43a
cefbdc7225bdeb6af5b01934be59b044cc468922
204800 F20110114_AABOWK lewis_d_Page_035.jpg
a8dba7d7b173ebafe4bec903cd7f4e44
76c104419af23db08883ab3916baef681b730198
3236 F20110114_AABOVV lewis_d_Page_010.txt
607d5e1bfe5efcf4cd721c06349b5e95
02dd792d816d6298e966a8aff52ec7930edac1a4
96748 F20110114_AABPCE lewis_d_Page_048.jp2
3e6066f8b24af8f67c12b2ac5a154eb7
3dd5dd0f36062d22d84504800d6ff90ca38e6445
106 F20110114_AABPBQ lewis_d_Page_002.txt
17649fa772d7c31c2b777e7d52671225
8b6921d7148bf0f4e2a8a002fdec294485c6bf48
37441 F20110114_AABOWL lewis_d_Page_132.pro
6f8d3a28d15225f7e9fa3228161cb320
c7fc1a5a28470845414c92cd9a9a9ef0e6c4fa2c
94001 F20110114_AABOVW lewis_d_Page_152.QC.jpg
15fb6d480549291917aca6101d88c6f9
5f69b4cd0f025358899d20a25d3dfa61cb91ba2d
F20110114_AABPCF lewis_d_Page_075.tif
ecb69f9fcabe353d20e99ae8478f1a15
f3f479a45a057f88e6537c5e418da846b6098f4a
1350 F20110114_AABPBR lewis_d_Page_020.txt
fa0f84737eeaa5c85bda81e8b0ba1407
45e62cd5e9d2b74bc10d6e34b877fa347e5fa9ea
79439 F20110114_AABOWM lewis_d_Page_011.QC.jpg
bc1602bc8eef4c7af58341385c4af269
f320652d0baf2155ad5c8dd11273c09b65753491
546668 F20110114_AABOVX lewis_d_Page_074.jp2
3df8c0306440d1d5959d40dcca49f46d
1c027a302f55802e5526bb6def47628763bdabf0
179345 F20110114_AABPCG lewis_d_Page_091.jpg
9e293a9bbbe692dac428aaf67dc5176c
34715353a96d7c0a06862cffc76c494b1abbfddd
1163 F20110114_AABOXA lewis_d_Page_097.txt
b5cdd9ed9309acd9ad82c75c85e92a79
4910aa5592106a3dea4899ba00f1d6fdce8cda81
47067 F20110114_AABPBS lewis_d_Page_017thm.jpg
66b78bc249ca4ec1442b68d717e9187c
4ecf88d653be7daaf0e34d3d172de3d5389e8713
1684 F20110114_AABOWN lewis_d_Page_133.txt
e5e9919755891ab660f5c9ab7a350a62
ea483ef004575f328e455e177152e9ebc0bbe1f6
5566 F20110114_AABOVY lewis_d_Page_002.QC.jpg
4b784ae820d25b55e2334e87c9b84933
b11d5aca35a44c76a169cdc34b6233a1341f3e66
50348 F20110114_AABPCH lewis_d_Page_106thm.jpg
add06ea6c0b5f74a40930f5ff3abae1b
5fb52b8cc679ab1799985001f46c717a7b7425b5
202500 F20110114_AABOXB lewis_d_Page_079.jpg
675654c4a1933d0443fb1d50dda7f26a
67c7461ba3de7c2ba234244cc724774844e6e25c
42455 F20110114_AABPBT lewis_d_Page_051thm.jpg
6f79b55e2a9a777fe01f16abb77bff3f
ba112a5535ae3d8d2570ed9c03fda11917c15c8e
131031 F20110114_AABOWO lewis_d_Page_128.jpg
2e0a0ce43b742da39c10c232d6511803
0970a42c1e97f1a63cb40724315826bb0b18c68c
213438 F20110114_AABOVZ lewis_d_Page_102.jpg
a315b6072a9974d6f8bf51194729a356
fa32d57c79b53951110849e45ef0506d2e9ad362
19449 F20110114_AABPCI lewis_d_Page_064.pro
c9936b70047335296326ae37b1d29bda
9912a340675448340bcd291426834b84acd6d9c6
F20110114_AABOXC lewis_d_Page_009.tif
128aa55c16062b3aae612845bdf68c6d
0e6f7c396ef3e7d5c4f6bc1e7d4ed7e8a568e3d4
32337 F20110114_AABPBU lewis_d_Page_095thm.jpg
5a98b2478c9712e2a3f67d005e2c57d6
eaa82bda0eae9c1bea9918e11b54b11cf8b22bac
93520 F20110114_AABOWP lewis_d_Page_112.QC.jpg
9c529d7ed5f3c92b4a2ba19a370c49b9
f2a0b794eb7295d5b8c6dd276d5f24e10349594c
F20110114_AABPCJ lewis_d_Page_122.tif
135ee91864213166392c13055137479d
e4050f67d43c1c74b59dcaa36d76618701511464
39527 F20110114_AABOXD lewis_d_Page_138thm.jpg
341477377ad9389498180fd4b694a51c
867d23e6b2fd18f7b3bbcc7248ad8e69ed82bd90
73612 F20110114_AABPBV lewis_d_Page_044.QC.jpg
d668a556a1ffe749b47414508ad164f4
d4003224676b633edfad72661379a55a47efed18
F20110114_AABOWQ lewis_d_Page_012.tif
a01aebb8abda4c7daebd78ed4b8f159a
1051155918f5253dcefab548d4eda82f95d8f6cb
8250 F20110114_AABPCK lewis_d_Page_065.pro
5177001c470dbeaaef44446219c9e2a0
2a3261e9b382b7893581a35120a163ffa950d1be
80027 F20110114_AABOXE lewis_d_Page_066.QC.jpg
292ef5d2d16ae7aae9380c014dce8cba
42f7082ebfa7ea8dbe1d0bf808a5da64c7f3c41f
F20110114_AABPBW lewis_d_Page_023.tif
d13dd3a7d2a1d3f77799db9b0a90692a
955ece9770e4cde4e0d432a0014c8d455bd857e2
90523 F20110114_AABOWR lewis_d_Page_031.jp2
99f94916ca64133d66058ddc746996ac
e0b892db520d576602f47c306b914da243fbedf8
1051929 F20110114_AABPCL lewis_d_Page_124.jp2
8ae4681a07f23819e477600834656734
60107e5b362a025b403d58cfc74d8828f3aa80f0
44644 F20110114_AABOXF lewis_d_Page_073thm.jpg
3fee012fc33d7589834f25c330e73121
ffe7484a80e680f9ad2ebe4744daaa8292bc431c
10995 F20110114_AABPBX lewis_d_Page_003.pro
2954ebfaf63cbcf9e814718970cffc04
1b20037855f32b41554c73f57a62020d5a1ff97f
251132 F20110114_AABPDA lewis_d_Page_106.jpg
c28ed45aab242534310c0262490a4a50
783207af79f3a441f6aba97a74df68d1b740495c
62250 F20110114_AABPCM lewis_d_Page_059.jp2
92bec3c8ab241cc71d0a1b6b61c2b6a1
0ad47f0f07e2658e364181d9304f93d03dfaae9c
F20110114_AABOXG lewis_d_Page_079.tif
183591cf58bf2fd62205a5af39d12290
3c695b62edaf2112b276bc446a05648ac1109756
193874 F20110114_AABPBY lewis_d_Page_120.jpg
cceda24002fbfb6c111c5db46b0de212
42c8c8c31041955726814767f06253e2695a9510
F20110114_AABOWS lewis_d_Page_015.tif
e00e72378f1b441ff52c10537273c353
3c4bc60bfa60a06fdb96be2da593cc0b219820ed
43179 F20110114_AABPDB lewis_d_Page_058thm.jpg
9e0a02ca7af62ff845d36a983fa5c289
f1b0ab3787318649fd40e40585234e7a7b638a9f
1185 F20110114_AABPCN lewis_d_Page_129.txt
28a069bd269fe61b13adea8bb05c78b4
c438fb10bfd0a3be7ea740a2898bfb3bdb3d942d
1658 F20110114_AABOXH lewis_d_Page_056.txt
76cf40483450f73d8fb16f71ba6ff2b1
beaa4ab032568d4208851029b96535e7f8b0254c
56214 F20110114_AABPBZ lewis_d_Page_080.QC.jpg
6d59e60cb6899c750eb56d8b875a7216
61bc1b23f0947553db76b14969c0bb3a20563c5f
645 F20110114_AABOWT lewis_d_Page_037.txt
28766bb4d6cf4b20f2107bcc472284c9
3d27de4a36359f75c52b64ef886b094b9430e0a5
153256 F20110114_AABPDC lewis_d_Page_049.jpg
4bc7c283ef7148b4eeeb8206c693a00a
d3e86547380aeeada2c727e84c3dcbf5c871e5f6
506341 F20110114_AABPCO lewis_d_Page_142.jp2
34f01fbdb8213bbea3e9f9bbd185dac6
ad83250e7f0840419b5a68477df938d5503168c6
48707 F20110114_AABOXI lewis_d_Page_101thm.jpg
508a25dd262a9074b14a8d255dd4df56
a5b5294f584a93d78f39f57a555c69bc49029188
611716 F20110114_AABOWU lewis_d_Page_077.jp2
5fdb0046d7e4564e9400debebb89c074
fe7b4e67292acefe72b2c3c997e6f3bbe86d9885
234093 F20110114_AABPDD lewis_d_Page_095.jp2
78d6801522f055cffcdc6700dad92173
63ed4ebbad9b22a3be31e0c504485b8dd7d73a69
201768 F20110114_AABPCP lewis_d_Page_087.jpg
55c939236c301477e314651d0681db55
40afd3be9795ac2b6216985247f33cca86fec8db
60309 F20110114_AABOXJ lewis_d_Page_056.QC.jpg
503dc90964c3b50f77f75954bb87c3aa
dd3e32a143aab372ef1494aabd14ffc51c775ad3
F20110114_AABOWV lewis_d_Page_022.tif
5be820d6a97ae00c96488730fe2782e0
a8cc0581970048a5806a7270284bcb5b9524593c
F20110114_AABPDE lewis_d_Page_144.tif
bd4a52fb7c5fa13ff76383f224b250ed
4c771b84af3f2f0e8e0ba6430f07edacc7019843
F20110114_AABPCQ lewis_d_Page_094.tif
f122e91efc771f2b0e780ea5874df226
23f451752c369ee8e99437683a56d99f7f4aba69
225570 F20110114_AABOXK lewis_d_Page_103.jpg
79ac8e3d6319e8509dd8ec65359358dc
d1334383d67247266f9c8011e9ec77c8354c5470
196307 F20110114_AABOWW lewis_d_Page_016.jpg
8615178aab78ab961c04abcc875a443e
5338f8596c14999c16fce3030b6ee714ecaa4a07
796620 F20110114_AABPDF lewis_d_Page_073.jp2
60688ffb106d933c8ec4a7c6939e8cf5
f78f5f8389e956077c1d992b49252cd93332b83e
F20110114_AABPCR lewis_d_Page_107.tif
a1e4bf62537f42887f22f8eac71b0cad
39247d8a6a6f53dc3784cbca3651b29e9ec59529
46975 F20110114_AABOXL lewis_d_Page_152.pro
2e9df79b56492ef8bbfe455567b6174f
8313bc90d74fa2a86958378240f5686bf3335473
23763 F20110114_AABOWX lewis_d_Page_066.pro
c06c3331dcf694414910fecfc1bfc99c
6793fcd067f85eb3385bec8cdce77feee129e6fc
F20110114_AABPDG lewis_d_Page_048.tif
5cf0ed6f468fd44cc3cf4f0039d0ced7
1d1775d4bbf1adc3d4aea7e7499ae408bdd483d4
1836 F20110114_AABOYA lewis_d_Page_120.txt
d4c33d682d36d19145a473ac903807e8
dbbec18b805688a1941624f2d23a83c3b2c5043d
60037 F20110114_AABPCS lewis_d_Page_012.QC.jpg
cbe04d9152f8e38f676c2afa54ec3947
5eaaf7762a08764e329f8fecf21838f80c9d85f9
380 F20110114_AABOXM lewis_d_Page_065.txt
0057049331abe5ee1e64bd875f12abb4
31ef12d3ebc1de1fceeaf9d6ea7af4a67724f5a4
43524 F20110114_AABOWY lewis_d_Page_131thm.jpg
515cffc45eff797542bbf01de21b1552
40b0e33ea74997abacba16631957a6f45f056659
9160 F20110114_AABPDH lewis_d_Page_063.pro
409d209f2fe442afc6f85968f0a1b1eb
f87a63a1bbbec167464101c1b40916ffd19e321b
48335 F20110114_AABOYB lewis_d_Page_034thm.jpg
26307569371689ea2254e17c8b9283fc
2e7c3c358145fb8b710de9e46aef010b08fdbebb
10255 F20110114_AABPCT lewis_d_Page_153.QC.jpg
4db668b006be16a15ce23eeeaf737579
586f33728359d5065bad2a6e55caaee2236ef9a5
195704 F20110114_AABOXN lewis_d_Page_144.jpg
150e9672bcff2b3acc79a3a54547bc2a
6fec3c13b8bd11b9f7908ac9c139f7fa3fa02597
222385 F20110114_AABOWZ lewis_d_Page_034.jpg
89b2ac24d2184a3122c76bfc27a1b9c4
1240d18164731db9f1ce3d7e7e4bd4562123e1a4
78969 F20110114_AABPDI lewis_d_Page_080.jp2
01182bbd45134bf0e32b4d09bf6bcb25
517243f5b7b3cab9caec6e13cab3a9b6f7d971db
F20110114_AABOYC lewis_d_Page_087.tif
da265bf953a1327ad1f75280125e0758
9d3650c204cc8e8200f0c1f8848eebc29061bc66
104000 F20110114_AABPCU lewis_d_Page_079.jp2
b9b97672b8cb08c79b5e6b2659fa2222
5ae2f2da8ff48707a5fd3319c33e993ce2a51b2c
19088 F20110114_AABOXO lewis_d_Page_056thm.jpg
c1782d54e33ccb3887c9924b86b14ddc
b0a2b8473fdaeaa101b5db7a8ed2f430e1343cad
71975 F20110114_AABPDJ lewis_d_Page_039.QC.jpg
27d34ab205fac361e9e0e340fdc16dba
ef03fc083c3b0cc4927c12abad4a1cf9d92aaa1a
35702 F20110114_AABOYD lewis_d_Page_065thm.jpg
6cdf9886c2f7acd5409a78481bd07724
cdbc3c2d2998a4476d9bc389021c2ce784cab939
90388 F20110114_AABPCV lewis_d_Page_018.QC.jpg
4772a0e6aad9f88f921c17e1e6f294a2
48bbc578c5762b15187c39da215f185e1fca64e5
77303 F20110114_AABOXP lewis_d_Page_006.QC.jpg
1f0beed9ded07213f352d5ade6f1ae5f
bdf950ef45ab2386c7353bb44073ebfbba418bb9
49620 F20110114_AABPDK lewis_d_Page_060.QC.jpg
e3fc251ae8a460fb7f942f414f13b850
2961c94d11acdeb86396ee187ff64dc99b327a00
41286 F20110114_AABOYE lewis_d_Page_049thm.jpg
05ea3742efc5997472872c68f334fa48
9acf07922a19008bee987d2441e43851fc13a575
180507 F20110114_AABPCW lewis_d_Page_013.jpg
4f10a5dd7d4408d9771b43d20bea7018
28459d821c852fe340b7695b3a1ed08dc6b2fb0d
1028 F20110114_AABOXQ lewis_d_Page_040.txt
e8c1fdd983015725a2dff39a95cb2889
ea2aa7fc4767796b4bfe4e57d22efc744b40be88
1739 F20110114_AABPDL lewis_d_Page_025.txt
ba1e3d184ed25d78eb18110c67fd18a8
4316b000634c533e630784a2ab385ea8ec5c4bb0
F20110114_AABOYF lewis_d_Page_017.tif
484b0c89c6563af7bf927bf272603c85
b7fa1233f4875549d090e37b69f320dfe8608b9f
1524 F20110114_AABPCX lewis_d_Page_017.txt
5ff469c7c5f4c5d7dc09c9eb5a1937a4
30306c2a8a370272ca5ae77e49a11c1b51f4d109
41154 F20110114_AABOXR lewis_d_Page_062thm.jpg
6d17ada478a1d45133d269647b009514
a9c4f86e42a6d4b1d5944c79ff658a5308ec2a50
1675 F20110114_AABPDM lewis_d_Page_106.txt
e436e81a930487f2e3eaa6a41ef01c6e
1a07f569908552d829f9be80c83a97ddf0d9bc96
932604 F20110114_AABOYG lewis_d_Page_112.jp2
535f992f5ea321722da2276a112a5306
7da0a0d437c8daeb3788bc4eb0cf8df8beb5f47f
90680 F20110114_AABPCY lewis_d_Page_107.jp2
962ec1ae3e5bfc39526598c72018c779
1bc0ef72bc26b897072d3b0d22cbd9f0efc02641
52713 F20110114_AABOXS lewis_d_Page_010thm.jpg
756412938288e9869fd44612cc77cb26
845eef506e950bf08bffdce5750b47d574c7bd36
63813 F20110114_AABPEA lewis_d_Page_096.QC.jpg
7d5cca5f5cb97ca3e2393b5fdee58a5b
e818758e4f68713ed7d42cd2f13745ea41a41432
1070 F20110114_AABPDN lewis_d_Page_002.pro
dbe7ca0f2c0aa7532ebce9c8c79e5111
f1afc5d9ff8d36d9ca15b1625d2f5fc7a394a108
2424 F20110114_AABOYH lewis_d_Page_104.txt
437e7ea197340965038d224bb396864a
65888186a859f83a1425bfacd4ff3b1efffa286a
1663 F20110114_AABPCZ lewis_d_Page_057.txt
516c672b682752692eb8ef90f5303f83
418a5756ce40b77ac7ffb6ba43235f5a403a2868
F20110114_AABPEB lewis_d_Page_105.tif
b163d21718ba4c421f49d5535956944e
eed5645fd7c1a310c0f337aa5e52f408510054c5
32465 F20110114_AABPDO lewis_d_Page_151.QC.jpg
8db742fcc660395ac2e748bb4820f5ca
0d48fc80c7e360d9f50b6bff71aaf1fe1cd63608
1393 F20110114_AABOYI lewis_d_Page_047.txt
7e4ea42fb553bcfa67307578f2bab6e6
296c142058bf4b59a02d067365493ebc2bc90c86
42614 F20110114_AABOXT lewis_d_Page_112.pro
b3d43c7aaf638bbf95516e736f62dc53
b375af692f690f48dadc2b3f848c2c1481b8078f
128943 F20110114_AABPEC lewis_d_Page_138.jpg
3a8ae62daa0249f2b39b27c632414fc3
0063718db28fb82c63dc51b5968cd86063fc25d8
38840 F20110114_AABPDP lewis_d_Page_122.pro
32df3ed6ccf8cd93d27622d07035fa28
3a4edfb5d70ff8ecd951146de8091d3aa366affc
2495 F20110114_AABOYJ lewis_d_Page_007.txt
783943fd832db94dd765320f6e440caf
c0f3968cc4cf28725ecc71110b164135212d049c
979 F20110114_AABOXU lewis_d_Page_044.txt
990aa22a9a725afe777aad06691bea6b
6f2013046165922b0723c0194f3e395912081548
1051981 F20110114_AABPED lewis_d_Page_028.jp2
0aa523826280308d85cc93d9a9725806
01db6265dfd1540bcad59d10204b95e3c2d73020
165986 F20110114_AABPDQ lewis_d_Page_046.jpg
3ab9e6fc6154c37e4653746f62e65dce
58b2bd9ded5c7342e290aa74b4c43d048b6418ad
128649 F20110114_AABOYK lewis_d_Page_069.jpg
02a78b57d4f5c5ff0d2b58ae199e3c93
247e24c349ba8475af3eb3303c4d0d35916843ff
F20110114_AABOXV lewis_d_Page_041.tif
dcfabce06ddaa82896d3dc9dbc8ddafe
aaa29d06d2f7a5cd168e1c81abb59f9693d841d1
47137 F20110114_AABPEE lewis_d_Page_033thm.jpg
dadc90dcdd15e53b46515a65ee881d5b
00d5870f4661981ec951844333bdc04321e09f52
43555 F20110114_AABPDR lewis_d_Page_070thm.jpg
edf65f755d34bc8aabd41a4a02d7ebf6
6c1aec98a648ae575f319840bd4fe9ec72cdc147
41426 F20110114_AABOYL lewis_d_Page_039thm.jpg
7f91dc0f3c744eea85baa176af3b023a
102fb8dee66f65326c27b88cd05f9e580d6d8ee7
4681 F20110114_AABOXW lewis_d_Page_143.pro
3b1120da0b49ead7a27f4738b2c3f9cd
143aa2f5e9aa6233abad3db3378e404bd5c5cf36
F20110114_AABPEF lewis_d_Page_146.tif
4973ca6ab3346d6ada59c8f937377e29
44be45882e64ac88cf1401b3c5d82505e11ee654
729892 F20110114_AABOZA lewis_d_Page_139.jp2
428f8a30e7a91323958f6939a16cad3b
5cf0a6b04b5a745b332e19ce8cf2d23dba6a474c
69274 F20110114_AABPDS lewis_d_Page_006.pro
5d480b8cfc4d800aa81704bcb58c5bd8
c4dde486c63aa7d1752b58e4730048957dc0cfac
41721 F20110114_AABOYM lewis_d_Page_030.QC.jpg
da8ecaf541deb651e2c55956bb824c62
d6b9118cf3278be5b3964e45e2f5ce6c514848ea
F20110114_AABOXX lewis_d_Page_026.tif
5a8fa2f153fe4b60ca8f610a1f9f223a
d7e6631d5cd04ccf53be3c70b267b6ccfded4a8a
43665 F20110114_AABPEG lewis_d_Page_038thm.jpg
e376dcd152376593614f20c6f6699458
5fb171f899b9ad58ad0a8af5dc6ac7096baccbd1
1982 F20110114_AABOZB lewis_d_Page_148.txt
d55ba35fdbdf1224838932324564a665
c81850ff12ca00ce210260c568c6c86749e19b9d
2302 F20110114_AABPDT lewis_d_Page_093.txt
eacbd321774d28cb0509b2cdf8268ae6
177ddbd41dc6560f6a4eadce694d4d5e1b352dc5
207862 F20110114_AABOYN lewis_d_Page_130.jpg
588e2b78ee88cf9c4c3e64ee4cdbc733
f0b8f3d82e2fca27720bb03ab2fb5cd0908b6066
F20110114_AABOXY lewis_d_Page_137.tif
c2e9dfc29efa93c24612090f4e7c91f7
d1db42ce63b1b818d518e52204cff88364a0092d
40439 F20110114_AABPEH lewis_d_Page_119thm.jpg
6e0d2246f6bde17c0a97a55130773ea8
cd89633959426a29f102c9ee8508b07bd690d8b4
F20110114_AABOZC lewis_d_Page_002.tif
e115eb46cfa414cd45643f996b72ac61
637223a5e23bda367e59b8cfaed4c0f5dacd30d9
192922 F20110114_AABPDU lewis_d_Page_124.jpg
ca0a99bfefb6a90c35e9f085aa841f6d
c38ac2de643323087f713f651881ef7fea254e69
F20110114_AABOYO lewis_d_Page_140.tif
299514ea6e7bf2173d1a96db09966a60
905b9958d6e4e9ae3cd3dd34232431d3053c86e1
F20110114_AABOXZ lewis_d_Page_083.tif
dc4f0fa6eb443c6650b3b31542c02295
58b36704de535be0b5340f427aeecbb4dac1ada9
20987 F20110114_AABPEI lewis_d_Page_082thm.jpg
24789c0ec8b458ef5b71cb016593b929
4de053f542f32bd2eeee50ece5594bcfbee7bb65
F20110114_AABOZD lewis_d_Page_025.tif
39b4be35732cf56c526ee6836053ed78
7bc8c93c4439e06bfc3d789cdf10014cbd8e74f7
736 F20110114_AABPDV lewis_d_Page_039.txt
0202e3bae3bfdd46621a3176882f07c0
c1ce9bf9014f22198421eeaf48fcb1669c8c42c4
42552 F20110114_AABOYP lewis_d_Page_006thm.jpg
adc030d160b6aa192bcf418df0885fd1
8e007f08c8a8a638258102520114c4f41e23c1a0
F20110114_AABPEJ lewis_d_Page_009.jp2
799e0233883c6855fca498c83ead7a75
ebaf78b40f22b7ae4b8f64ae1da2bc8a1c28914e
701325 F20110114_AABOZE lewis_d_Page_021.jp2
f8b12178250b042eb9b594ab82e1fc5e
40cbbe9b0db74cf12fde7dde80764840b034ce7e
171809 F20110114_AABPDW lewis_d_Page_139.jpg
4f592fc3830a1e1797bfb35a31ecda87
b4901d0ff40fd75a0a28a195c5ab54f789208569
55204 F20110114_AABOYQ lewis_d_Page_148.QC.jpg
3340e260c51236d8b3f1a4fcfdf7ed27
03088b14991affeafe5f27939893dd98cbc952aa
82253 F20110114_AABPEK lewis_d_Page_014.QC.jpg
53b8871c632a9d451bebbefd4fa13248
b9893e9c73611a3416c46616bcda1b1fed955d34
F20110114_AABOZF lewis_d_Page_131.tif
ae8bd30fc5ed845d71111ff89a3d872a
f415f23807598a988810dca6d43315e5cc426d4f
579162 F20110114_AABPDX lewis_d_Page_138.jp2
869f0ca28c37f5e4f8992b6c17a71942
7dce3edf2d6c7e7cdb397ca21a07cefc5f67278e
48578 F20110114_AABOYR lewis_d_Page_151.jp2
234a4c2c8af43948f26c3ac34343b07a
0d2c54118b3ad693a7044378970028c2639e24b7
F20110114_AABPEL lewis_d_Page_084.txt
367518309806b790fd9aa1fce509f283
fcfb0068975f0bcc243e522e81eb25ac0b1f5919
74833 F20110114_AABPDY lewis_d_Page_136.QC.jpg
abefd0805f5aa30aeb453ba6ae89daeb
e78494abab9bd9668f067bc6790bc76450a23878
783116 F20110114_AABOYS lewis_d_Page_040.jp2
e18cc7545069d27792a36cfed7468496
13d1893528907a4de051249c5fd336e57ca38905
46640 F20110114_AABPFA lewis_d_Page_035thm.jpg
afc3ba151b9ec617384750547040ce5b
e05548ed92362a7826fa1c0909a0fa85a876959a
F20110114_AABPEM lewis_d_Page_029.tif
f87c05ddcac08d1ee3047960df61e90d
fe61bc6fc1b6deb4c24730ae6dd934eb6255deaa
167320 F20110114_AABOZG lewis_d_Page_114.jpg
6bc3786d5ab3fe8d7f97afa957dbfa61
746811c7fd921ad7e8df280b83366950679377b4
35490 F20110114_AABPDZ lewis_d_Page_143thm.jpg
24b6cc384874c59acca988f6dc21cb8f
ac5f042c8b725f4b48cd019d2bb767b3992d7ddf
16735 F20110114_AABOYT lewis_d_Page_069.pro
34543a61fe353a8c3e0e508855b4d5ad
688ea6d2c0996734562b143799ecf9af649cca52
1051774 F20110114_AABPFB lewis_d_Page_043.jp2
76451de2df6b175a40cfb67a8c4b12e5
bac71fb5c9792b18a0bdff00d6362990d8ec9b70
173980 F20110114_AABPEN lewis_d_Page_117.jpg
0fb3c7c6741e506099df8c03cdd2c4e4
b6f9ebb11955876d8b2fec0cc1056ea92553134e
34099 F20110114_AABOZH lewis_d_Page_028.pro
f3a24413d7d9b3b2232a61fcb05c6e30
4febb53e839ca74b95c178607a86559e992350ae
1051973 F20110114_AABPFC lewis_d_Page_109.jp2
105f0f90bc52558db8602309c1e71b08
eb55eca55dc1aeaca87b3bb3d543e4630987b2f0
34007 F20110114_AABPEO lewis_d_Page_091.pro
b7ef909c3cb086e0cfbb79dce7eb075d
d5fe5eedcbd551aa4deefc574845ed98c4f667e3
205447 F20110114_AABOZI lewis_d_Page_149.jpg
bff1f49e2ecbf95b06591048c142ece8
1d91e98a6db343564bab28e67db79471aeb5fc73
67411 F20110114_AABOYU lewis_d_Page_064.QC.jpg
9efce06ad13f08d3d39b1a0d516c788c
a3adfda91a896508dd449defeddc62e521113304
56200 F20110114_AABPFD lewis_d_Page_023.QC.jpg
06f9c4a77a396ba8df3b4ad3e22ebd44
4df671d2636d97f74d88c93eeed234c7aed8b792
30675 F20110114_AABPEP lewis_d_Page_117.pro
5cde4917958cd0f8874c560f78cfa2d9
068aea1d45c26347f120577e3468c74071fa7a8a
38307 F20110114_AABOZJ lewis_d_Page_017.pro
d2e2b9d5605d6c820623c6248a714ea8
64421ab820c6b51206e046412a82bca9d2c83604
49840 F20110114_AABOYV lewis_d_Page_075.pro
187700c576b4612e7676aebce2a31fbe
1b529b27211192213d49557441d949b936f33331
1051857 F20110114_AABPFE lewis_d_Page_029.jp2
7b4f0e74083da8b0c05699365edbecd1
8e6f953635974f7d054b28cbf479b90552785a51
F20110114_AABPEQ lewis_d_Page_136.tif
1a20a177bb654d13ab5b578223e7f440
1693f1bf4a47cf6ff6308e11246c64bc51a497c8
175580 F20110114_AABOZK lewis_d_Page_066.jpg
000d11ea1c0ba840a804c6a026f9ce60
3656630c58f8d986ad2092688893ac9ea5780d71
138531 F20110114_AABOYW lewis_d_Page_041.jpg
5e16be90f64078eb50c532fdad8fb033
22773797c0208fca9c1d21187f49356693b5f0ae
1714 F20110114_AABPFF lewis_d_Page_094.txt
8de7d6314176c8e637fa8d01b84409f8
7ad1a4b905ffec3c5719337f8d36a19aac51ffcd
47246 F20110114_AABPER lewis_d_Page_028thm.jpg
9b2d23fe8fdd934cee26c16c78724b2b
0fd0e510c3bb7fb0f089e264f53e61c52c84be8a
86138 F20110114_AABOZL lewis_d_Page_148.jp2
3070b25074fd3d2875f8463d6cd658b1
bcf7bdfd0c656423cc79c0d905a1ffba16b6eef9
7094 F20110114_AABOYX lewis_d_Page_154thm.jpg
c533945f1001b6aed151b45e3d8f382d
7cec0f971c2b9845255abb0ac8f6899155213256
37454 F20110114_AABPFG lewis_d_Page_080.pro
33a883833fbbc341c912e66194e49808
d50a9d7dae5808fd9a698c0a4459d7227cf619ec
63293 F20110114_AABPES lewis_d_Page_084.QC.jpg
8ec0320225545ef65cd78497c0a5f7f3
adae9f472ff2237b6c18626ca7dc15268ae92d9f
F20110114_AABOZM lewis_d_Page_035.tif
24ecf72c9091687941c8206f26732025
a5575aeb329ac662e4911a0bbe5b40af9ddd5b55
F20110114_AABOYY lewis_d_Page_113.tif
f8b1cb5bf2eb5d95f655ef3d01a351a3
b05c467db74a1ab30c64e3bb03c412a8e4cbafa0
F20110114_AABPFH lewis_d_Page_011.tif
3111080327802496ff12a18242a92416
15a0e47fb56c310551c870c2e5ae5016a84ad5f9
212188 F20110114_AABPET lewis_d_Page_108.jpg
5403164cb9ec90f820807afe4d2378a0
fe43f284f0640f4748556c7ffc3329001c0f784c
87150 F20110114_AABOZN lewis_d_Page_118.jp2
847370b7fb95ddc81ef357dbc6c9a8ec
cbafc6eaa95d498ad2245539ee1b3a2be0d3b318
95199 F20110114_AABOYZ lewis_d_Page_102.QC.jpg
f60039b796f2db55e954a8cfa7d92a8b
2d2b6aecaabd7d8311e1542fc91ce4ba8304d9e9
978 F20110114_AABPFI lewis_d_Page_074.txt
1342b4f8974078e80a525673dc71a7ec
98922d40b3ad8fcf0b7f7ef45eb4dae5dcea70c5
1646 F20110114_AABPEU lewis_d_Page_078.txt
ce8c025a26ccf3e58dc193bb3697bedc
d7b1e40e85acfcb1673d2f89c62b549c1f9dff29
17355 F20110114_AABOZO lewis_d_Page_077.pro
1b53fc8ec2f2c7f8acb317c77aedaf4e
ba23b9ff09ba9cd4eb4ad9742a779f1b02f4ccac
F20110114_AABPFJ lewis_d_Page_100.tif
48479f48474c632833e1ac8b625ff3a6
1201bc7a30366a38ac92dcf68c75ad8faac0ba60
101030 F20110114_AABPEV lewis_d_Page_075.QC.jpg
6a392db590df6791d4650913f7b93452
34856e38a02c7f19d912c18ff86059611314faed
409186 F20110114_AABOZP lewis_d_Page_063.jp2
c2bb9a7b9d4e520216b36052fd1b78f4
6f2572ecf7e69b79fbe37b332dfa42bf0ff376e7
58662 F20110114_AABPFK lewis_d_Page_150.QC.jpg
49ea4793ebb9dbc4a6341d17739fd126
fef61a62980af55fb9c2b0d1a6b6e57da30c2fd9
24679 F20110114_AABPEW lewis_d_Page_130thm.jpg
17ea442b61fcb17d40c6363113634d9c
4cbd0a9caf49ec3c655cb9f62925e1668b61b121
226760 F20110114_AABOZQ lewis_d_Page_143.jp2
ba832f46e683bd937bd360eb89959cfa
643ef4b4df12dea1bf8aacdee16a3735d995209a
22012 F20110114_AABPGA lewis_d_Page_111thm.jpg
6e9553ba62b8d3c22c609dcc745f635c
a4567c9d407294ceac96eeebd3d72e9c3e1ad407
102261 F20110114_AABPFL lewis_d_Page_106.QC.jpg
8f15dc894edf1f3150295d81ef94fbb4
71a911c3e2ad91347dd1d241b279eac5536fe07a
25663 F20110114_AABPEX lewis_d_Page_099.pro
b8bc113321c7a33dba342715e5d652cb
7322e5f1863e75c7ba1b1c1562f92cc09f650dff
1954 F20110114_AABOZR lewis_d_Page_123.txt
4a12c4882ec65599aec8ccfb9732edf1
d8b9cb9614498d282bbc3c2931cc1f17e82342d1
173321 F20110114_AABPFM lewis_d_Page_096.jpg
753b742e2163f4bcae6670a81ba8f196
00d8b23a6eeecb498ed1c47bb9e5bf8c91767703
47251 F20110114_AABPEY lewis_d_Page_129thm.jpg
c7d3e9ae2483668689e2bc04683a2913
6b34b94fea6a43763a31baaf0aea6cd9196bf358
211239 F20110114_AABOZS lewis_d_Page_146.jpg
57df1ab872781be9261bc943d94558a9
220d62bd9c1f94de794e7e2daf6bfdaffed45df8
F20110114_AABPGB lewis_d_Page_142.tif
5e5c9fc20b4732fa1f9264f872159ac7
d1103c2b9aee08c777e7ec07e264857494cfb62d
3830 F20110114_AABPFN lewis_d_Page_153thm.jpg
c2a002895867a046b293b2283e82b3d0
1901734c96f78c46771a9ab3d8c30c00a08a81f0
F20110114_AABPEZ lewis_d_Page_078.tif
7668f1ab332a7021a1b43c48334fd564
cc69e6466c09b041cad6fe4b7069065eae35b308
38087 F20110114_AABOZT lewis_d_Page_035.pro
949bc12e69203f504048bc8378e0f281
b1c1c308a5585e89a74cdc1577da61fd27fa2ae1
27551 F20110114_AABPGC lewis_d_Page_147.QC.jpg
8090912b7c128114bf3f456d0133f485
105725b9ab4722caae7280c5c52a4a6b255e292d
F20110114_AABPFO lewis_d_Page_099.tif
404c85843bafd1667d88c47e2c622aa2
a613b45fe497617fc4cf95c286c4ee7bcf2da368
F20110114_AABOZU lewis_d_Page_125.tif
8fa8aed99274d0dfc2a3bb58689604b4
081a6364ef236b111346b69d973db36492f8deac
74403 F20110114_AABPFP lewis_d_Page_045.QC.jpg
fa8b6d6e9b719fdf5f01a778faaa3647
5050cc162ba3e69ed3c680c81e5605edb497edc0
672 F20110114_AABPGD lewis_d_Page_067.txt
b8f6e73a63a6303ee16cad71c6780bc0
276dd66e49ed7d1b033ff037e9c887af0862ba7e
36584 F20110114_AABPFQ lewis_d_Page_078.pro
2514dcce6ab68653edac270d8ea5b42d
081bf96ad6fedb7e01ff692bd0e3f9db3fa12306
73482 F20110114_AABOZV lewis_d_Page_086.QC.jpg
eb6311640afb9f8733d5767b1ae73bf6
42beaa5ea5509f79a03089b9d5ea1bccf5210490
26929 F20110114_AABPGE lewis_d_Page_139.pro
a2cd58b2fb0bb3d1413eb8278ebd91f0
ff24f9a51beb3bd18780d2ae8058fedbed939212
53418 F20110114_AABPFR lewis_d_Page_030.jp2
c41eb2e3f3e98352963dc20b200ede4a
5d252bc739218995e6ce2bccc4b20e1180567b29
19191 F20110114_AABOZW lewis_d_Page_025thm.jpg
95994b96bf859d3edb9c79f0a0787b65
10fc990a6048eba5b830718270e45697da0cbd8a
38251 F20110114_AABPGF lewis_d_Page_053thm.jpg
9891c5c40ca3dcffa3917099a5ade57e
75673779dfedd386d259d2679e31f5a13f90cb35
90167 F20110114_AABPFS lewis_d_Page_004.pro
040e7f439d963bb948f1ca44f50ffd0c
2317ae938a524a9e6a2ca82d93fed389628b8b3b
96875 F20110114_AABOZX lewis_d_Page_028.QC.jpg
6774584b87e18fffd14faf3084726afc
c1ac82e4f39a60d9830073c5555989c8ecbe2344
F20110114_AABPGG lewis_d_Page_019.tif
c8ef38c4597d6920e3d2e7fbc00996da
213032bd5336a2a219480b00134b65b9fe811d86
596982 F20110114_AABPFT lewis_d_Page_137.jp2
209994fde694511f170b752e3986680a
e5601de6e81428c64b58cfcf4d72fb4cc2da3a83
29973 F20110114_AABOZY lewis_d_Page_136.pro
0260f12cde12df6c1fef57be3a5e91f7
fad0f97dd54161aed11f96ff2c4653b30799fba5
F20110114_AABPGH lewis_d_Page_129.tif
4ef55a84e0862a0ab70de870ff0c0f89
03d362ff7ec2c0b27b03cd977c5dc72fb3488d05
831330 F20110114_AABPFU lewis_d_Page_132.jp2
822f50542ff26ac00629df38f4de8389
294d13381b85cc74c9414c5ba4e8a152b2419395
69929 F20110114_AABOZZ lewis_d_Page_081.QC.jpg
1b6c1610a704e77b41bf35ad08a2a9e0
28c393eb5af9107ff26202dfcc4a743055a29a7c
11229 F20110114_AABPGI lewis_d_Page_060.pro
a2b6df34675d1b4c6dbbd816e1b6d530
d759ae2c05a7b9d20794519b8279b8634f4d6fba
1839 F20110114_AABPFV lewis_d_Page_076.pro
e9bff738866c351148799873ff0be155
aad3b249638c15d03fde1b325c5983de93efd413
21411 F20110114_AABPGJ lewis_d_Page_134.pro
553bee5b6d5f8b312bace48799ad768d
189537f2bf3524ead90d77e8eaf897e770d2a241
177609 F20110114_AABPFW lewis_d_Page_107.jpg
29a1580015cd4f9a82eac96aebcdd3ed
9116ba71b8e7c4006e281d53f7c94faab214d0ea
22339 F20110114_AABPGK lewis_d_Page_054.pro
b565c4df10521c3cc71ddc4bad8d73c2
bc2e51b3b04a4d3f6edf1a9d934c43656adb4db1
1935 F20110114_AABPFX lewis_d_Page_130.txt
046e88ce6cec8ca961e8a92c9a547b20
a8f356e59ec4c80008b6e4612b7ca5431ff3f2b0
26979 F20110114_AABPHA lewis_d_Page_015.pro
0075d578a96862793f2d4355fcdc9b1e
2a00237f45298bfb31b4ec8ed9a2b2780b4f2a08
F20110114_AABPGL lewis_d_Page_005.tif
6b177f53e66751bd8d70e87e9b270169
0a769f3add95df82b15b5e0d139a6dba19554a7e
75724 F20110114_AABPFY lewis_d_Page_105.pro
e18b18d1aeb830514cfc4ec3362d9b5a
5f5082ab6267b8a592a1b1bc014a82b237d84b80
62839 F20110114_AABPHB lewis_d_Page_138.QC.jpg
537a1fc8892e8bbff989cc64b42a83f8
0e45cb9902916210bdeba9493df2288219250ea1
1832 F20110114_AABPGM lewis_d_Page_072.txt
d47a191b5faa677b7795dd6fbb7c0dcf
b811c22bcf2dbbff3d846c80c6807eb831e875d1
43331 F20110114_AABPFZ lewis_d_Page_124thm.jpg
ffa2679298252516528972c9d393e092
b13a52671124ff3b8350be8d17fed5ee37cffeaa
3415 F20110114_AABPGN lewis_d_Page_145.txt
a5f258767ec125a3901c13d3b005254d
5af146ec50cb7d3478283db71a8a7b5526a1b27d
2964108 F20110114_AABPHC lewis_d.pdf
2fb5fd76483c16e257c88d2fab56afd9
795b169aebd65f403f779d13af422e4e20f7c3c6
220097 F20110114_AABPGO lewis_d_Page_110.jpg
6dad479dd33ae78dc16c20ab241281ac
aac7a29a265e22990bf777b58f2b1e56eeed53f7
122045 F20110114_AABPHD lewis_d_Page_068.jpg
ee19375dc978845af6892b70d80c7999
6b6254d35cc9193acf8fc91746955611a0f5b1d2
197381 F20110114_AABPGP lewis_d_Page_109.jpg
89eedddd425fb64e43fae738c6833468
d09766e18028ec9412b4dc663bd101f0a4ee8698
89726 F20110114_AABPHE lewis_d_Page_099.QC.jpg
0c40af527ec55fd09db94bd9db3bb8bb
243e2b02b727a3faadf29be83d083c7485c847cb
1042467 F20110114_AABPGQ lewis_d_Page_131.jp2
28aefa4f792e922f4e72a9dc8b6cfe50
7af6f878b48bdd663cda07d5b1cb82aad260cf5a
75967 F20110114_AABPHF lewis_d_Page_123.QC.jpg
34c5e2f50656b12d25f865d1f2fecd38
d55f07ccef85c0ef59325388ad0994e2c17185d0
246504 F20110114_AABPGR lewis_d_Page_024.jpg
0165d1d5adfd3a71a9295c4c4607a60a
72d8fafeb2d95dcf1abb5b8055e0583dcd185627
F20110114_AABPHG lewis_d_Page_060.tif
b9b8335de49d9d159a0cd16fe9a13d26
d029575126507e984c578371931eaf6ec76a4437
26206 F20110114_AABPGS lewis_d_Page_105thm.jpg
b38497379f9a12a5fb6a4b03a3da59f5
4331754ee689cb64f65c8fe9c1d7e81a78bd7764
38310 F20110114_AABPHH lewis_d_Page_034.pro
85c50f543e41a17f7f5241e800df98cb
78c72289c3fb3c19f564e9d77b80838c16aab3db
F20110114_AABPGT lewis_d_Page_036.tif
f4172b7759f771ee17891c6cc8e8a781
0fc17dd4fd1a4d9b0d03a16c439b0f1482cb8db7
F20110114_AABPHI lewis_d_Page_034.tif
fefcd3d2c4059b0a40db6b069893e06e
79e731c4b2e5040128cc50e7a9dbcf11977e27ef
159859 F20110114_AABPGU lewis_d_Page_022.jpg
6ce5f54cee5f738bb6e110ee2684ec3a
45edda2e4c2d2a6375545fce30877b844a86ba36
26497 F20110114_AABPHJ lewis_d_Page_058.pro
29d57031867be94097ff46e38bb4351d
720b81502c43e3ceda6eee802e86a5b40db6baef
15183 F20110114_AABPGV lewis_d_Page_059thm.jpg
a4a183a618c9416bc1c04b91591bf8b5
78b43cc8db3f82e11f2856198e91a4b0f68a8fdc
70607 F20110114_AABPHK lewis_d_Page_125.QC.jpg
94eeaac47206c24a9bfc48cc4c62a10b
ce23732c984b4f8154712dd1d8012d8268ec777e
40477 F20110114_AABPGW lewis_d_Page_021thm.jpg
42edffb99c538e3fc0865c2dad3890bd
363434724285e9be35914a2d39ae25a4501ea791
10156 F20110114_AABPIA lewis_d_Page_128.pro
b213b430bfe8d12c8f1f75641b2417ba
e5f22ede6cdaa2530eda11a0ead1bc0258bd817a
F20110114_AABPHL lewis_d_Page_077.tif
cde2e2bf9f163258f8a57248df23be98
dc77c4e562f2c3b37c8281680851bbdd1072af73
F20110114_AABPGX lewis_d_Page_089.tif
3062051b6c5ccdea53e06ab911c47cc1
21b74bedcf82a666ee5ed44eab0729ae9a439502
1669 F20110114_AABPIB lewis_d_Page_107.txt
ac4101142233c34a0941adb19eade088
47f0cd596aa5ef83679afe17d690402fc5832cf7
229238 F20110114_AABPHM lewis_d_Page_101.jpg
d19cc972a8df0978347eccaad25572d4
6e194bbe3e4a8f69e3d9ac2bf407981525825027
14269 F20110114_AABPGY lewis_d_Page_002.jpg
9f054657953c88ebc6b3e8b8b8228315
f60f867ab53f87d8eb792f1aca967c740d007417
77146 F20110114_AABPIC lewis_d_Page_079.QC.jpg
859cf3c47f6b76e8f07a280db10d65a3
f8f1d50b106cdd73868570a60744cec05cac3b88
1051958 F20110114_AABPHN lewis_d_Page_101.jp2
0cf5f9692b233e105ec8c71c124db002
eed6ef7ac62ef797d2291ca06bb9996f437cc0c9
43485 F20110114_AABPGZ lewis_d_Page_043thm.jpg
a6ebe5799a81d78142e999caa3768ef9
70f59a4604dff3d161ac3c0b1e60eaf264817676
24153 F20110114_AABPHO lewis_d_Page_055.pro
a47fae5597895647d7fab954a1362da2
ffc3f69991704b6238f51dfbfc7986cc8550dffb
19886 F20110114_AABPID lewis_d_Page_133thm.jpg
eb76d3abe97ebfa747a9aa533f1012d7
ebbee300bfa6a70fa49726dae82330fe58c197aa
17396 F20110114_AABPHP lewis_d_Page_115thm.jpg
335df25860aafa612ab4f96ecc5451cc
e4e177a3f90c883778e1c816c6c538bf91ce797e
1051975 F20110114_AABPIE lewis_d_Page_007.jp2
3e5f24464fd26cec6ac1f9127525c86d
ae9539f16903fcd947d25dda9d9f7464cb2f2b11
114702 F20110114_AABPHQ lewis_d_Page_093.jp2
9ae358b5698d7e86bcb0348b5e411bd2
8a13d9e02625d638abe42693feadedd12dacc18b
73617 F20110114_AABPIF lewis_d_Page_055.QC.jpg
3416d2c7fe3854ed19d018fdd9424378
13844558eb0b0014edd572a024cb35add3c11115
155124 F20110114_AABPHR lewis_d_Page_023.jpg
3baf1a356a1f53d4c7838237173d4411
1f7a885e1af7a92e1d1e355837c25c69be277973
799881 F20110114_AABPIG lewis_d_Page_098.jp2
22b0c273941510c056bf6a1e485fa91f
1cccbe708447bae99590493c450e42f1a5071cfa
1345 F20110114_AABPHS lewis_d_Page_058.txt
049e8a3ebe3f6e18efc06de0ff68bd1e
8d4200223ab349d6326a4b67985cfb3d3996a492
84580 F20110114_AABPIH lewis_d_Page_015.QC.jpg
b69ba2e1c8aeee7778e3d1746b72f5d5
1363a9e9303aa03b43f619a75106b8607dd91a11
146749 F20110114_AABPHT lewis_d_Page_125.jpg
b7b65c263de36004c5afe79bed74e9f8
c44a27c5dd953641f84e88d6bf44cd7b40417d83
1051964 F20110114_AABPII lewis_d_Page_004.jp2
f3ff8b2ea759edaa7b2c590c2e8b3b1a
09bf343d46606577a63b30056c94e84b4938b356
65087 F20110114_AABPHU lewis_d_Page_092.QC.jpg
b8ce2aef7806be93ac209fcbfea5ce59
06c9bb83dffc7984e8b935594512bf0c424d8853
1578 F20110114_AABPIJ lewis_d_Page_090.txt
b6e096576089429fa7be1d3237895211
9e9774d756b29ef3e230c6b8299daafa28587020
60324 F20110114_AABPHV lewis_d_Page_104.pro
05710e6ba28dbaa7e84a0e631a85c3a1
32f4900993c05e17597d8e766479b75a4c4d83fc
77815 F20110114_AABPIK lewis_d_Page_146.jp2
2bd8bba3bf9bdd3948115fd40a0974a1
58a1389ea177b7ec0943a8e24cbf8682ae24d307
48504 F20110114_AABPHW lewis_d_Page_005thm.jpg
1aa0fc515bfca8c2761c123db02b4ae4
18d1f751df96277309a5b198e0ea8c667edbbdc0
1009 F20110114_AABPIL lewis_d_Page_069.txt
975c6bde610f6d5914c74ee9c74e47e6
f708b8508c698a2487b5b935f5334f86c1ae4a26
F20110114_AABPHX lewis_d_Page_044.tif
8b7697d6931ef533cc5ae0c25e0ebaa9
eeb9b77e36ba06314415ae0ea36220179afede1f
1438 F20110114_AABPJA lewis_d_Page_126.txt
dca0c0c2178b69c6a6f4fa94b092cff8
3154574fe0d9b298855115919a5153e160421823
11982 F20110114_AABPIM lewis_d_Page_137.pro
582e2410fe556709d131eac58a5ac719
f7126e40530dae4a579ea6253550989f79337363
F20110114_AABPHY lewis_d_Page_053.tif
76b3051da00006468f1b80d5527d2ed6
64d5aa09c65c0abe07746875f4b4410cfcb1f114
105726 F20110114_AABPJB lewis_d_Page_130.jp2
40a76d3aa7e0c4d0af5187afca629343
9e30d1af98176732404087d783454a6ffd1cbfe0
F20110114_AABPIN lewis_d_Page_068.tif
f28af51be26f10e857f5f667919d63ce
d5c3b28b6efbe0e2b595ee4360a814c5612fbbbc
61324 F20110114_AABPHZ lewis_d_Page_024.pro
d592210c7a63dc2185a4252bbb433259
d43e3730e7d4a229c746bb860a99548f2e4710d7
45221 F20110114_AABPJC lewis_d_Page_120thm.jpg
b66d97f47471e5fe585ee4483a504ff1
825c3d00608ee5c009690704ca035d46b070929d
F20110114_AABPIO lewis_d_Page_092.pro
9049d1718755de813d0bb62ebb7afbd4
da447fc75ff534f49eb342dcdf8fdb33afc0b8d6
F20110114_AABPJD lewis_d_Page_115.tif
16219f4ef267476ec71d7ef3b247d369
8b69dabbd7741c9da62dd70547f78c0c98c3079b
38915 F20110114_AABPIP lewis_d_Page_120.pro
79cf619f6fc02c9dc596ee8a3e34245a
8883ccc82b72dc8c0629c5d39280d093f0efeb16
2080 F20110114_AABPIQ lewis_d_Page_087.txt
c9baeb3299a4ff23765773f353620c35
c31cd3d9c03d34effbd4cbeccf3e2944909cef22
18894 F20110114_AABPJE lewis_d_Page_154.QC.jpg
9586e2e4fa4dac9ad85e732f8c95eea4
91ee17202867fca7e00082a26a2b4f35acd95b33
F20110114_AABPIR lewis_d_Page_101.tif
864ed3e8ac1739a5ff949ace127487c4
cef01e612fd0d1790394f1aea3ef7adc6e5effb2
168981 F20110114_AABPJF lewis_d_Page_118.jpg
f3f15de48c944b5374fbe4436c6047e3
8e86041eec21ac0ceff5d65cd79f5b6239907b53
56159 F20110114_AABPIS lewis_d_Page_140.QC.jpg
90e8094f8c8b14ab614f4919f4bbe7fb
17cd9260d2d3c554648bb815de51778206c0b1bc
663456 F20110114_AABPJG lewis_d_Page_049.jp2
f3bf99bef69424ec1e5806c83f13bfdf
5a19cfe45c3840886ecfdfdffcfd718170b8955e
F20110114_AABPIT lewis_d_Page_057.tif
82d47d0694f70afda40e55d282f2282a
48e357feefbf634350938d13d62263f6d5ca1326
401 F20110114_AABPJH lewis_d_Page_001.txt
983640a0c32f181e2022391acde6a976
0ca60b707ce1fa90d0864fa058834b9cfcedf0b0
121421 F20110114_AABPIU lewis_d_Page_088.jp2
ab2be673088a9e583cc6801f584ceeae
2acd005d23d67488784ec57add5ac53c7cfe9ae1
159061 F20110114_AABPJI lewis_d_Page_058.jpg
151269da0e0379e1fc2529ad86d156df
240c03e1b38240988f55c23459568617612a9211
37322 F20110114_AABPIV lewis_d_Page_140.pro
40653ae94fad5fa7d0a85f75df50714b
7d5c2a7783b2989f16f040e77c4c6ed14120c2fb
40239 F20110114_AABPJJ lewis_d_Page_045thm.jpg
2bf48e1fdddac1028fea650c929ee753
ecb35d139a4987987f92794a2ac91f7c2e8f8210
81484 F20110114_AABPIW lewis_d_Page_121.jp2
97d57ab73228d03e48a8232f1bd7d890
3c54ca419dbcb86560ff7f2dd582b26ce5c3a3af
18151 F20110114_AABPJK lewis_d_Page_080thm.jpg
81baa994de87e65d50a0ed6edf8fe30e
e926a52b71bd02cc777c65e6187e1b92b4e9f8d4
56771 F20110114_AABPIX lewis_d_Page_126.QC.jpg
6b2309dfe108d19082bef2c9104500fb
fa3646238b1c8397ef16f7a7e40fb343377b5deb
549 F20110114_AABPKA lewis_d_Page_137.txt
bf5d20fe9a5c556e6d7ea9a9b7e029de
b892ed4fc91b8a57e6039f5022523d734f5ae7b8
13282 F20110114_AABPJL lewis_d_Page_030thm.jpg
246da5ecdf98786bb81f09cd0ea83681
769e3c9110cde979086ef9bf9ea70c2fcf5db5dd
45828 F20110114_AABPIY lewis_d_Page_122thm.jpg
5ff4d05f579e8e52dc37ef706e2eee2a
f3ea07c8440c6b8730aa746c35b4c08f5a013417
169173 F20110114_AABPKB lewis_d_Page_133.jpg
b2ed755aade77dffda6e62aa67874b79
c1984ca8bf1b5cad8b5ff8a44a69c10421384fad
1078 F20110114_AABPJM lewis_d_Page_043.txt
88b2abfe96ebad05c82a2948348f34e7
6be02315aa40cd023ec06a865d736048c918d6a0
24304 F20110114_AABPIZ lewis_d_Page_104thm.jpg
28fef32bf8427b622b49ffab330b4f9f
3f2ed5807df1d71da640eacbe347f292bdd3bc55
93691 F20110114_AABPKC lewis_d_Page_082.jp2
26364294fc03fc2548a5b25053f79f62
3ae3490cb882df5e85416f24bfa08a97bdc883f3
657493 F20110114_AABPJN lewis_d_Page_067.jp2
eab5359b8570e6c58cd2e195c98412a1
78ecc9484537e0d5e3a08b6f01f3035e6da80e54
F20110114_AABPKD lewis_d_Page_040.tif
baba3bf6be322004d8bbbef4b090963f
cce33aa9f641aa8ead0e433bb172fd2340824701
1043273 F20110114_AABPJO lewis_d_Page_075.jp2
1cd3feaa9c999e97a8d464424cfcb776
1da08744a68e35aa6b9ed5d979735f4753220f17
43685 F20110114_AABPKE lewis_d_Page_098thm.jpg
04725f26f7d3b881405fb479e76ed5c0
96246858cd086c95905f82c40d6a64ca61754c42
F20110114_AABPJP lewis_d_Page_042.tif
1830471eab9b6d3bc5c292b9c142876e
ae04909e0aa56c93b5cab601576222234cd79730
615 F20110114_AABPJQ lewis_d_Page_062.txt
2b52a8bf710917a6b5d8d3115d733bf1
54677a6c2c0a155584ec1cd6df157a43f5e082cc
43814 F20110114_AABPKF lewis_d_Page_072.pro
b42bf050ee3d04f04647751f172d0944
e7c9491ff64bf344e5e3ca2e1ff62198b265bbaf
21296 F20110114_AABPJR lewis_d_Page_092thm.jpg
eb7ebcda3231e9f5d762664a46fd17ce
4f0b04f32141904fff2742ee98c30d930b135e7f
151359 F20110114_AABPKG lewis_d_Page_119.jpg
ef1b41a7d8b08d178ff66a6ea57a84a4
9bab1f558456e0ae59f20f04eaf65be0a6226871
87965 F20110114_AABPJS lewis_d_Page_060.jpg
2654e31f6177cc97fd9075b1f85e1189
fe2b416743e18303ff3ed2142c16e4f663df23b6
41380 F20110114_AABPKH lewis_d_Page_044thm.jpg
16849c4156e69611bb8f1a80bfa31620
4f4195fe2b92afa73e321fc092da88c097fdea3c
32710 F20110114_AABPJT lewis_d_Page_047.pro
dd88d5cc02c7253fb55d8bf72076b04b
c7b02774a94ce32504e51d7f1321caa433728cb3
3013 F20110114_AABPKI lewis_d_Page_002thm.jpg
f75e769c582b1fda6340f2a1d2dde2a7
8a8d6d262b6971db54d72a543627bc87c08dccd8
83335 F20110114_AABPJU lewis_d_Page_043.QC.jpg
873a28d99ce8302d399730c8e7bc80b6
f26faf917addfc95b7ba0005f39fc992f3dd3454
23384 F20110114_AABPKJ lewis_d_Page_093thm.jpg
6780dd8c1c55509a46ae613bdd45e02a
cdc331128b5015375480967d2b32240ca35320ec
F20110114_AABPJV lewis_d_Page_062.tif
76ba8a14f202dacbbc01da611d707a11
68d5473408113ba1fe63a802f2006bd1732e653c
36744 F20110114_AABPKK lewis_d_Page_029.pro
556f15e90a5e91328319011d2c31922a
88a120e9551499aa63a5d07723fd7cb110b2a339
720 F20110114_AABPJW lewis_d_Page_131.txt
3e5b2c9e79219626b184a39b14a47160
c48642cfbf5660799800b22d0ffcd0ae7a1b66b0
486818 F20110114_AABPLA lewis_d_Page_084.jp2
0a89a35850c93e8dcb5b0cf362993b44
f997c4c894bddd0ab62aedc348c587deed4d74e3
F20110114_AABPKL lewis_d_Page_065.tif
39ea588e1f88b9074787d52002ba6cf8
e317e0ed2019dfa590ef1e5dc02ba7fc3c1a84bc
7085 F20110114_AABPJX lewis_d_Page_001.pro
49557238a6ccd51db9def262859bfa16
b61907556910ec249a2f2c8c7b095d3b87b57898
1386 F20110114_AABPKM lewis_d_Page_028.txt
e3b307b2174ce65a48868556f712993a
d79edfcbf432ea8e6e4a54b33bae7975df3ca90a
91261 F20110114_AABPJY lewis_d_Page_150.jp2
f33db3d5b424dc64003c8cd925122814
f4e60b753783a277535289bfddf4d065903b1d02
63739 F20110114_AABPLB lewis_d_Page_037.QC.jpg
83b2924f12c900b8d0ec794a935d9fb6
ea60d5cadb5e45050538c5634dba71fa23242c69
82482 F20110114_AABPKN lewis_d_Page_097.QC.jpg
6090f6c7724461029ba383f26c374d9a
5ea26d1a582f3d9dddbb46e14e9f14a94637c5ee
226417 F20110114_AABPJZ lewis_d_Page_075.jpg
116b0adc0c60db67f243ec46364fab56
9bc4616253d56d9e40df4a5ea4d4e9fe987c3afa
45772 F20110114_AABPLC lewis_d_Page_019thm.jpg
a208bfa513d1a7f79a5a148418c17703
4c133ee4b30775b328f0a0d480685a7646d6f830
F20110114_AABPKO lewis_d_Page_016.tif
a6a66cbee611f61419e6fe90ba2617d1
5c096ccfc1ccc372901057faaae3adbffd0bc63e
F20110114_AABPLD lewis_d_Page_032.tif
44118f7458471a2e261acb8c54f304aa
e04c23418e7e36d7a6f0d3f5bf394299b4515795
5337 F20110114_AABPKP lewis_d_Page_002.jp2
72a84a4c2ce6aab6108b98c0fff54110
9573797145f3f03672ccb7c97386ee5fad503e95
109094 F20110114_AABPLE lewis_d_Page_086.jp2
07969073333333968f6de3d3f565e3ff
339b7b40f91aa6dd32d88f3e7464ecf546d4e76d
1051978 F20110114_AABPKQ lewis_d_Page_006.jp2
18017f3658cc692ee45ee910744595e8
522b7a71276682c5d5504b64ab71942d2199bf21
96587 F20110114_AABPLF lewis_d_Page_078.QC.jpg
191610b354bdc376aa6dbf49c5a99d3c
c54052816b0a53b791203db1deb8c6515be68e81
F20110114_AABPKR lewis_d_Page_124.tif
03fd8667bda74159a52fdd0103643660
42dc6256d84728e2b88b0216f59ba48d6e6e2d33
42540 F20110114_AABPKS lewis_d_Page_136thm.jpg
fd34f7a6856564cee9b248a410d593b0
3f10ce18be04c1e67412bcca8de9aa02d7757006
24787 F20110114_AABPLG lewis_d_Page_040.pro
e6340da824b4e1f4e5819e48ffb70fa4
b1854e6848bcbbacc6860cd300fc55ab202a03aa
43940 F20110114_AABPKT lewis_d_Page_114thm.jpg
9236cea45f265c6e0cafec5b0c6b2f76
a523148ed10cc000cdeb9798739750568f7ab3f5
48381 F20110114_AABPLH lewis_d_Page_078thm.jpg
af2ed41da5379a1861cce95da3e6ce68
ecb5673ff4eaa7bb4d765a37fe85a80d599fa627
28260 F20110114_AABPKU lewis_d_Page_114.pro
0bb4f014352acd34b55f318b54b3628d
31c275cd594ea9359486b3b5732c3e97788f81c1
66070 F20110114_AABPLI lewis_d_Page_145.QC.jpg
0b1955e6c1de39f5d4864292ea5df96c
83b7006285fbf2a490433575a4006bebf358aa92
975160 F20110114_AABPKV lewis_d_Page_033.jp2
ac5c88ad631d59505270c7ed004b9de4
467e5d97b0f7095025362de59624b3879107b11e
F20110114_AABPLJ lewis_d_Page_008.tif
7201a35c7f7653a080ef0f982e80aaa4
1efe590c7643de269f97736a3311cbdd5d29cf8d
F20110114_AABPKW lewis_d_Page_153.tif
ece1577948c8ff7dd5bdf2d28fee5a1d
468506bf8c36a9c886578cbd7282a1c668a86ab3
208085 F20110114_AABPLK lewis_d_Page_086.jpg
f2a6679f84d6317cfde0aeb23b088451
bdf758773a48b09745f2ac0c0e9cdd7504e5165d
273181 F20110114_AABPKX lewis_d_Page_009.jpg
09f7130064a32555074027fa3ee832a5
38b853ae37cc772f863daf8ffa3e8a5b3bf0a7ff
83138 F20110114_AABPMA lewis_d_Page_124.QC.jpg
47e2f67c32e3948ae137af046954be17
4d5a10ef68c94227299082e6869a347076a6a540
186006 F20110114_AABPLL lewis_d_Page_043.jpg
59e5f3e8a350d82192b80a920f357236
5178d4483ad1a8f1d9f6418fa21f5e7a7ac9a007
183520 F20110114_AABPKY lewis_d_Page_070.jpg
34bddb4440ccfceb03047bef22c0e45f
6beafeee9f913e2bc8b0a4258be5eb6f436bd4b4
1537 F20110114_AABPMB lewis_d_Page_100.txt
508b3bdf61db22a91e8dd3a0288e35b8
9f2cee1c608b0bbd57fa99dac0955448dabb52ed
149292 F20110114_AABPLM lewis_d_Page_080.jpg
9468a8027e5c2312ee090a560eed7391
249da2738e09a86bfb55fbbbfa8fe93630f23e00
103036 F20110114_AABPKZ lewis_d_Page_123.jp2
fddd010d83ffe1928a22b75b1e086ff9
9cc1619f5e1086dc800a9bb1f5fc2c319893697e
81414 F20110114_AABPMC lewis_d_Page_139.QC.jpg
53700e188d635b3cffc6d79a31c27537
92ef69f4664593fda76e30a8acf059942d5fd616
42386 F20110114_AABPLN lewis_d_Page_011thm.jpg
c1626f4f18751c68f3ea5cc15a7b9eff
e199e2150d7fff0909491991c4ac03477339160b
21888 F20110114_AABPMD lewis_d_Page_036.pro
50f628be963122749f42df5c55d3e3f6
5657149a90d805b131a6582c2bbbb2fe9998d79c
F20110114_AABPLO lewis_d_Page_021.txt
036a0ab516418a6d439eb1b3be66c8bb
8c07e2f6023181fdc117894ae1334bde242f4c34
29161 F20110114_AABOJC lewis_d_Page_129.pro
19042403f96ce16f061090b5ed4d068c
2bce108fb5b360950f34ff1c3a88514620777ac1
106626 F20110114_AABPME lewis_d_Page_009.QC.jpg
1c86ff70fa91749df146b19120551730
f759a9347f1046f105c5248f163dfd18fe175d90
63516 F20110114_AABPLP lewis_d_Page_067.QC.jpg
2979e85b6b99b44576d32f3a37a29511
36de927c7950ff72fe1985a4dfa0cb17b7b2b710
159766 F20110114_AABOJD lewis_d_Page_055.jpg
3d47a166cff4580998589109eafd6d48
541fce7a6b2d6f67797272303a6e62f05f7c7a53
1780 F20110114_AABPMF lewis_d_Page_082.txt
2e8cf770d75bad998468bacd5a982525
535bf3b29afbcae41fdd64ab226a71b0f66088ce
162510 F20110114_AABPLQ lewis_d_Page_050.jpg
344c1a52a29d604179503b405a595ee3
65cf5a29f320ad95819f81fd93fbe5ab886c1400
2462 F20110114_AABOJE lewis_d_Page_095.pro
4e819d5f03282cbe79dd6c1376109443
f79697f64694e027501d94609d2147c1aed19f33
12846 F20110114_AABPMG lewis_d_Page_153.jp2
81230589329bb0aeefa5ef8dd73376f5
57110138458aefaeb3b0239f35dea4e4cb23b7f4
F20110114_AABPLR lewis_d_Page_150.tif
e81ceef497304869b6958d4d4128e20f
dc5a1737502bcbbaccf8b0411a88a88f960ceebb
F20110114_AABPLS lewis_d_Page_130.tif
c0379045038aa799544708aa3f69decb
916e2867775f552bf907cc9c0d61a5e69379b9bc
27824 F20110114_AABOJF lewis_d_Page_003.jp2
58be0d2cb4d5cef102c8d12931189c1b
79662f951365510f58d0f9d53521dfa802dc97b9
191720 F20110114_AABPMH lewis_d_Page_019.jpg
0b15cc1710541c7ca26a4eed34505bb9
729b8457bf26472237f29ce3009c2fd503355868
1009468 F20110114_AABPLT lewis_d_Page_052.jp2
4859ce21d32da32c222bef09c5f9479c
04b187006713b6489bec802896e8cdedd034e352
987 F20110114_AABOJG lewis_d_Page_022.txt
bce4e07688b359be17bf8d42b901115e
b0135aa365b639b3f9988e3c5cca9d4e43a29d09
844461 F20110114_AABPMI lewis_d_Page_127.jp2
5da54e5103f5745797dcc733b85052a9
e939b3563f728fe58de534404eb494364372e92f
234422 F20110114_AABPLU lewis_d_Page_088.jpg
84d0bd58005f848e9760e67d55cfd5fb
b2e6d1afcab0244030007f568a870cd84eea36fc
61123 F20110114_AABOJH lewis_d_Page_061.QC.jpg
eb47baefde7751776fb2c4cfd535f112
be58a10e819179dab27521898e9469826f7fbc62
10631 F20110114_AABPMJ lewis_d_Page_050.pro
f38d252f4b8f3fe9351b488abdb95e8f
2a64509109701aa913ab659edbec3ad4c61f637a
91365 F20110114_AABPLV lewis_d_Page_013.jp2
3411d6b00a50cc831b80b6f2dcb0fe2d
7079cb1de85bfc66ae2b425e3949e27259caf0c9
49313 F20110114_AABOJI lewis_d_Page_123.pro
200433e5188f361fcaf0431cd8f06bfa
dc858d6041d73a4c91fdcfcaf4bcbd9488fed99e
2459 F20110114_AABPMK lewis_d_Page_144.txt
8a1b2600db62fef9121ade2edcb03044
291a4afdb72d9e0d1f0de3602aeb7c583e3f7390
64497 F20110114_AABPLW lewis_d_Page_144.QC.jpg
76630c0d59fb6c74f7b22faf0d490035
a8e8e0ca1f1c9b58d4009d4bc98dcfb9897b5d9e
918675 F20110114_AABPNA lewis_d_Page_135.jp2
8200af8583182a98f9369a3657722945
d95ae024a90ea9ffcd2793376be7c13fbfea79e8
1338 F20110114_AABOJJ lewis_d_Page_046.txt
8436a001b213c84ae0926813f56f3c1b
88d6ce7e0943ad57b52f4a38e4cccd56d51cc401
24168 F20110114_AABPML lewis_d_Page_154.jp2
b515b3f705ec079618d6c4113dc8195f
cb68bfef80f9825b4c90d4b988e153de9e74d628
F20110114_AABPLX lewis_d_Page_046.tif
bdaf02c61482f4a78ee7005b5e582853
6109fe22043002385dec6c1cf965dd8c83509e19
F20110114_AABPNB lewis_d_Page_091.tif
9d794dbca9e243f9f80766c1e07933ad
6705a138f04d8269de15156fdb239f8ed1dc90e7
1015 F20110114_AABOJK lewis_d_Page_066.txt
db13ed83675ac83c2a0e15c3f79126fd
b9d4095c9c0d0b46a85c197b11d2d5e0a2c06735
65189 F20110114_AABPMM lewis_d_Page_013.QC.jpg
bd988d79f5dda70401f184fd8de8a90c
2e529f6b82ca4535fb71d1eb9baede6d6ea6982d
26571 F20110114_AABPLY lewis_d_Page_141.pro
a90128a3e7cd77aa33db002bebf5b579
68b6eacf1db9cd17509dbc276439f88ab8fa9357
F20110114_AABOKA lewis_d_Page_004.tif
0d48b1da15eb88836f0a072fb83ce13d
b97a333b96c7e65c55203c27e30d4d8d4875aa1a
83045 F20110114_AABPNC lewis_d_Page_012.jp2
8a7c57ebbcde50be0f3188cbd1a568c8
0250e90325abe76dca7b5e20797870e3af9bae80
F20110114_AABOJL lewis_d_Page_090.tif
196a812602905b23855e1201f22c5abc
231695ffbd9c4bbe06980d95d983624939df2c6b
25320 F20110114_AABPMN lewis_d_Page_046.pro
db78a126f169259bb58dc68f37af68a5
c4f9179b523b9b310c2edff3d1ae9908e5eb84f9
163421 F20110114_AABPLZ lewis_d_Page_032.jpg
999645fd198fb316b0018a828356f2b3
113272c574aa3d46c3ee71868c83633271a29242
34234 F20110114_AABOKB lewis_d_Page_126.pro
0571a25b3ac50a279f625f55b6c2252d
4ba2b81b3534679d684a562399087d74fee006f5
248332 F20110114_AABPND lewis_d_Page_104.jpg
260ac8d74cc84cc40eb9e52e8b8422b1
1381b084b33df8694b1034cf50c97bbf5172a784
38538 F20110114_AABOJM lewis_d_Page_128thm.jpg
0c044f018f15b975c382bb4f4f6e6a31
2b832679cd291c2abae3199998b165fdeaba5fd7
89491 F20110114_AABPMO lewis_d_Page_016.QC.jpg
93f0bfe6b4542aae5d737e7d07b5659d
eab69104f8b04de0a6f9b545177fba2856fac492
F20110114_AABOKC lewis_d_Page_148.tif
74ba8754cecf9de6d488ebe680f0357a
2f46694bd48005a012882737ff3e86fb0171b038
651076 F20110114_AABPNE lewis_d_Page_081.jp2
58ba442981f91278aacfc3925aad0e05
afb05eeb7f2da991c86edd0eb8b95e222786a837
182227 F20110114_AABOJN lewis_d_Page_073.jpg
e1c513c698f8b2e167d75d2637245fe9
f39ef73a6f54a075692a75392a42cca87213760a
1046 F20110114_AABPMP lewis_d_Page_068.txt
87309432f657df4d35f696f4f6f25771
c7417e546f1e8ebf3f64619f11f0d2a5d6b50bef
F20110114_AABOKD lewis_d_Page_118.tif
85de5cec9d9c8bd8d1c9c2665c8aecdd
91ae61ff8e4ab0aaa1b347b6937684d97ae42025
23007 F20110114_AABPNF lewis_d_Page_086thm.jpg
10fd7856bd7f9e9d4c1f3d952f11ef8f
49915c9e42d2d8c18d827e8e067437188ab38843
F20110114_AABOJO lewis_d_Page_138.tif
a1684a38c57817bdb2cccb5fadf1373a
61ef329c3836dd48d8db06e7b5f88eaa83a4b20c
964 F20110114_AABPMQ lewis_d_Page_119.txt
79a27dd821a87f5c47bd16f7fd605a4f
d7c254c7d9f70eb87313a466960eb46d8975d68e
F20110114_AABOKE lewis_d_Page_104.tif
e27a9d895d879a9629a775c82fb4e5e4
7e5af4c85de95b9d194ce9522d382a304d882c9a
46157 F20110114_AABPNG lewis_d_Page_007thm.jpg
9ceab5998901b421a0a420aea2d3bdf5
9a4c7aa4509f1a7faf0a3ec2704a6993cde5a389
49293 F20110114_AABOJP lewis_d_Page_079.pro
781bce5cf0bf65026f040b06b5e3fc9d
e157abb4d13f9cab4d21c49f184187ac2fdd65b3
35487 F20110114_AABPMR lewis_d_Page_085thm.jpg
894f6ba069c9f879a4966a67f81e36de
d235b18ddae4c23d2b3dd4b7cf711a4c3485b010
279 F20110114_AABOKF lewis_d_Page_026.txt
8abbaa76b272f73f7a9ebf1d49be4db8
13a8f84fb68aae52563a747e0b25a14763e4e012
85868 F20110114_AABPNH lewis_d_Page_073.QC.jpg
cf854daa9a5f68bb38168460ca3dabee
de5682ad52257e8d718f7b20716d6ce599e6486e
F20110114_AABOJQ lewis_d_Page_149.txt
81ea56e91698e6aa256b777bc942342b
661da7cc49a09ff224501c7bd4ed84c09a9d4cd8
85143 F20110114_AABPMS lewis_d_Page_143.jpg
5bbc16bd25591828eaef5acf427a944a
f578febc8eb24f179a775ef6488d24909520d557
179263 F20110114_AABOJR lewis_d_Page_092.jpg
75beb4353eb7f462a05011e232214211
b33938cf8597f8c95322b5e4fcac27405f9b3f82
37533 F20110114_AABPMT lewis_d_Page_061thm.jpg
dcb5481a33382504ed7244643cef5df5
0ed17be0999c5aa79fedc5e08ffd6fc1c8dd595a
70834 F20110114_AABPNI lewis_d_Page_119.QC.jpg
11a906b10754a20aeaa4968128170400
c62d2cc0bac07b794b87a9d8755c03966aeddc49
58976 F20110114_AABOKG lewis_d_Page_121.QC.jpg
53e11257e77a4a2e4c60d88dd3c54457
f85dab4a6b8b222f185c20324f324ec5d3fb6687
105170 F20110114_AABOJS lewis_d_Page_076.jpg
1a64b60e1da71c67dfdb0d30dc6272f4
72a1b36a2024a1b747898891e80cb7ac061e4baf
50552 F20110114_AABPMU lewis_d_Page_143.QC.jpg
340652b54b3ee34c83f89875ba0c0122
11118d83d7d8106afc61fd0da2542e0dc9dd2e95
51075 F20110114_AABPNJ lewis_d_Page_065.QC.jpg
a32118c7d8b8fa1d41c135ccc1d6c90f
ffbaebab14e19ee06970d2913c75cbb42b0d620b
83307 F20110114_AABOKH lewis_d_Page_131.QC.jpg
67d617a31f346c8c1421cda57cafc20e
7b15f269f90ba2835d37707cc09f85fd7a53fe4a
2788 F20110114_AABOJT lewis_d_Page_006.txt
e60c81dfed734c32594ea90e888adca4
51e39db708dac59194d7857a37612ff5b64677e0
F20110114_AABPMV lewis_d_Page_005.jp2
75fc91420df11a2361fd7911ccd2b07e
1c51bd0bfb3737967d8f9bfe91d15bd3363bb05f
F20110114_AABPNK lewis_d_Page_031.tif
806f5f86744339ea1efcb98e6d5ebf40
611ba0999fa28b29d0881ade07f5e2acb24a3e28
4311 F20110114_AABOKI lewis_d_Page_153.pro
9b0f5dfd5946d1e9b4a6d8f0c82af9bb
cddf5e56f1194ff2bff73cb0271bfdbc27d0cacc
477 F20110114_AABOJU lewis_d_Page_050.txt
265e5d64bc5d94464237bd1fd4d201d8
6b53545d0cece862c2a9decb4156cd277f591077
124322 F20110114_AABPMW lewis_d_Page_104.jp2
c35b404b967723f8f8f84ec4f64c6e31
329d505ff350a68c58162fe4395230a56fce674e
724108 F20110114_AABPOA lewis_d_Page_045.jp2
2d621dcb1ddc6666c0dc892f5069c87c
fb93c915217365ebcbbc5a70b0a80e24dbb1720e
38955 F20110114_AABPNL lewis_d_Page_036thm.jpg
bafc100b6bf5a9258b7bfa3912aed5e9
06a3d8b6ce2fbd493e54f52285d77065edbc0984
1206 F20110114_AABOKJ lewis_d_Page_018.txt
82a0475bc4c9b053bb83c621d2488ec1
4774425ae3f65cfafca491aadafc1f195d8a7c0e
1556 F20110114_AABOJV lewis_d_Page_080.txt
9724be58dfde5bc3fcba322401237d2a
980c2a1f300a15cdc5718980e2af6026c2a6c8d6
33945 F20110114_AABPMX lewis_d_Page_033.pro
eab2b570842e0798faa199eff7fc4949
dfe44bf2be3bf755d7d768c912693a3327e1e0a9
F20110114_AABPOB lewis_d_Page_149.tif
2c4092c2fae160fd1a3fa65ab9fcdd9c
629957283485c02db3a585b97979b97edac49ecd
73753 F20110114_AABPNM lewis_d_Page_111.QC.jpg
974afce0d49d8295a8e48a4469b31a9b
e2e1001ba6d4e5229a3e2866f161972f682060e9
1424 F20110114_AABOKK lewis_d_Page_016.txt
c6cfe6d1fd6d53fd9b234be8d0e211da
29ceb44698eb363c61f343cbc060c0175736d45e
F20110114_AABOJW lewis_d_Page_095.tif
d4e26c52cc49717bf89c01bf2ab3ed97
119852540aa6a80483bafb8b509c5c16b3fd6452
48878 F20110114_AABPMY lewis_d_Page_075thm.jpg
ba12185d255ee02bf45c580a63a0d68b
2f00b4956dfc8b4862496407edf77041f90a0736
F20110114_AABPOC lewis_d_Page_123.tif
c41407633d4f6ba1447ed3e0bef1e650
88af41588d3d5425596037d49bc627eb941a592b
45318 F20110114_AABPNN lewis_d_Page_004thm.jpg
5046f95bd3cf5d9d1cdd2d1394391cb0
7cec490bdfec08a1666013ecd4418758029acc97
67063 F20110114_AABOKL lewis_d_Page_074.QC.jpg
f2f712493dbf03c29b0876513042489f
e1cce86dafa93540bab6b78a820689fa1aec8df6
214117 F20110114_AABOJX lewis_d_Page_017.jpg
277de50f8c20d8bf7922c2ce3e9b01b8
4f5238cb59abf9ce50145ba16a455388f00f60e0
186705 F20110114_AABPMZ lewis_d_Page_131.jpg
4cae26f8e49cfb1813fe2b6435ff30a9
464a91f117462adae71616ee4e0b73c80293c8ee
99238 F20110114_AABOLA lewis_d_Page_034.QC.jpg
709dc4c60a82386ef1eb3e5705c11113
15f560fe01f5a95cb884dd9f11f4330b2c084b75
36942 F20110114_AABPOD lewis_d_Page_121.pro
65a45c4672bd7115926255bab871b68d
b333163bca1988572f38c85de8a1ada401951815
21461 F20110114_AABPNO lewis_d_Page_113thm.jpg
6cc4ef6c37dcdfdcd53dd4e34a0088cb
ada76e5d65ad8aea455806aefe0600d2c41244fc
168910 F20110114_AABOKM lewis_d_Page_052.jpg
b175748404ce47be747a6806f63ea79d
4bfaf2cf2c9487579e3f651eb38e4ed5f9f17e58
309418 F20110114_AABOJY lewis_d_Page_010.jpg
20212a8aeb93e71f686435c6ab061b9b
bac0ad1f784cbe4d2ef75bfea3ea59b646700ae6
F20110114_AABOLB lewis_d_Page_063.tif
23a872bbb0cb631e8ea633de4962b401
b65c7c7c56c7c9a8753a036df57a2b5fb8052165
F20110114_AABPOE lewis_d_Page_049.tif
86d74ff256cb3660a17c4c370591d2e5
c77af2d917e7e95414fa4874bad09a4e20dbddf4
89366 F20110114_AABPNP lewis_d_Page_092.jp2
69c8fdb7c0c23c7661137d0c43da017e
5550587f5346b6c85bc3e99cc2b6d57cef5ef70e
928 F20110114_AABOKN lewis_d_Page_116.txt
01edfd9491a9a571d3320a7d802e00b5
8124758c0b28dc9f89f1c8423c25698f5eda5f2a
22246 F20110114_AABOJZ lewis_d_Page_048thm.jpg
7b3cf13e4844afdaaea332d21c8b466d
ce4076d72d74c7a10555f1e56db7bb70b701134b
F20110114_AABOLC lewis_d_Page_106.tif
82ea96f92dc8415f6d72182ba22fa2d5
ee8831c5939c88c18d397f24145d8a32055d5040
F20110114_AABPOF lewis_d_Page_103.tif
00af66191d433a18447d0461d6af25d2
8b820088134debc8564140bcfebec376240013f4
F20110114_AABPNQ lewis_d_Page_006.tif
83fb11e90a3c34879d2d2e57b7e4d3db
0ec0e5c2b220de70c7cfec9b9e5e4588bd7cdf0f
845274 F20110114_AABOLD lewis_d_Page_090.jp2
d3e9cfa8ecbbc954db6da4bf938eb5b9
d8506658b0bf3962d7dea2698b1782ca95111190
135389 F20110114_AABOKO lewis_d_Page_074.jpg
071fb7da5eee41ff4c56e9373c34f55d
c39b491fd84ef7c2698c1a03810933caeb8baebc
921797 F20110114_AABPOG lewis_d_Page_120.jp2
31677735d21b6f8ea061c5ee1df9763a
a3199034cdf3c78faf08cea3d930792d940e4c39
19024 F20110114_AABPNR lewis_d_Page_023thm.jpg
99ace4c1d5adcfb5e1fedc2c59ab0abd
b97e084acfe651eafa6442ca2d37bd211097eda5
170370 F20110114_AABOLE lewis_d_Page_047.jpg
7739b6bb4a2eb599d7056b48522c994f
ed0ddaee4cf803b05f70bf58266361aec6dceb11
5563 F20110114_AABOKP lewis_d_Page_026.pro
7bb837f9648a0906023ec10eceb4c99c
ec97b623dd1cc8642c62996197b125b2d316e162
24669 F20110114_AABPOH lewis_d_Page_022.pro
9d65c19b5a21f7ba76458389df24f1b3
eea8092b844963b1e9c6da7c3f727e1b2a8a8d5a
40235 F20110114_AABPNS lewis_d_Page_081thm.jpg
b776f159e6e2f325f5c47da1e36e7d29
11dc53cf321f5975127055867361e62ece35cf20
F20110114_AABOLF lewis_d_Page_072.tif
0897e610f8fc0cf45d1279e484483a34
6671c908f98d3c54142862e22308c5463911485a
2000 F20110114_AABOKQ lewis_d_Page_042.txt
361f754027c571ff27a49c58dc31b7cc
3240a11fe1019cc21abdbd40215a09c3cae5f5c3
39231 F20110114_AABPOI lewis_d_Page_069thm.jpg
ac8fc762c9a356ed4b336b4b8a411de0
90710b5701f4a86ac2158487bde68accd2ed428f
18640 F20110114_AABPNT lewis_d_Page_149thm.jpg
d35b37069862763963245024c56c688d
f34e6e1abe7be89082f97fd26806309819851c8f
F20110114_AABOLG lewis_d_Page_145.tif
969f29b96c67d366d41617e4039fa9c5
7e5a019547b69766c560ac3954f7c47f81f807e2
19359 F20110114_AABOKR lewis_d_Page_012thm.jpg
47f3f34386d7ebb4741a03997c83aa2d
6fcef65bb7a8f108a4d539beef081af4e04f9cd6
39772 F20110114_AABPNU lewis_d_Page_057.pro
5dbb34a3e774c59935d6da34dfaf3890
b02ac7ae82b2107b6531d6d0bb9493e63c433972
83144 F20110114_AABOKS lewis_d_Page_117.QC.jpg
3a1a7414fed6d4db8f7018fdde7cbd06
35d3731421a5ec2837d60a120cf62478dac10280
88106 F20110114_AABPOJ lewis_d_Page_096.jp2
0b7bd588e7011ef17456774b49dab061
181f2066eccf4109a7cdba5b14d9931b0bfd5773
165009 F20110114_AABPNV lewis_d_Page_012.jpg
0c030f030e22d6f2dc82d64b136b89cb
e9cc0539ed2dae119306c009ec617ca526baf65d
18819 F20110114_AABOLH lewis_d_Page_070.pro
65fb06d55d26f34d5839b8441a23b9e6
4e14858492e5ba5839723c933274848329902d12
125082 F20110114_AABOKT lewis_d_Page_061.jpg
960025dc80374f842a715c680a2ab36d
f6e9bf494bf9f954a9f79c1ad936797ae1d3bb9f
2502 F20110114_AABPOK lewis_d_Page_024.txt
3e5b7ce4e1daf6a0bfa0412e28e1dd19
545d33cac7abb162c7f613ea4180c709c49a452f
F20110114_AABPNW lewis_d_Page_008.jp2
e74691bb03e418b06bbe2dafe39344bd
30415ca0870b2300a7cb4335e7e79766682b1635
155354 F20110114_AABOLI lewis_d_Page_053.jpg
17931cf37c07eb63cf541c8f60452d00
33df9f2ee76d29c7aed60e7989273e8ec276cab1
79935 F20110114_AABOKU lewis_d_Page_047.QC.jpg
686195a9aa9fe8e7e18c3a4b12102350
84e1b533a75bb3cca1ffaf199b4062995ec79e67
40385 F20110114_AABPPA lewis_d_Page_077thm.jpg
1e16df56231a80fc5bacced04b28fb48
bd9ce90ae63535d6bfa7c2c85220395b0c0b1c33
46492 F20110114_AABPOL lewis_d_Page_152thm.jpg
605f98fded91704142788bf61e7cd18e
c5d3692dcaffb5175a33537e8953341471495dc5
70699 F20110114_AABPNX lewis_d_Page_146.pro
57109113b1e35caa66f90ad180cde311
bfaf80e305f6fd493ca7f6b4a6a81891fbd8a9f1
88479 F20110114_AABOLJ lewis_d_Page_025.jp2
6a23c4f022118df58935c94a04cfbe47
7c769fa29af7d066f3f38aadaec17bea5954a81a
F20110114_AABOKV lewis_d_Page_050.jp2
8d348570f4cd6cd7bc194f51cecf662d
0de1562bd496e5003d5548e41e27d096142cbd36
199181 F20110114_AABPPB lewis_d_Page_057.jpg
ad6e6e7b166143c7eb4a12f266077fc5
563f6519145b605c9f932bd37d9f071e4439e5db
39406 F20110114_AABPOM lewis_d_Page_056.pro
f8281074fa6e878fb5f9c35d35156e7e
cf44f4180f6c18cc3809b3188a9e9a416238615b
51759 F20110114_AABPNY lewis_d_Page_149.pro
81ad9343bc940e09bf7aff50387b4d9d
849318097e788940ba563ed6f5dcc655114d07e1
124838 F20110114_AABOLK lewis_d_Page_036.jpg
eba6d46c3c1815034db0ee655a913677
0a90344aea92486bd6e66aade34b72a2d5d03c96
39728 F20110114_AABOKW lewis_d_Page_050thm.jpg
33a11fc25fa5941010265b5dc778dafd
118b7b99cd1e6bd463d88a5d93b9edd3851b2c0e
37528 F20110114_AABPPC lewis_d_Page_068thm.jpg
f214c9cd496260b4f1bb7a56685d09a8
8bf684b7637c1e750a8a8d6c351fdfd934560ab3
76760 F20110114_AABPON lewis_d_Page_058.QC.jpg
361dc74dbcf9621ebfa3f29166d20c66
cdf0dd896b73b272bc92f9cad0482545296dfe8b
1519 F20110114_AABPNZ lewis_d_Page_035.txt
277f332d7d66654d678fac2d68f3f7e6
50c2b84d9d64d8e6aeaedb5124fa911703391d55
1884 F20110114_AABOMA lewis_d_Page_111.txt
e3b24b4d9e0ce704b67d9a3b939926ba
8b392005859e4eaa2ad5ab5df9bd6aeafbf09211
373 F20110114_AABOLL lewis_d_Page_085.txt
19b1eef9a77d25ba270ba8923fb55eb4
fc04855ac55ebf64aeea57de9e7fbabd1133f0f1
14925 F20110114_AABOKX lewis_d_Page_051.pro
6d0135a7ce3252e761e0bca876ee07a8
364fc2a149c834a78c957d98f1ab8634dc5ca702
F20110114_AABPPD lewis_d_Page_050.tif
3b1acbe47e98658d0dde1e9271ebca95
79b046ae122961694e79c7448aac28bab8487d0b
19624 F20110114_AABPOO lewis_d_Page_100thm.jpg
05944895efae2b5762cd7c37c4a4ba0c
14c0fd64301f227eae257e5510db9facfd906278
12653 F20110114_AABOMB lewis_d_Page_116thm.jpg
0532389b68ccf9fd3cd67a593488ac23
a21758a6f763982ea67a280317737cdc66de1a32
F20110114_AABOLM lewis_d_Page_127.tif
439deeeedb70b0a8767ca5484666bcfc
21c25f0af89d88277735900e85e4552ea203e9bb
67740 F20110114_AABOKY lewis_d_Page_050.QC.jpg
d6976c8329c8073aa66e4322dd4d2587
7988f09d193fbf3b9a540ce4407921d0b9a39789
41740 F20110114_AABPPE lewis_d_Page_031.pro
e1a36cd9c4cdbbe11251338b282ee81c
044e7e6fbcf12d7299c601f4fd4b79278bf3402e
58800 F20110114_AABPOP lewis_d_Page_088.pro
4eceac78b20688faee715758ddb1af34
4e249cc3d7cf23b778ee0cbed591701388df9415
1051938 F20110114_AABOMC lewis_d_Page_051.jp2
8adb68d082bdf9978cb5226027159dd3
585db8c3a6c75f363edc447129039d057b288cb8
14534 F20110114_AABOLN lewis_d_Page_037.pro
e713cc2aef84bdf583be0339a6296df1
c8d4c6cd0e823067300dfcc9fda017148f0f8228
84999 F20110114_AABOKZ lewis_d_Page_133.jp2
199988ec9cb964757db5402366b6772a
c5977d62f1b378fd7bfff02e8afc148b71f2612d
10568 F20110114_AABPPF lewis_d_Page_147thm.jpg
6cb6a44dee4c9c91e81e8be183522cf2
e5ea88ea8dac36966e28274bb2d9d621fd717e0f
307448 F20110114_AABPOQ lewis_d_Page_060.jp2
ec4f9634b5406d4cd4f7170073c95b14
821a699ea4f9a41c491d228b1984797953fa16c8
F20110114_AABOMD lewis_d_Page_071.tif
f692a158ce78c9233d9c379444acec6d
7e49c1020ee4c173f6731246b799fd3d1d992208
24980 F20110114_AABOLO lewis_d_Page_109.pro
92c3b80abc4543c6d1cc88fc2975f6b0
0c44943472695a46e294b7854602daca3e0c8bc7
F20110114_AABPPG lewis_d_Page_028.tif
e982f2bb928489168adead4ace483437
93d5b89a5f4369c21cf865fcc45f99ca46733eab
44916 F20110114_AABPOR lewis_d_Page_097thm.jpg
c1d594db7de7219af65baef2b7f0bde4
a11b77d04221bea1edf810d6f890f394e3cabd59
83151 F20110114_AABOME lewis_d_Page_070.QC.jpg
f84a9ee30117d4e26134d8cebfb96943
f63bc1950d6de8fa7187aea6a63c61f4d5395801
190282 F20110114_AABOLP lewis_d_Page_051.jpg
12ec3feecca3b30f0b72e08df10256b1
65b1c8acafa732ff7f5595f9d3fd7dff90044e55
40897 F20110114_AABPPH lewis_d_Page_096.pro
710fe0abdb71196eb96f830c16a8dd53
925b0b97bbfa2cbb35cb3121fad9fd700395e0f3
103304 F20110114_AABPOS lewis_d_Page_089.QC.jpg
e4a15ab571f8d604e6ac38cdde958efb
044e712966861d732198218801196c2c0e60c7a9
474332 F20110114_AABOMF lewis_d_Page_061.jp2
9e771ffcb69de6acdec6045bfd51c67c
bc42cc30762f97109f465692f8660fd98ad45dfd
F20110114_AABOLQ lewis_d_Page_073.tif
89c074509e5475b8ff1ffc257542ece4
edcc83ed44034a3e32b1cfbf19dcb746023ede9f
89506 F20110114_AABPPI lewis_d_Page_029.QC.jpg
51074142660eb1cc9b380b5cb6f31476
cf933feae5621f7b3c9a178370d69a0c8ddbebc8
88353 F20110114_AABPOT lewis_d_Page_004.QC.jpg
692c112a14277c2286e9009d5548dc0b
5f4506e7f1cec48832920b238119499c34dab06e
18055 F20110114_AABOMG lewis_d_Page_126thm.jpg
e4c4aadef0efc6fb53e302e5a580c1d7
fa479536d85d451ca82d0a599603e6556147ec72
103908 F20110114_AABOLR lewis_d_Page_008.QC.jpg
6c841f00a0a68e3549f985a42e9e7d60
6d5d5ce05b9058764c1e51d21b86a3e3e2f67030
54764 F20110114_AABPPJ lewis_d_Page_076.QC.jpg
d6051fb778c32669e310931e1ace8a29
81a01bf7329db580d3c35860e894fdf9a235cf9f
530240 F20110114_AABPOU lewis_d_Page_071.jp2
a4b5f1ebffeddc7ac86dd90a73845563
ae2acec2497102090a8ead3349ebb4c3b1db24fc
77843 F20110114_AABOMH lewis_d_Page_093.QC.jpg
2fc3ca4b67119c745822a7983344f128
0a5ba894bee467725e33a8c90371b8f357942b51
200281 F20110114_AABOLS lewis_d_Page_029.jpg
343eeecc952b424504303af7dae5f8cd
df4b023e3ff1b7db74e42f419825d34950d1685c
137412 F20110114_AABPOV lewis_d_Page_142.jpg
8174802d94cb604e9addebb8b669bcd7
ad35fdf8959bffb8b7f296f84ab3fbafd1ebd46c
70248 F20110114_AABOLT lewis_d_Page_142.QC.jpg
51a674886d5746dc23ddc3bbffdf02e9
c960b9388791f74253d703e63136a17204381011
83979 F20110114_AABPPK lewis_d_Page_098.QC.jpg
3433912968d21496e411fad443436a28
35efbc1a6ab711900c2dd9a3debf249e1fcff5a4
40930 F20110114_AABPOW lewis_d_Page_026thm.jpg
97d4dbd82d752d6a9e58fb330d187bcc
cee3a52f38f66b0547a6c9ebab8df135715b7d1d
F20110114_AABOMI lewis_d_Page_135.tif
9d96f306231f4f92c3c4ab57180a8bbc
99041680eb68d9b70b8cd6b122d3b05f42116940
15371 F20110114_AABOLU lewis_d_Page_067.pro
70425218f6c457df380f741ee2028b25
7a3dc154be8735dbe4f0d88f364e9c3e92dc87b5
43339 F20110114_AABPQA lewis_d_Page_046thm.jpg
3f15240bcfa263f50eb34f6bd9f0f517
11dd114ef5b933b6ca5d23622110ea098f01df12
207542 F20110114_AABPPL lewis_d_Page_112.jpg
0ca150d6ccd395eed08ceef8ad0961b9
e67e672240b13476ee700dc6ddd8cf77dc26d587
68173 F20110114_AABPOX lewis_d_Page_072.QC.jpg
e1d364abb56c4affdb522383511cc5ea
cd87a37429cfa13d86b843cf140833c63b8e308f
75941 F20110114_AABOMJ lewis_d_Page_114.QC.jpg
d80b8dff47b8f356666d179a1b5ae7f8
b29c7e79724b47c70ef9bd3ae4ee0ec2a751a8d7
F20110114_AABOLV lewis_d_Page_054.tif
36fb1ef7fa1288065a9671deff011082
daf6c7fb49ff7a34f3595a3d8b873489979161f5
F20110114_AABPQB lewis_d_Page_013.txt
b18a06a1fc7537f6c9f06d2ce2f653fd
18d72c9724742fc866375d9586061523b8e1ad47
686031 F20110114_AABPPM lewis_d_Page_046.jp2
c3b93b7224945949c31be37ccf0e5a19
a9844e9ec9115a57f5892f45fa8625575e294234
26838 F20110114_AABPOY lewis_d_Page_032.pro
4d7850d4b26685fcf0b6db9e2931b95a
ab74496f87e54de412e4b1b8c9ac3756d3b4dad8
F20110114_AABOMK lewis_d_Page_112.tif
81dc34ef6cfa49c71f7a2fdddb9815f3
9cd9be47937f6aa98504082bc9f5408786357878
3666 F20110114_AABOLW lewis_d_Page_004.txt
54c45ec8f7c3c8839bad8c5e683afda0
284630edf5f8613b03974909a78effdf3e3e6bc7
F20110114_AABPQC lewis_d_Page_066.tif
b5cbac07db0b5e4389270eec3ce36866
cdd16460687371411f0e91c0684d420a541c3806
1263 F20110114_AABPPN lewis_d_Page_117.txt
f162acad6d51220c6a0caf1b192ee4f7
5e426f8ea2a1e268167d38196d220aefd058d3c2
776457 F20110114_AABPOZ lewis_d_Page_117.jp2
e99f53408ad3dc92c6240145175ae1ea
99f578b69a42ed97ecab4fefa865f46b895aca2c
916586 F20110114_AABOML lewis_d_Page_057.jp2
ddfe27d6e3a0e710e7bf9155b9b58413
aa01f4c8fdc5c30da0e93b2e2d26b31d49f8d383
F20110114_AABOLX lewis_d_Page_134.tif
b02c9393b45bf0e641a663ec1090bc7b
8fa6df2e0d37c2a83846ce80f95c80ed2e277848
769284 F20110114_AABONA lewis_d_Page_114.jp2
e5efd0c2e14a7bfb85d08b22a07c05a1
e66ed48084d2187cc8efc07f25ec2ad03b059a9d
117012 F20110114_AABPQD lewis_d_Page_137.jpg
3cde51dd85fa1dde491ac2b0a7f0b77c
49c7b658563179b9232197861103e708dc087d35
849691 F20110114_AABPPO lewis_d_Page_097.jp2
f6520c9d888fa4bfa104abd074d31664
583607351c38ad3ddee013e82f9ccc86937e4f3c
18896 F20110114_AABOMM lewis_d_Page_144thm.jpg
69acef7ea30cd307e265e519d36a7da8
f9e28c6f3f1dceed20c1a8a3c5992473429957db
22500 F20110114_AABOLY lewis_d_Page_027.pro
3bcf19dd2aa0b8ff44b895cc3d1731bb
e5a8fdf0daa47c35e06d9f07f53871e7b6b7943d
1406 F20110114_AABONB lewis_d_Page_091.txt
687d2bb3fe46baa360698bf65b0b17c0
e3929ebc793a6919c2f155f1353c26ba6aa670ab
1051817 F20110114_AABPQE lewis_d_Page_106.jp2
146d2dc4dccbaa636a7d7434ea7c816f
b350e65012ce249e6abd54b930784347e11415bb
79681 F20110114_AABPPP lewis_d_Page_038.QC.jpg
b877ee46049829c6e8d2856e4f7f75cc
292a126c90b415256ace5e108a7152fd27488a15
F20110114_AABOMN lewis_d_Page_147.tif
d5ab3055028fe13b57c3835de809da09
4b98bb502a0117a9a21d8ffce21b0c12542525bc
950 F20110114_AABOLZ lewis_d_Page_054.txt
f0492441392a758866522fc9989ce6f1
d348937c57d9f27ac18be5edb4fa4e35b08a10f4
40906 F20110114_AABONC lewis_d_Page_094.pro
5c414b7527db73223badbab251bd3afe
4840f2211d2555b84381c8adbebf473fbc7721d4
418 F20110114_AABPQF lewis_d_Page_128.txt
67ca5840daab08a61c673758b966116f
e0820fd81d35422de6b9ae140483779ece9aa0a3
266225 F20110114_AABPPQ lewis_d_Page_089.jpg
abc999bbd588c2349582f2ead99f0cf0
19141a8c810d3de97936bc209afb62adb18f1e63
40152 F20110114_AABOMO lewis_d_Page_106.pro
39bfe3bb9e2b8706644c407b7001de89
8bb6bb131b6b703dc3d42a19894f447dbd2af3ad
1246 F20110114_AABOND lewis_d_Page_101.txt
e0df738920ab347057db3bb51feeeaad
e5e4c775f792406895d9176f4dcf6bd1ba9419c8
89664 F20110114_AABPQG lewis_d_Page_042.QC.jpg
34994cbbbc60218fd8a366a167c08028
525f3e3e4f1a4def515506f23aa4415d3f604871
107695 F20110114_AABPPR lewis_d_Page_030.jpg
f4abf9ca7dfdfb625ce9e4ffb7e78d3f
104c7619cd302a7da67ccd53c3d8728a56f6bcbe
1744 F20110114_AABOMP lewis_d_Page_113.txt
7bdd1f44916a7c5d3467a3f08294423f
82d133b899e4756e05160deac9fd8f97040f4001
192291 F20110114_AABONE lewis_d_Page_020.jpg
bda692543706e724051a68cbd6956038
144c37d450cf7c6af91c93ef559bec9b5aea7056
190508 F20110114_AABPQH lewis_d_Page_090.jpg
954ac606baaccab6a64150f16cf9d1c0
ac7150c0877de82a782601aac41b40f5b5beb061
F20110114_AABPPS lewis_d_Page_074.tif
b0c786f4a2aec4dd5f4a7a6bfb54ce16
01090e50b14f9f091c7a29ff65f734b1f2edefa0
F20110114_AABOMQ lewis_d_Page_141.tif
74693dcc306e9eb2ff50cb5da8c9dd0f
e1ade5210b8af421bfd639960b9dcf42d4632d7f
17086 F20110114_AABONF lewis_d_Page_131.pro
2785da178cfecc8ad4dd10bfebf94bfa
066db5ce0a5a5e348cf7de96ae7aaea24559db1d
1051985 F20110114_AABPQI lewis_d_Page_010.jp2
1ae7486ba5b677655102f92e0953fdd3
add61be1c3ae552138ff98e8fd71c2b9e5cbe28d
50464 F20110114_AABPPT lewis_d_Page_087.pro
597e5f95eed4a2316907fb51592ca962
5322a8dd58422330dc59113a7648f9eb4796c580
160703 F20110114_AABOMR lewis_d_Page_100.jpg
68a1cc687bf93b39dc9cae098dade624
14201cbffcc3867a1142337a18e488d11dd9939b
83006 F20110114_AABONG lewis_d_Page_024.QC.jpg
47a63f54ad78e5b71d16e7913391cb70
34af8a3aab9018c3b4d0be98aaa6e9347f954799
96167 F20110114_AABPQJ lewis_d_Page_017.QC.jpg
ba81519c02becfa6e28eedc9bbba1577
890419b09773776221a8785cbdf80b5ac0b83ec9
76787 F20110114_AABPPU lewis_d_Page_130.QC.jpg
12b90d61ff4850e4c4e802673b2a1a23
b0cc927746218166fe545ff27147337074de18a7
81376 F20110114_AABOMS lewis_d_Page_088.QC.jpg
34ff1d4dc0c21cbfa34936a6f61fdb28
a4c0f13890823b62dba7d8aee73c557b6fcf37e3
34103 F20110114_AABONH lewis_d_Page_016.pro
f626f0569514379228220644f7f14d80
6b9eff6dc8e558e6c6f7b56655dfcb9a0bbed81a
516 F20110114_AABPQK lewis_d_Page_060.txt
6f2fde7afbeba6e0271139413d2991c3
ad134bcbea94d45080d6e759bab5eef10c5ce02f
1337 F20110114_AABPPV lewis_d_Page_098.txt
7e0a761c32a6eb346f521242d769986d
20cfd0c78b1fa1f2a06332e64b48aaabc1d57b7a
724808 F20110114_AABOMT lewis_d_Page_027.jp2
c1b3d895ee2a3f9022b595c6bdfcfb5f
54a26544efc95c6ee38c832e27d488ca311dce2c
44127 F20110114_AABONI lewis_d_Page_048.pro
06d271609dd5cc5ad4adf5eb25362a99
a6ca0203d11f045001ca05a00701154cdc645e44
856 F20110114_AABPPW lewis_d_Page_077.txt
acfa0e494686a8c8a18b87c3bbfd08ec
bab833d7c783749e17dc2f9fdca0259abf6bea26
20563 F20110114_AABOMU lewis_d_Page_096thm.jpg
aac266ea2a821e73977a26c3cc8a2b9c
50272ef68d2bc3f93fc7696357026051cf56fbab
20802 F20110114_AABPRA lewis_d_Page_074.pro
3d0a7c59a93dfb41db008ef0997fb812
cb2f2d3d81fb87cb1b981b2d81fc76a1b3920636
64379 F20110114_AABPQL lewis_d_Page_128.QC.jpg
8179e79e259c67ddb03ae2854e6a1148
1c9735596162ce2a5291725a888ea0c17647c054
70483 F20110114_AABPPX lewis_d_Page_048.QC.jpg
14cd2254ec84baa5204ebac275476f46
1cd71e5c92823984cd23ffa2ac7eae6922d4745e
224436 F20110114_AABOMV lewis_d_Page_028.jpg
343b84ade8327a502860f58d680ca1f7
57cbe0aff80b6dd967a16a4393769ebec94b873e
F20110114_AABONJ lewis_d_Page_111.tif
4f4472117494a8bdffa8f4b53941eb07
0f5483efdc277f1f8da63c37de17129f4c2f439b
F20110114_AABPRB lewis_d_Page_038.tif
505a73a3b6edd11e9a75d8bdbb7d9eec
bb1b673ed9c7124ff6b1c22e092b45ef1363a44b
94362 F20110114_AABPQM lewis_d_Page_113.jp2
d0986d6c1a411adda51e5e4145b2382b
fe1619af22fe72430353570d1549f77efc4638af
F20110114_AABPPY lewis_d_Page_073.txt
2d203ea632efb32bcd1cfe31e44d22f7
113c6d683fb575a22c0bd1d3e37b2b18821ff3cd
17134 F20110114_AABOMW lewis_d_Page_039.pro
c99ad7a213c219d672fe3ebbe2f6282c
7100ab57da38c9797c8183af94e7c6754d23aa17
42105 F20110114_AABONK lewis_d_Page_107.pro
c7927a5769833007707dcc77ab6bcf3e
82b1d2fc0d9c1558d83c0d1670c4b069dda20b42
8090 F20110114_AABPRC lewis_d_Page_085.pro
1f342c91a8091cec8450bf805e4ede8e
0ae1d3e5455d7a8bc96c3ea46f799bc7b08834e6
89587 F20110114_AABPQN lewis_d_Page_147.jpg
8acde53ee7b8f34b760f5104e800a665
c0c27a81acd11316966fef89c2e944e9950ed9dc
F20110114_AABPPZ lewis_d_Page_096.tif
ed1ba3b89a2373fe5650a5073cee19aa
ade19acff3ac15835f5560694515886a4e00d203
3072 F20110114_AABOMX lewis_d_Page_105.txt
a29b7ded3451015f27ddb516c7d0cbf4
67a64f01359a32f96b1b7de8ea5aa7dfbed87a45
93959 F20110114_AABOOA lewis_d_Page_072.jp2
5e306c03cc5d9699033aea62b19fc4fe
67965b52889312e7032df1c042bc7fbcf22d7da2
60797 F20110114_AABONL lewis_d_Page_025.QC.jpg
07205d4455804b2b9c9268e9b3614179
e5ecb21c3f0746ade0bff932ec49ae65d9cb0fe1
F20110114_AABPRD lewis_d_Page_086.tif
3102247cbb22c607ff75010228b919ee
fd4bd787b11940c144b0fe0e27f65acb76028e5c
66744 F20110114_AABPQO lewis_d_Page_107.QC.jpg
82d0cdf4983bb044a305d99ea1247a72
d3589dd186a5909522f0ddd37f6ad989d40fdc95
176353 F20110114_AABOMY lewis_d_Page_072.jpg
e9be2d4641af3814a7c74b0c8a9a8384
e31ff30a27a981a4cd890b5a87fb5f19e2f4cfed
F20110114_AABOOB lewis_d_Page_082.tif
20beca4d947bbe9f4e5ac5ffc1cf1a92
50733add9e8d8d5973f31c1f4fa950d27e290e8e
26019 F20110114_AABONM lewis_d_Page_014thm.jpg
ad1b363e2c33bb07b5ac0de9f8f847b1
16e5bbe25f262cf1ec0fd5315dd1d817fca1b9d4
2033 F20110114_AABPRE lewis_d_Page_150.txt
202274a09df07cd8d814e35d9554fd6d
ad539f657a8b6a1fb93da04f6e32dbaf65e04795
61860 F20110114_AABPQP lewis_d_Page_094.QC.jpg
6e885e583a56ecee5958d5084eed9e78
1643b975ec75c882f157d7831c968de689d7038d
1856 F20110114_AABOMZ lewis_d_Page_112.txt
98aab4eb6dcb9d75356ae2dcff06c440
844280e30b3a8dad93c63b303fb18a8344b8f260
88131 F20110114_AABOOC lewis_d_Page_007.QC.jpg
3f944f88339f3cabaf430247671605dc
e6e64e79c37538792c7746257b5fdb8c456b013b
45445 F20110114_AABONN lewis_d_Page_018thm.jpg
4449fa6115f6d3ca6a2b0cb87c49c89e
2a3cbdbd60cb2abf4404ca8b2f14732f75f4ca42
633 F20110114_AABPRF lewis_d_Page_124.txt
21f8fd98a00821d24bee647eb42c423b
86142fc1252ff89eb0419a267cf625dde0e78f00
680 F20110114_AABPQQ lewis_d_Page_045.txt
1ab34c0087dc9fd6ddfe02fb87a7c1d7
de9400530a4fd56abb192fdb7b64bb04637b2e78
37854 F20110114_AABOOD lewis_d_Page_012.pro
81e3517076b07059a8ed7b2a2df86bdf
ad8a77686b7b718664b8fc493ccafa7e86f2416e
48250 F20110114_AABONO lewis_d_Page_103thm.jpg
e098142d2b11685bc9b759f85ead9d4f
5edfc2271708a61fd6e2659492ca66d4f64ef38d
27616 F20110114_AABPRG lewis_d_Page_083.pro
51b38ec768e8a2acb63d0164c6314421
28138e0e992c6f7489caa673427ff9ed1b1be6b4
28017 F20110114_AABPQR lewis_d_Page_153.jpg
d132dfa8115af66a6ec596c42f9e1b9d
1e6f4c2f4b791a4ebe1da3480b29688bd051e401
99832 F20110114_AABOOE lewis_d_Page_111.jp2
3ee5cf0e3536ab5ae35a28a2c5e161dd
65cbb1ba95db7526e7bbdaee2ad94d80934b3766
21100 F20110114_AABONP lewis_d_Page_145thm.jpg
2a93a6bdd861b0cb3ab75890e6225290
8b43aa8c904e28852f47832fa63507ed8d77fb63
41056 F20110114_AABPRH lewis_d_Page_134thm.jpg
ad58448cc6f282db09ae21c778a55b53
b3c716ad3217af9357fac5f3ceb8a5f012eea85c
173882 F20110114_AABPQS lewis_d_Page_083.jpg
b82054ee03ff026915e90379d8f467b0
d91bb97b2f38aa38701b4f70f722c03e56bbd3e9
45482 F20110114_AABOOF lewis_d_Page_135thm.jpg
9af147fdc7f8ec1a544701611106f8c5
b6ee4f6e838065d103f48e5fbce89828f98a8c31
102836 F20110114_AABONQ lewis_d_Page_005.QC.jpg
c4cab82134a8a06d68e49ecd02964409
e6188f6607d493b5e6287cbf009547d15342eb63
F20110114_AABPRI lewis_d_Page_108.tif
f0ef0f74d842ced9f5a13f071d883ffb
aa598483961aa2043369337532eddc46879a8d57
379293 F20110114_AABPQT lewis_d_Page_076.jp2
78a2a4bb8e0e4322dba4c919de698633
4d4086908e2126a3cae50910f06ddbcd2990c52e
170874 F20110114_AABOOG lewis_d_Page_148.jpg
76f4c0ba28cd5db912bdb03978f5fc98
b8fdad807d9fd16f8d5562d2af33ad34e2e056e3
41758 F20110114_AABONR lewis_d_Page_148.pro
939821fb1adf9a3d8c77e068d3a884a5
4f586f71ada40f7f40c15558f2ccac4a4b767f5b
30686 F20110114_AABPRJ lewis_d_Page_101.pro
4667e9444f3c9da9a37a881d80205aea
9fff66881860e71b474bd1e71601464fb9d3796c
84620 F20110114_AABPQU lewis_d_Page_145.jp2
6207b106591ca4329d4e652fdecece46
61c44083610ce9a1af77b07167c7c5bac2647e79
91354 F20110114_AABOOH lewis_d_Page_057.QC.jpg
6c0497331f36cf0990a628b4b5a44a39
539bff5b85b882c7309cae212df63a977e954b27
851243 F20110114_AABONS lewis_d_Page_026.jp2
b7d11089532b32a2857be07739451655
8088b8a547edd1c1e9af83a5d77f5d2b0b18c167
20717 F20110114_AABPRK lewis_d_Page_146thm.jpg
92f735d1c484962107076013f111f4ae
4747dc667bcb0e946c8c5ebaaf118d739f2baebe
88890 F20110114_AABPQV lewis_d_Page_127.QC.jpg
c2b4d2b5a0671e3b18d3010598a37aab
e1fcced0e127c5804a8db62ce40a5c19a0a5c035
129737 F20110114_AABOOI lewis_d_Page_037.jpg
79c4b8c9165b263c836ec06345fdf8ba
67624efefcfe025c10d31ac188da23801bc0e89f
72025 F20110114_AABONT lewis_d_Page_062.QC.jpg
f977c7e47392c21d85ba72218445401d
fecf51d8437e809dea551b50be063f4bc9671f81
1051954 F20110114_AABPRL lewis_d_Page_089.jp2
648b849795734b511405e3ab05a86dce
9e195dd12b822f8184f019965d76a1e03b570e84
156041 F20110114_AABPQW lewis_d_Page_027.jpg
0bea87eb7685733431e465084761e26a
50c7882eabff1e21c0c1ff88b044811e5465436e
F20110114_AABOOJ lewis_d_Page_139.txt
4c2ff49097c202adcb05b3c8d36187f6
b0f2f6fb2c19e5cb72ecc48f299efc318abc974a
1047183 F20110114_AABONU lewis_d_Page_034.jp2
f23e816451a142f3822ce8d8b0d692ef
043740541f16cc71d8f965bdc4f14bb09658d793
884079 F20110114_AABPSA lewis_d_Page_053.jp2
2c75c2a90529a0fba207d14c12866db7
6f93391785fef81517c4e8d39e85cbcebc1784cc
821261 F20110114_AABPQX lewis_d_Page_083.jp2
f4e744943779510404cc44348abbc777
102a48f3369d8e232064d2c9e962151371d18e33
F20110114_AABONV lewis_d_Page_069.tif
f5e164a6a2f73e0192232f236d370633
b10c8c147b67c1206c5693237b9520b1fee7640a
F20110114_AABPSB lewis_d_Page_039.tif
17b909117489d079d91410a19a8f8bb7
ca352ef2d02b46695d612af00ad9e41713841104
899741 F20110114_AABPRM lewis_d_Page_054.jp2
dd4c77c77a423c0c96b8a42082707139
16c6de4e75bce3c0e67c517896c5bdb989e3c4db
F20110114_AABPQY lewis_d_Page_133.tif
4e762414c9b57318f434bb8e73450de1
321c96a0dafe533ad6e73d9602d91cf87de440be
754997 F20110114_AABOOK lewis_d_Page_136.jp2
de1533a39d471c52be372f21f9ec21e4
b0d7a2b61e6a0c9d403489b0ec4f8cc8c02a9742
43891 F20110114_AABONW lewis_d_Page_095.QC.jpg
7efce70f2c90ed44e04bb50b0c0a1567
3ec20c051885d5f4ce2f23eea6588db0714ca2ae
2197 F20110114_AABPSC lewis_d_Page_086.txt
800ce7939a81edc8f4f921a01859d664
2428a806b89ca7a83958d381671dc01ce6a2c9c6
40013 F20110114_AABPRN lewis_d_Page_142thm.jpg
59c7331eaed3e1ad8af1ed97e4aa5760
9a322e6e5cc90a4963de58348eecafc7d6302110
191093 F20110114_AABPQZ lewis_d_Page_127.jpg
d606d4e228e4f8fb0fc223d0b17268e8
6ac4e81a25de419bb940c3f3493a3ec0ef627200
81578 F20110114_AABOPA lewis_d_Page_091.QC.jpg
49897fcc2fefbbb8c345b3c29f494cd7
6e492d8d43b6393a274767f60481c1c0927f813f
F20110114_AABOOL lewis_d_Page_033.tif
4151b0eb8ca710cb2985a04e9c41a58f
4a352752132a289a5cd159c899728eb6800b7800
150267 F20110114_AABONX lewis_d_Page_126.jpg
f37bd72ec447f53cac471cf9acfe3cdc
0b645fcf638702a7d05d3976ac091805c74ceba9
796 F20110114_AABPSD lewis_d_Page_070.txt
a40521343ea37fe424bd6bbc1f1e6ecf
b8361d34c57a173ef3bfc6fe2caf0e8dd8cd1d53
85308 F20110114_AABPRO lewis_d_Page_020.QC.jpg
5b326c9be5c12cc235797776a96301db
32922e6b48c01f84e19004419912cd8815aeccfe
1127 F20110114_AABOPB lewis_d_Page_059.txt
19fe99d79b4d94e85760629af2163a7b
2f5e50ae6825049039cbf0844d191d79ae1af675
F20110114_AABOOM lewis_d_Page_152.tif
9ad63617ecb1690a535bc9c4fcbb7106
db51c96a8bf7ec29ce5af204106594cc123e26fd
258 F20110114_AABONY lewis_d_Page_143.txt
0577a68ef572680290581a30f4aec33d
e58abdd5c442965d8b52064f8d65fe34d1deee4f
110836 F20110114_AABPSE lewis_d_Page_110.jp2
276340d2cf6a12ac63d87f0a96f5b2e0
ed30670c605acec673230670791073c1195eb310
81110 F20110114_AABPRP lewis_d_Page_140.jp2
8c692d33ba711efd4e3b4799a395273c
6f6a0ce2aa8abe2271e5fef40215ec359a81d5f7
93509 F20110114_AABOPC lewis_d_Page_151.jpg
037bb90aedfb121ab0b440b0a2922ff1
18be1c75d4a0de31aa7f85bf45d3a79bd9b6097f
2018 F20110114_AABOON lewis_d_Page_029.txt
f4dd63ad7273813ab6dafd065db78472
6a0e4a588705941c4c7de0ea3a90238faaf883b8
509 F20110114_AABONZ lewis_d_Page_061.txt
42ae594314c73c9b59c1264a848fa006
7a18cb848f8bcf26a7c44516b22aefe1bc72e7fe
F20110114_AABPSF lewis_d_Page_088.tif
bcc4d317737a08d33cb4e2abb512bfaf
87dd171ae02d50d103230aed467f8879b1248929
909774 F20110114_AABPRQ lewis_d_Page_038.jp2
d84d992392bff6e6ad4f833ef2cdc72b
5a01ec02721fd85487b130d2500ee2f4b50fcae0
1528 F20110114_AABOPD lewis_d_Page_034.txt
deac5aecf8113e5752df31ef162c46b4
31efa4e39bf045ed6c548e96aaa002c00b64f874
F20110114_AABOOO lewis_d_Page_080.tif
4538bc990c95c241ce1fad6830e55241
1cfb0a57f3bd2c1d5bd66114522d0f3790eeca6b
1188 F20110114_AABPSG lewis_d_Page_015.txt
d2fe2664a613dd9a0ad4bc39f49bedb2
4c93a91726993ee9af44702bc218638a7280f389
45055 F20110114_AABPRR lewis_d_Page_054thm.jpg
e6335070c2ba092c147c6663842df45a
e4446658e5e75d5c04c6ab96b0198f132bc8f3d8
40507 F20110114_AABOPE lewis_d_Page_125thm.jpg
b25a897e661fbcf270d46f071228ac27
029831e2c4d29fffb4b422e7a0f223d53cd7b297
F20110114_AABOOP lewis_d_Page_139.tif
d03944f43e8c8d38ba2a7843d7ca4a53
e70c733091dae6a62a114a8cc4bbcb5c0613f35a
56111 F20110114_AABPSH lewis_d_Page_003.jpg
71e224c73d88b6931161382e6e777e21
efed042f7e3de0fb5f28655191b18006670225d6
620680 F20110114_AABPRS lewis_d_Page_041.jp2
a591f68d3fd51569e4768d41345fe3db
6d7784c4b484085ad1a82d1bd35b1d1a42343e49
155742 F20110114_AABOPF lewis_d_Page_062.jpg
bec7a81f0723b57c2e52c7fb1a0280ea
fa303f28841d0216ca7f1371c6e2ae82431f3df7
2638 F20110114_AABOOQ lewis_d_Page_014.txt
4c3ee62eda6f6ac8691306e59036920c
6189c43acd4be3995e18ac0450ea5cf810d34059
F20110114_AABPSI lewis_d_Page_064.tif
f736cab2adb92fb92e6be3362bd0b19b
c4f9d38bfc513c5308bfd6249b846d3cc3cb2e15
149833 F20110114_AABPRT lewis_d_Page_021.jpg
726dd4155aaef8565a03148915cb4046
5129a85c8e57da44c29c6b1c26d046ef109725aa
207734 F20110114_AABOPG lewis_d_Page_099.jpg
d88021351747d77422179d1f29312e90
18e7ce03986eed53c3391aed6b93e0e77bd4bda2
26827 F20110114_AABOOR lewis_d_Page_043.pro
cb88a45b0719dc83be0ed93627dd4a38
a6aa23190e866418576ba9a811a2efa2a104d569
201898 F20110114_AABPSJ lewis_d_Page_123.jpg
9b0e5a25e797c8b5cbe35359f4a28495
20ed38f77ec76982551e0d249216a453bcc35d1c
43446 F20110114_AABPRU lewis_d_Page_117thm.jpg
1853cde223065f184138f56cbfa8c170
c7111e75c50737d0555dd4f6b4fd63d1865c28b7
176373 F20110114_AABOPH lewis_d_Page_031.jpg
f90485a7c781cc73994c7c127f1d88d4
c806b30d52012a232761023929f0ff2cde93a380
23777 F20110114_AABOOS lewis_d_Page_079thm.jpg
b8c462d11966032273cfcc570293a6cf
26fcc40d122925761871bf29445cd6f084d19a0d
82174 F20110114_AABPSK lewis_d_Page_083.QC.jpg
67a0dbc700e7a3ad4eac424de2edda5e
1d5fe5e8e0afc2ccd7aa654cb26ad04733d5c0db
632696 F20110114_AABPRV lewis_d_Page_037.jp2
979337fc831a7141accba4aceab36f81
e6ac2df84828cb7efc75db97408e067e88066eff
70233 F20110114_AABOPI lewis_d_Page_110.QC.jpg
fab2e5e7c0430c1c51e45e1cf31fb7d8
030f658e66fbbf8d88ac34951549e2be7659b72d
45698 F20110114_AABOOT lewis_d_Page_099thm.jpg
7df27225290ec2dc4a4bb9fd0c5b0603
11a696307e617710378fca3e17a68dc3af697f4e
983816 F20110114_AABPSL lewis_d_Page_042.jp2
4ec10e1580263d9d751161696b67060e
82e4f81acf8d065af9f909d8bd4bdba4964c2237
31375 F20110114_AABPRW lewis_d_Page_089.pro
f1824a3e9fb6d689c065307788eea6e0
40ea00f0b059fb0223ad04b0231cd114b7b87b93
415 F20110114_AABOPJ lewis_d_Page_063.txt
edae5fa3ed60211ab00d5f7f1acb0ce7
c35e0bbb3e403311044f4a63d1994beceedf3295
161767 F20110114_AABOOU lewis_d_Page_045.jpg
fd292eedbf05830ffee1db56e644eeda
ec96f89996fe9feff5c735e06943ada6d4bfb30b



PAGE 1

THREE-DIMENSIONAL VISUALIZATION DISPLAY DEVICE By DAVID LEWIS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by David Lewis

PAGE 3

ACKNOWLEDGMENTS I would like to thank Dr. Carl Crane for the opportunity to work at CIMAR, DOE for their sponsorship, Shannon Ridgeway and Jose Santiago for their advice, Kayla Spellman and Matt Borondy for their superb craftsmanship, my parents for their encouragement, Lora Gay for her moral and technical support, and most especially Eric Spellman for always having the time to help with any problem I had.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF FIGURES..........................................................................................................vii ABSTRACT......................................................................................................................x ii CHAPTER 1 INTRODUCTION.........................................................................................................1 Mayer’s 3-D Display Device........................................................................................2 Similar Projects.............................................................................................................4 2 PROJECT APPROA CH AND OUTLINE..................................................................11 Objective.....................................................................................................................1 1 Goals.......................................................................................................................... .11 Evolution.............................................................................................................11 Reevaluation........................................................................................................12 Resulting Machine......................................................................................................13 System Architecture and Overview of 3 Systems...............................................13 How It Works......................................................................................................16 3 PHYSICAL DISPLAY................................................................................................19 Skeletal Design...........................................................................................................19 Pin Motion Inhibitors..................................................................................................20 Possible Actuators...............................................................................................20 Miniature rotational solenoid.......................................................................20 Magnetorheological dampening...................................................................21 Shape memory alloy tubes...........................................................................21 Shape memory alloy wires...........................................................................23 Clamp Creation....................................................................................................24 Home design.................................................................................................24 Pencil clamps................................................................................................26 Evolution of Design.............................................................................................28 BioMetal Micro Helixes......................................................................................31 Hole Pattern................................................................................................................35

PAGE 5

v Pin Sensors..................................................................................................................35 Hall-Effect Sensors..............................................................................................35 Contact Traces and Steel Wool...........................................................................36 Trace Pattern and Board Creation...............................................................................37 Pin Movement System and Frame..............................................................................42 Motor Selection...................................................................................................42 Solenoid Selection...............................................................................................44 Shaft Selection.....................................................................................................46 4 CONTROL BOX.........................................................................................................60 Raster Scanning..........................................................................................................60 Mapping......................................................................................................................62 5 COMPUTER................................................................................................................65 Overview.....................................................................................................................65 Graphical User Interface.............................................................................................66 Path Planner................................................................................................................67 Digital I/O...................................................................................................................6 9 6 ASSEMBLY................................................................................................................70 Control Box................................................................................................................70 Pre-Clamp Physical Display.......................................................................................71 Frame...................................................................................................................71 Motor/Shaft Systems...........................................................................................74 Brackets...............................................................................................................75 Plates....................................................................................................................75 Solenoids.............................................................................................................75 Pin-Clamp Systems.....................................................................................................75 Mechanical Pencil Disassembling.......................................................................76 Brass Cup and Wire Wrap...................................................................................77 Diode Curling......................................................................................................78 Guide Tubes.........................................................................................................79 Steel Wool Attachment........................................................................................79 SMA System Assembly.......................................................................................80 Final Attachment Process....................................................................................81 Elimination of the Fan................................................................................................82 7 TESTING AND RESULTS.........................................................................................84 Tests.......................................................................................................................... ..84 Clamp System Tests............................................................................................84 First SMA test..............................................................................................84 Micro helix tests...........................................................................................85 Electrical Tests............................................................................................................87

PAGE 6

vi Raster Scan Test..................................................................................................87 Sensor Tests................................................................................................................88 Hall Effect...........................................................................................................88 Contact Sensors...................................................................................................89 Testing the Steppers, Driv ers, and Digital I/O....................................................90 Final Results...............................................................................................................91 Physical Visualization.........................................................................................92 Quick Fixes..........................................................................................................93 Virtual Visualization...........................................................................................94 8 FUTURE WORK.........................................................................................................98 General Changes.........................................................................................................98 Major Changes............................................................................................................99 Clamping.............................................................................................................99 Sensing..............................................................................................................100 Conclusions...............................................................................................................102 Opinions and Personal Advice..................................................................................103 APPENDIX A DEVELOPMENT OF THE FIRST PIN CLAMP.....................................................104 B THE RASTER SCAN TEST REPORT.....................................................................114 C ODDS AND ENDS....................................................................................................121 D SPECIFICS OF WIRING..........................................................................................132 E PROGRAM CODE....................................................................................................136 LIST OF REFERENCES.................................................................................................140 BIOGRAPHICAL SKETCH...........................................................................................141

PAGE 7

vii LIST OF FIGURES Figure page 1-1 Jeremy Mayer’s 3-D display de vice (translating pins not shown)...........................3 1-2 Mayer’s 3-D display device had eight phys ical pins and eight virtual pins that could mirror one another..........................................................................................4 1-3 The LeviTABLE......................................................................................................5 1-4 Tactile shape display using RC servomotors...........................................................6 1-5 Servomotor positioning............................................................................................7 1-6 The FEELEX 1........................................................................................................8 1-7 The FEELEX 2........................................................................................................9 1-8 High Density Tactile Display...................................................................................9 1-9 Refreshable Braille Display...................................................................................10 2-1 The 3-D visualization device.................................................................................14 2-2 System architecture................................................................................................14 2-3 Control Box............................................................................................................15 2-4 Physical display.....................................................................................................16 2-5 Graphical User Interface........................................................................................17 3-1 The initial skeletal design......................................................................................20 3-2 A miniature rotational solenoid.............................................................................21 3-3 The SMA tubing is bent and then returns to its original shape..............................22 3-4 Attempts at cutting SMA tubing resulted in irreversible damage..........................23 3-5 The clamping process............................................................................................24

PAGE 8

viii 3-6 Initial clamp shape.................................................................................................25 3-7 FEA analysis for material choice...........................................................................25 3-8 Manufacturing of a clamp......................................................................................26 3-9 Example clamps from mechanical pencils.............................................................27 3-10 Two clamp systems................................................................................................27 3-11 Clamp actuation testing model...............................................................................28 3-12 The first hole pattern..............................................................................................29 3-13 Close up of electrical connection...........................................................................29 3-14 Plate arrangement...................................................................................................30 3-15 BioMetal micro helixes..........................................................................................31 3-16 Helix model 2.........................................................................................................32 3-17 Multiple pin clamping model.................................................................................33 3-18 Final pin-clamp system design...............................................................................34 3-19 Hall-effect sensors.................................................................................................35 3-20 Steel wool attached to the bottom of a pin.............................................................37 3-21 Digital bottom plate...............................................................................................38 3-22 Digital ground plate...............................................................................................39 3-23 Digital power plate.................................................................................................40 3-24 Digital support plate...............................................................................................41 3-25 Digital top plate......................................................................................................42 3-26 Thompson 55M048D1B Bi polar Stepper Motor...................................................43 3-27 Free body diagram of forces on the screw.............................................................43 3-28 Magnet-Schultz DC Voltage Cylindrical Push Solenoid.......................................45 3-29 Shaft / motor system..............................................................................................46 3-30 Two-dimensional drawing of th e shaft/ motor system alignment.........................48

PAGE 9

ix 3-31 Skeletal structure of 80/20.....................................................................................49 3-32 Assembled 3-D display frame with fan.................................................................50 3-33 Motor mount..........................................................................................................51 3-34 Upper bearing mount.............................................................................................52 3-35 Push solenoid mount..............................................................................................53 3-36 Preliminary assembly drawing...............................................................................54 3-37 Top plate bracket....................................................................................................55 3-38 Ground plate bracket..............................................................................................56 3-39 Power plate bracket................................................................................................57 3-40 Sensor plate bracket...............................................................................................58 3-41 The fully assembled physical display....................................................................59 4-1 An example 3x3 grid of actuators in a raster scanning sequence..........................61 4-2 An example 33 grid of sensor s in a raster scanning sequence.............................62 4-3 Direction of signals through the control box.........................................................63 4-4 Basic mapping of the control box..........................................................................64 5-1 Computer setup......................................................................................................65 5-2 Graphical user interface (a) Example GUI display, (b) GUI reading a file...........66 5-3 PCI 7296 digital I/O card.......................................................................................69 6-1 The sections of the frame where the red circle encompasses a waste section, the blue circle encompasses an arm section, and the green circle encompasses a leg section............................................................................................................72 6-2 Waist section of the frame.....................................................................................72 6-3 Legs section of the frame.......................................................................................73 6-4 Arms section of the frame......................................................................................73 6-5 Bins of separated materials....................................................................................77 6-6 A diode before assembly and a diode after assembly............................................78

PAGE 10

x 6-7 On the top-left is the heat-shrink tubing, towards the center is the clear tubing, and on the right is the assembled guide tube.........................................................79 6-8 Physical display with fan.......................................................................................83 7-1 Clamp actuation testing model...............................................................................85 7-2 Micro helix tests (a) Helix model 1, (b) Helix model 2.........................................86 7-3 The raster scan test setup .......................................................................................87 7-4 Sensor test setup ....................................................................................................89 7-5 Stepper motor wired to a driver.............................................................................90 7-6 Path planner test setup...........................................................................................91 7-7 Results of a physical visualization test where (a) is desired product that was formed on the GUI and (b) is the final product produced by the physical display....................................................................................................................94 7-8 The four possible arrangements for incorrect sensor readings..............................96 7-9 The results of a virtual visualization test where only the front row was tested; (a) shows the initial arrangement of the physical pins and (b) shows the virtual results.....................................................................................................................97 8-1 Pulley system of clamp actuation where the left block moves with the pins rows and the right block move s with the pins columns......................................100 8-2 A two-level sensor reading system in a raster scan arrangement where the red ovals represent sensor holes, the blue squares represent the single-level first read points, and the green dots represent the multi-level set of second read points....................................................................................................................102 A-1 Initial clamp design..............................................................................................105 A-2 FEA analysis for material choice.........................................................................107 A-3 Node placement on beam analysis for Nub 1 placement.....................................108 A-4. Node placement on beam analysis for Nub 2 placement.....................................110 A-5 Completed clamps................................................................................................112 A-6 Evolution of clamp creation.................................................................................112 A-7 Manufactured clamps (a) Completed clamps, (b) Clamp changes through production............................................................................................................113

PAGE 11

xi B-1 Configuration of LEDs.......................................................................................115 B-2 The Wiring Diagram For the LED Display.........................................................116 B-3 The Assembled Electrical System.......................................................................117 B-4 Five LEDs Being Lit............................................................................................119 B-5 Nine LEDs Being Lit...........................................................................................119 C-1 The three hole patterns that were analyzed..........................................................122 C-2 Table Arrangement..............................................................................................123 C-3 Close-up of el ectrical connection.........................................................................124 C-4 Design 1 concept drawing....................................................................................125 C-5 Design 1 detail drawing.......................................................................................125 C-6 Design 2 concept drawing....................................................................................126 C-7 Design 2 detail drawing.......................................................................................126 C-8 SMA spring actuation with inelastic band...........................................................128 C-9 Ideal spring SMA actuation design......................................................................129 C-10 Compressing SMA spring actuation design.........................................................130 C-11 Miniature compressing SM A spring actuation design.........................................131

PAGE 12

xii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THREE-DIMENSIONAL VISUALIZATION DISPLAY DEVICE By David Lewis August 2004 Chair: Carl D. Crane, III Major Department: Mechanic al and Aerospace Engineering This research focuses on the design of a device used for presenting threedimensional images. The machine was built with the purpose of performing two particular tasks. The first task was to be able to adjust an imag e on a computer screen and have a physical display mirror that image. The second task was to be able to adjust a physical display and have the computer screen mirror the image. Such a device would be useful for rapid prototyp ing, virtual modeling, topogra phy, measurement, and touchscreens. The device was built with the hope of es caping the limitations typically involved with virtually creating an object and the time typically involv ed in physically creating an object. This device utilizes three main components: a computer, a control box, and a physical display. The physical display uses a gr id of parallel pins that can be moved in a common direction to create a “moving wall.” Some of the key issues in designing such apparatus include minimizing the power cons umption, packing the pins as tightly as possible, overall use of space, and the accuracy of images.

PAGE 13

1 CHAPTER 1 INTRODUCTION In this age of information, society has placed great value upon methods for transferring large amounts of information as quickly and tightly packed as possible. One of the best ways to instantly transmit a lot of information is to deliver an image. A threedimensional visualization is an ideal way of presenting every aspect of an object in a single, complete image. One of the most popular methods of 3-D visualization is through computer graphics displays. Many of these machines can lead to problems where the user is required to wear bulky headgear and only one user can view the image at a time. In 2001, Jeremy Mayer attempted to solve this problem by creating a machine that completed the following three objectives (Mayer 2001): 1. Presents a three-dimensional representation of the object being displayed 2. Does not require the user to wear any special glasses or head mounted tracking system 3. Allows multiple users to simultaneously view the same object. Such a device will greatly aid an operator (or team of operators) in tele-operation and tele-supervision operations. Mayer’s project aimed to create three-dimensional images out of a “moving wall” where the “moving wall” was composed of thin parallel pins that could be moved up and down individually. He successfully created a prototype display device that moved eight parallel pins up and down. It was the desire of this project to create a machine that would improve the capabilities of the prototype design while being more efficient with its use of size, power, and cost.

PAGE 14

2 Such a machine could be of use in the following areas: Rapid Prototyping: An instant three-dimensional model could be made for an object designed on a CAD program. This would allow a designer to quickly provide others with a model of a product that they can view and walk around. Measurement: An object could be pressed into the pins and its dimensions would be recorded in the process of making a virtual image. These dimensions would be easily accessible for the user. Virtual Modeling: Virtual models could be made of real life objects. Awkwardly shaped containers and objects could be virtually rendered after being pressed against the pins. Then a computer could be used to solve premium packing situations. Topography: The display could be used to show large areas of land. Such a device could be useful to anyone from engin eers deciding where to place a bridge to military officers deciding where to position their men. To get such a tool currently, these people need to have a model sculpted from hand. This device would pair well with a new technology that allows an airplane to reflect radar over a section of land and a computer program uses the reading to render a CAD image of the land. This image would then be easily transferred to the physical display. Picture: This would be a new way for friends and families to send one another images. A grandmother could see the three-dimensional hand print of her newborn grandchild that is hundreds of miles away. Touch-Screen: The display could be adjusted into a grid of buttons and when the user presses a button, the device would be able to detect the adjustment. An image could be shown down upon the display to help aid in the visual, as well. This would allow for various interactions between user and machine that would be useful in such devices as ATMs or Information Kiosks. Mayer’s 3-D Display Device Jeremy Mayer’s machine, seen in Figure 1-1, moved eight pins that were aligned in a row. Depending upon the user’s desire, the pins could be moved manually and a computer image could imitate the pattern, or the computer image could be altered and the pins would imitate the image’s pattern. This is represented in Figure 1-2. The device’s actions were controlled by a graphical user interface (GUI) where the user could adjust the virtual image of pins and tell physical display which task to run. The physical pins

PAGE 15

3 rested upon push-pull cables, which were allowed and prevented from moving by solenoid-controlled swivels. The push-pull cables were composed of a metal wire held within a plastic sleeve by friction. The top end of each sleeve was attached to a stable plate, which the physical pins rose out of. The bottom of each sleeve was attached to a plate, which could be moved vertically, wher e the metal wire hung out loosely. To create a physical pattern of pins, the pins began at their bottom position while all the swivels were activated and pinched the metal wires into a stable position. The moving table could then be lowered by a stepper motor, causing the sleeve to lower, but not the metal wire; in turn, the pins would be raised. When a pin reaches its desired height, the corresponding solenoid is deactivated. This rotates the swivel and the pin is held in place by the friction of the push-pull cable. Figure 1-1. Jeremy Mayer’s 3-D displa y device (translating pins not shown)

PAGE 16

4 Figure 1-2. Mayer’s 3-D display device had eight physical pins and eight virtual pins that could mirror one another To create a virtual pattern of pins to mirror a physical arrangement, the pins are moved to their highest positions and the swivels are opened. The user can physically adjust the pins and the friction of the push-pull cables will hold them in place. Each hanging metal wire is then gently pressed against an individual switch by the springloaded solenoid. The moving plate is translated upward by the stepper motor and the hanging metal wires are pulled up. Eventually, each hanging metal wire will be pulled above the swivel and the spring-loaded solenoi d will push the swivel into the switch. The switch will be activated and the relative location of moving plate will tell the controlling computer where the pin was located. The GUI will do the conversions and render a mirror image of pins. Similar Projects Research was invested into finding similar projects and inventions in existence. Understanding the designs, successes, and failures from these works could help set an efficient path for the next generation display and prevent previous errors from being repeated. The following are some of the most similar and practical related machines that were found:

PAGE 17

5 The LeviTABLE, seen in Figure 1-3, is being worked on at MIT (Kelliher et al. 2001). This apparatus is an “electromagnetic topographical display” which basically means that it uses electromagnets to create a physical, three-dimensional feature. The device contains a 6.5"6.5" actuation platform with an array of electromagnets placed in an 88 formation. Above the actuation platfo rm is a table system containing 64 drinking straws, also arranged in an 88 pattern. The straws, eight inches from top to bottom, move freely in the vertical direction and contain neodymium magnets within them. Varying the strength and polarity of each of the electromagnets in the actuation platform allows one to conform the grid of straws into desirable topographical displays. The display can be changed as quickly as the desired magnetic states can be relayed to the electromagnets and the footprint remains equivalent to the space taken by the grid of straws. However, the straws are limited to 1.5" of movement. The project’s most blatant flaw is that although a theoretical movement tolerance of 1.0 mm is claimed, the creators have not yet been able to implement any more than a movement of the straws from one extreme position to the other. Furthermore, the machine requires a continuous supply of energy for all times of use, and the thickness of the straws along with the spacing distances do not allow for a very "tight" image. Figure 1-3. The LeviTABLE

PAGE 18

6 A Harvard project resulted in the creation of a tactile shape display using RC servomotors, seen in Figure 1-4, with the purpose of “conveying small-scale force and shape information to the tip of the finger” (Wagner et al. 2002). The device is a 66 pin display, with 2mm maximum displacement and is an "attempt to realistically simulate skin deformations that occur when interacting with real objects by transmitting smallscale shape information to the fingertip." RC servomotors (small, high-performance ball bearing servos) actuate the mechanical pins as seen in Figure 1-5. Positive features include a low parts cost, simple construction, and the ability to move in time to a specific height. However, the machine also has limited movement and scales in size poorly when an increased number of pins are desired. Also, servos rotate so the pin is being moved in two directions, but since the movement is so small in comparison to the pin size the developers might be thinking the pin crookedness is negligible; however, this will be detrimental if the pin movement is made greater. Figure 1-4. Tactile shape display using RC servomotors

PAGE 19

7 Figure 1-5. Servomotor positioning The University of Tsukubas FEELEX Project aims to combine haptic sensation with computer graphics (Iwata et al. 2001). Basically, the shape and image of an object will change as a result of outside forces on the object. The basic system is comprised of an array of actuators beneath a flexible screen. A projector (72020) shines a twodimensional image down onto the screen, which in turn becomes three-dimensional when the actuators deform the screen. Sensors on the actuators can determine the force applied against the screen by an outside source. The readings can be used to determine a response in the shape of screen. Two prototypes for this system structure were created. FEELEX 1, shown in Figure 1-6, was "designed to enable double-handed interaction using the whole of the palms." The flexible screen (composed of a 3 mm thick rubber plate and a white nylon cloth) is 24cm 24cm in size with a 6 linear actuator array beneath it. Each actuator is made of a screw mechanism driven by a DC motor. The machine is PC controlled with two strain gauges atop each actuator acting as force sensors. The apparatus moves in time, is structurally durable, and has a linear ratio between grid size and footprint size. Unfortunately, the machine has a high power

PAGE 20

8 requirement, poor resolution of shape, and there are shape limitations, as sharp edges cannot be created. Figure 1-6. The FEELEX 1 FEELEX 2, shown in Figure 1-7, was created with the intention of improving the resolution of the previous design so that interaction can occur with the fingers. This system has a 50mm 50mm screen with a desired resolution of 8mm. Because the servomotors used are much larger than the desired resolution a piston-crank system was used to move an array of rods (six inches in diameter) that change the shape of the screen. Force is measured by reading the electric current going into each servomotor. This machines best feature is its tight resolution. Its drawbacks are that the piston-crank system creates size of footprint difficultie s for pin expansion, it requires a high-energy intake, and (like the FEELEX 1) there are shape limitations, as sharp edges cannot be created. The “high density tactile display,” created at John Hopkins University and presented in Figure 1-8, is intended for use in neurophysiological and psychophysical experimentation (Pawluk et al. 1998). It contai ns a 20 20 square arra y of pins. There is 0.5 mm between each pin. The pins are moved by linear motors stacked in layers of 100

PAGE 21

9 Figure 1-7. The FEELEX 2 motors above the array. Each 100 motor layer controls a 10 10 section of the pins. The pins can be moved 2.5 mm lengthwise. The machine has the ability to move the pins in time to any desired position as well as maintain a desired pressure from the pins. This design makes it spatially difficult to scale to a larger number of pins (as shown in Figure 1-8), is limiting on the available length of pin movement, requires large supply of power, and is considerably expensive. Figure 1-8. High Density Tactile Display

PAGE 22

10 Refreshable Braille Displays, pictured in Figure 1-9, were the last similar device that was discovered during the literature search (Gallagher). These devices can be purchased commercially from various vendors. These machines are commonly used by the blind for the purpose of reading text that would be displayed by a computer monitor. The Refreshable Braille Display has a row of "soft cells." Each cell has 8 holes filled with linearly actuated pins. The pins can be moved up or down (using either miniature solenoids or piezo-electric outputs) so that each cell contains a Braille character. The user can scroll through lines of text changing the character displays instantly. The pin movement is limited as they can only be located at one of two positions (up or down). The machines are also expensive, as an average, commercially sold displays run in the tens of thousands of dollars. Figure 1-9. Refreshable Braille Display

PAGE 23

11 CHAPTER 2 PROJECT APPROACH AND OUTLINE The knowledge of other work performed on similar three-dimensional displays helped create guidelines to steer this project. It was important that past errors not be repeated and as much improvement be made as possible. This chapter reviews the project’s initial aims to its final product. Objective This project began with and was always governed by the three main goals that were created and used by Mayer in his project: 1. Presents a three-dimensional representation of the object being displayed 2. Does not require the user to wear any special glasses or head mounted tracking system 3. Allows multiple users to simultaneously view the same object. Such a device will greatly aid an operator (or team of operators) in tele-operation and tele-supervision operations. Goals Evolution For the next generation display, a new set of subsidiary goals had to be created that would make the display more effective and more efficient while still maintaining the original three objectives as the top priorities. A set of secondary objectives inspired by the experiences and achievements of the previous machine were created. These goals were not set in stone, but merely guidelines meant to help provide direction and an ambitious aim for the project. What follows are the original secondary goals listed in order of importance:

PAGE 24

12 Pin Pattern: The most vital modification to the new machine is to go from a linear pattern of pins to a 2-D array of pins Actions: The apparatus will be capable of displaying the shape of the pins upon the computer screen as well as displaying a computer image physically with the pins Pin Grid Size vs. Total Machine Footprint: The device will be designed so that adding additional pins will increase the pin grid size the same amount that it will increase the overall machine footprint. Number of Pins: The minimum number of pins will be twenty-five; however, the pin count will ideally be in the thousands Spacing of Pins: The pins have a diameter of 0.045”. They will be kept within 2 mm (center to center) of one another. Power Allotment: No more than 1.5 Watts will be allotted to the actuation of each pin. Height: The top of the machine will be less than 3 feet high so that a person of average height could look down upon the display. Footprint: The machine will be within a footprint of 2’2’ so that it can be easily moved through average size doorways. Reevaluation Next it was important to evaluate the existing 8-pin machine. An analysis was performed to decide which of the current features should be altered in order to achieve the desired goals for the next machine and alleviate the difficulties from the first one. The following is a list and description of those decisions: Actuation System: Renovating this feature is the most vital action of the project. While the moving table system worked well in the 8-pin display, the push-pull cables caused a number of problems. It wasn’t always reliable as changes in friction occasionally occurred with movement, resulting in display inaccuracies. The rotational solenoids (which supplied far more pressure then was necessary) used for clamping were too large and required too much power to be practical for a machine with an expanded grid of pins. These actuators will have to be replaced with something smaller, cheaper, and less energy draining. Power Supply and Control: Having hundreds of pins where each pin needed to be actuated individually would require a more conservative system that requires less power per pin. A new wiring system will also have to be implemented where the

PAGE 25

13 outputs from the controllers will be activating multiple actuators, so as to minimize the number of necessary outputs. Sensors: The current sensors are contact sensors that provide acceptable values; however, using something more precise, such as optical sensors, could provide more exact readings of the pins and reduce measurement errors. Interface: The interface can be improved with the addition of location axis and data tables giving the exact location of each pin. The tables could be used for both reading where the pin is located and writing where the pin’s next desired location is. It would be convenient to give the disp lay an option so that the heads of the pins are seen as a single skin and a section of a CAD drawing could be transposed onto the physical pin display. Resulting Machine System Architecture and Overview of 3 Systems Figure 2-1 shows a picture of the resulting machine. It is composed of three main systems, which can be broken down into the system architecture that is shown in Figure 2-2. The three main components of the setup are the computer, the control box, and the display device. The Computer contains the Graphical User Interface (GUI), the path planner, and the digital input/outputs. The GUI is a display that will appear on the computer monitor and allow the user to interact with the rest of the machine. The data distributor is a program code that translates data instructions between the GUI and the digital I/O. The digital I/O is a PCI card that sends and receives multiple electric signals from the computer to the control box.

PAGE 26

14 Figure 2-1. The 3-D visualization device Figure 2-2. System architecture

PAGE 27

15 The control box, shown in Figure 2-3, is made up of opto-isolators, power sources and motor controllers, as well as the numerous wires, cables and terminal boards tying everything together. The motor controllers will regulate the speed and position of the stepper motors. The opto-isolators will regulate when specific sections of the display device receive power. Figure 2-3. Control Box The display device, pictured in Figure 2-4, uses parallel pins to form a threedimensional image that is visible to the user. The pins are moved and monitored using hard plastic plates, clamps, and sensors. The plates are moved by solenoids and stepper motors. The clamps are toggled by shape memory alloy (SMA) wires. The following three chapters will go into the details of th e research and development for each of these sections, as well as how they work.

PAGE 28

16 Figure 2-4. Physical display How It Works The machine performs two main tasks. One is that it can be provided a virtual arrangement of pins through the computer and then arrange the pins symmetrically on the physical display. The other task is that the pins on the physical display can be arranged into a particular pattern and the computer will render a mirrored pattern of pins. The first task of physical visualization is performed in the following manner. Employing the GUI shown in Figure 2-5, the user will arrange the pins to their desired locations. This can be done by clicking on the rendered pins with the mouse and dragging them, by typing in the exact height of each pin, or by loading a pre-written file with all the pin heights in it. Once happy with the arrangement, the user will click on the “Send” button of the GUI. At this point the pins will be raised to their top positions by the sensor table and the solenoids will open and release clamps in order to loosen their grips on the pins. Next a new window will appear on the GUI asking if the clamps need to be opened again. If any clamps are still preventing movement of any pins, then the user will continue to click the “yes” button until all the pins are loose. Once all the pins are loose,

PAGE 29

17 the user clicks the “no” button. Next, the table that the pins are resting on will lower all the pins to the location of the height of the highest pin. At this location, the table will pause while actuators are activated to hold all of the pins that have that desired height into place. Once activated, the table will lower the pins to the next highest pin location. The pins with activated actuators will be held at their previous locations and new actuators will be activated to hold new pins at the new location. This pattern of lowering, pausing, and activating actuators will be repeated until all the pins have been moved to their desired positions and the table is at its lowest location. The user can then leave the pins at the current configuration or perform another task. Figure 2-5. Graphical User Interface The second task of virtual visualization is performed in the following manner: The user will click the “Receive” button on the GUI. The sensor table will move the pins to their highest possible position, the clamps will be toggled into a locking position by the solenoids, and the table will move back to its lowest possible position. The pins will

PAGE 30

18 remain held at their highest position and a window will appear on the GUI asking the user to adjust the pins to their desired positions. At this point, the user will physically move the pins into the arrangement he or she wants displayed. Next, the user will click the “OK” button on the new window. This will trigger the sensor board to start moving upward again. The board will move to its top position in a single slow and smooth motion, moving each pin it comes in contact with upward with it. When the board stops, all the pins will be back at their top positions and the GUI will show a rendered image of the user’s arrangement of the pins. The user then has the choice of leaving the machine in its current configuration, recording the pin arrangement in to a *.pins file, or performing another task. Further details on how to use the GUI with detailed flowcharts of the task processes are presented in Chapter 5.

PAGE 31

19 CHAPTER 3 PHYSICAL DISPLAY The Physical Display is the most complicated of the three main components found within the System architecture. This system is a display device that accomplishes the three main goals of the project. The display contains a grid of movable pins that can be adjusted to create a desired image. The pins can reach their desired positions as a result of two methods: the display will receive position commands from the computer and adjust the pins itself or the user will manually adjust the pins so that the display can detect the positions of the pins and an image of the pin positions can be rendered on the computer screen. This chapter details the components of the physical display, how they were designed, and how they work. Skeletal Design The design of the physical display essentially began after a thorough evaluation of Mayer’s project and related work. With a guiding set of goals in place, it was decided that the display device would have the basic design that is shown in Figure 3-1. The pins would pass through a device that would be capable of preventing their motion and rest upon a movable plate. Both tasks would begin with the pins at their highest position. For physical visualization, the plate would lower and the pins would follow until the motion of each pin was prevented. For virtual visualization, the motion would be prevented, the table would lower out of the way, the user would adjust the pins positions, and a separate system would detect the positions of the pins. This meant that three systems had to be

PAGE 32

20 designed: a system for preventing motion of the pins, a system for moving the table, and a system for sensing the position of the pins. Figure 3-1. The initial skeletal design Pin Motion Inhibitors Possible Actuators The entire design of the physical display revolved around the fact that some form of actuator was going to be necessary to either push, pull, or pinch the movable pins in the display. The type of actuator used would have a direct outcome on the display device’s pin-to-footprint-size ratio and its power drainage. The following methods of actuation were initially analyzed and compared in order to find the best method of actuation. Miniature rotational solenoid A miniature rotational solenoid, such as the one in Figure 3-2, could have been used to pinch the pins in the same fashion as the actuators in Mayer’s design except less space would be necessary for the actuators and only 0.3 Watts would be used per actuator

PAGE 33

21 as opposed to the 5 Watts per actuator used by his machine. The worries with this actuation were that as the number of pins was increased, they might be too expensive and still consume too much space. This was a tempting option because the solenoid greatly decreased necessary power and was proven to work quite effectively with Mayer’s design. Figure 3-2. A miniature rotational solenoid Magnetorheological dampening This method would have had the cables moving through a magnetorheological fluid. When a magnetic field is applied to the fluid, the fluid hardens and grips the pin and when the field is removed, the pin is released. This method could not only have been messy, but it could have lead to corrosion and electrical problems. A system using this method could not be thought up where either the movement of pins or the physical well being of the machine would not be harmed. Shape memory alloy tubes This method would have had the pins passing through a shape memory alloy tube. A shape memory alloy (SMA) is a metallic material that holds a specific shape but can be stretched or bent into another shape. As seen in Figure 3-3, when the temperature of the material is changed above or below a specific temperature (this temperature can be hot or

PAGE 34

22 cold and depends on how the SMA was manufactured), the material returns to its original shape. In this case, when the tube is heated with an electrical current, it shrinks and grips the cables. When the current is stopped the tube returns to its original state, allowing free movement of the cables. It uses little power (1.2Watts) and takes up little space. The worries were that the SMAs may take a while to react to the current as well as to reset themselves. There was also the possibility that heat from an SMA with current running through it might affect the SMAs nearby. Figure 3-3. The SMA tubing is bent and then returns to its original shape Because of its economical cost and structural simplicity, the SMA tubing actuation initially appeared to be the best method available; therefore, work was begun to design and create model that would implement it. The design that was created used three stable plates stacked on one another. The top and bo ttom plates would have grids of holes just big enough for the pins to move through freely; however, the middle plate would have thicker holes to hold miniature SMA tubes. A current could be passed from the top plate to the bottom plate to heat the tubes. So basically, the pins would be going through the triple-plate stack, able to move freely, but resting on a fourth plate. The fourth plate would begin lowering and each time a pin reached its desired location, its SMA tube

PAGE 35

23 would be heated, thus gripping it and stoppi ng the pin’s motion. This would continue until all pins could no longer be lowered. This plan had to quickly be abandoned, when an attempt to build a model proved that the SMA tubes were very difficult to cut. Each attempt to cut the tubes reliably resulted in chipping, cracking, and irreversible bending, as shown in Figure 3-4. Because pre-cut tubing was too expensive, a new method had to be sought out. Figure 3-4. Attempts at cutting SMA tubing resulted in irreversible damage Shape memory alloy wires While the SMA tubing was a disappointment, the best route still appeared to be to use SMA material. Shape memory alloy wires have the ability to be stretched to a certain length and then heating or cooling makes them return to their original length (Gilbertson 1994). An initial method of actuation was devised using this material. Figure 3-5 presents that method which would be the beginnings of the final method. Basically, the pins would pass through a clamp and cylinder. When left alone, the clamp would prevent any movement of the pin; however, when the cylinder, which allows the pin to pass through freely, is pressed into the clamp, the clamp opens and the pin can move freely. The cylinder could be removed from the clamp when an SMA, attached to the cylinder on one end and a stable plate on the other, is heated; thus, contracting the SMA and pulling

PAGE 36

24 on the cylinder. All the cylinders could be pus hed back into their respective clamps when a moving plate, with holes allowing the pins and SMAs to pass through, pushes them into the clamps and re-stretches the SMAs. Figure 3-5. The clamping process Clamp Creation Home design With a basic structure for the actuation process established by the SMA wire method, it was necessary to create a more detailed design of the clamp. It was initially decided that the clamps would be designed from scratch and manufactured in-house. Figure 3-6 shows the prototype shape that was created for the clamp.

PAGE 37

25 Figure 3-6. Initial clamp shape A finite element analysis (FEA) was performed using the Ideas software to find ideal clamp dimensions that would create the most effective and long-lasting clamp. This analysis helped determine the clamps dimensions, shape, and material. Figure 3-7 shows an example of one of the analyses. Figure 3-7. FEA analysis for material choice Next, a complicated manufacturing process was developed for actually creating physical clamps and several prototypes were made, as seen in Figure 3-8. Appendix A

PAGE 38

26 has the full detailed report on the shape design, the FEA analysis and the manufacturing process. Figure 3-8. Manufacturing of a clamp Pencil clamps After refining the manufacturing process for clamp production, it was determined that if clamps were being produced every single day at eight hours per day, it would still take anywhere from five months to a year to produce the desired one thousand clamps. Because that was a longer time investment than was desired, the next best option was to use an out-of-house machinist who possessed more powerful tools. When no machinists were willing to take the job, it was decided to explore further areas for clamp production. It was observed that a majority of commercially sold mechanical pencils are composed of clamps, presented in Figure 3-9, with similar size and shape.

PAGE 39

27 Figure 3-9. Example clamps from mechanical pencils Pencil clamps that hold 1.3 mm lead were found to provide the results desired from the original clamp; however, the pencil clamps use a system mechanically reversed to the original design. The pencil clamps place the hollowed cylinder on the outside of the clamp while the original design had the cylinder moving inside the clamp (as shown in Figure 3-10). It was decided to proceed by replacing the original design with the pencil clamp system. Figure 3-10. Two clamp systems

PAGE 40

28 Evolution of Design Now that a clamp design was complete and a basic setup existed, tests were begun in order to evolve the clamp system design. A one-pin model, shown in Figure 3-11, was created to test the method of using clamps from a mechanical pencil to hold a pin in place. This test allowed for design concepts to be confirmed, tinkered with, and refined. It proved that the 3.5” SMA wire could comfortably pull the cylinder the desired distance while the keeping the clamp closed with the desired pressure. It also provided helpful guidance related to spacing issues. Figure 3-11. Clamp actuation testing model The next step would be a complicated design process to decide upon a hole pattern. The goal was to maximize the density of pins without sacrificing the structural integrity of the plates and corresponding parts. Figure 3-12 shows the pattern that was chosen and Appendix C contains details of the analysis.

PAGE 41

29 Figure 3-12. The first hole pattern It was desired to use vented screws to attach the SMA wires to the board as shown in Figure 3-13. Because off-the-shelf vented screws were too expensive, two alternate vented screws were designed that better met the needs of the system. The thought behind this was that it would be cheaper to have a manufacturer produce a large quantity of one of the two designs. Appendix C recounts the design of these screws. Figure 3-13. Close up of electrical connection An extra level of four plates would be added to the machine so that the number of necessary holes for vented screws would be cut in half on each plate. Figure 3-14 shows how the refinements evolved the system’s design.

PAGE 42

30 Figure 3-14. Plate arrangement The five types of plates in the design were as follows: 1. Clamp-holders: These plates have the clamps press-fit into their holes. These will prevent and allow for motion of the pins. 2. Clamp-openers: These plates will open all the clamps and allow the pins and the SMA to pass through without contact, but not the cylinders around the clamps. When the solenoids beneath these plates push them up, the cylinders will be pushed up, the clamps will open, and the pins will be able to move freely. 3. Power-suppliers: These plates will contain the vented screws that supply power to the SMA wires. 4. Grounds-providers: These plates will contain the vented screws that provide a ground to the SMA wires. 5. Pin-mover: This plate will have no holes, but will have sensors attached. The pins will rest on it and be moved by it. At this point, trying to move forward became very difficult as multiple problems started surfacing. It was realized that crea ting the various hole configurations necessary for the multiple plates of the machine would be highly difficult if not impossible to create with the University facilities ( especially those that required electrical traces). Therefore, it was decided to seek help from professional machine shops. Several redesigns of the plates were created to match the capabilities of the different machine shops. It was found

PAGE 43

31 that the plate design was going to have to be compromised or the plate creation would cost significantly higher than would have been desired. Shortly after it was found that making the vented screws would be too expensive as well. BioMetal Micro Helixes The build up of difficulty led to the system’s design taking a step back in order to be capable of moving forward. Problems with board and screw creation led to a reevaluation of the pin clamping actuation de sign. Discovery of a spring-shaped SMA opened the way for new, more efficient designs. Several new designs using this material were brainstormed and compared in detail. These analyses can be read in Appendix C. Some of the better designs were protot yped and tested using BioMetal Micro Helixes, shown in Figure 3-15. The micro helixes are an improvement on the SMA wire because they require less power, take up less space, move a greater distance than wire strands of the same length, and are easier to attach. Figure 3-15. BioMetal micro helixes

PAGE 44

32 The most effective of the designs using the micro helix models is seen in Figure 33-16. The idea here was similar to the initial design in that the clamp was still held by a stable plate and there was a stable power giving plate and another ground plate. However, the micro helix was so small that it could be attached to the power and ground plates with a normal wire which could be soldered or wound for any attachment type desired. Also, the ground plate could double as a moving plate for opening the clamp. This model performed the gripping actions better than the previous design, was easier to assemble and created far less clutter. Figure 3-16. Helix model 2 Unfortunately, the micro helix design proved to be too costly for this project and focus had to be returned to the SMA wire design that was originally in place. Even so,

PAGE 45

33 the micro helix design was very helpful in evolving the SMA wire design. Figure 3-17 shows the three different designs together. With the new design, one end of a strand of SMA wire is attached to a plate below the clamp and the other is attached to the cup and a regular wire. The other end of the regular wire is attached to a plate above the clamp. Power and ground are supplied from the two plates and when heated the cup is pulled down and the clamp is tightened. The upper plate can move upward and pull on the regular wire, which in turn lifts the cup and releases the clamps grip on the pin. Figure 3-17. Multiple pin clamping model

PAGE 46

34 While the model proved this system to be reliably functional for a single pin, prototyping a system with multiple pins and clamps within an accurate environment lead to several new problems. There were numerous locations where short circuits occurred, hand assembly was difficult, and several connections were structurally unreliable. Updating to the design shown in Figure 3-18 reduced the severities of the problems. This final design allowed for the system to be assembled in a more convenient step-by-step process and minimized the chance of short circuits. Figure 3-18. Final pin-clamp system design The use of the shrink-wrap tubing eliminated most current shorting difficulties and actually allowed for a smoother flow of the pin. Using looped and curled wire ends

PAGE 47

35 eliminated the need for crimps and allowed for a simpler assembly and a better situation for maintenance. Hole Pattern With the changing of the design the screw holes were eliminated and it was decided to do a simple row offset pattern in a 3030 array. When it was found that additional space was necessary for assembly, every other hole was skipped and the hole pattern became square. Pin Sensors In order to read the position of the pins, the bottom plate needed to be lined with sensors. The sensors needed to require a minimal amount of power and take up a minimal amount of space. Most sensors looke d at were either too big or not precise enough. The first sensor seriously considered was a hall-effect sensor. Hall-Effect Sensors Miniature Hall effect sensors, shown in Figure 3-19, were ordered and tested for use in the machine. The idea was that, during virtual visualization, actual pins would be arranged in a specific manner above the bottom plate of sensors. The plate would be raised and the sensors would detect when each pin comes in contact with the plate. Knowing the contact time and the location of the table at all times will result in the knowledge of all the pin placements. Figure 3-19. Hall-effect sensors

PAGE 48

36 Extensive testing was performed in trying to find the best manner to use the sensors to detect the pins. A successful setup was found where miniature magnets would be placed on the bottoms of the pins, and their presence when within a desired distance would change the magnetical field enough to alter the output of the sensor. Unfortunately, while tests showed that the miniature hall-effect sensors were capable of sensing when individual pins came into contact with the moving plate, several other difficulties were discovered. The detection signals, which the sensors output, were so small that a complicated interpretation system would have been necessary for the computer to notice them. More importantly, the process of actually attaching 900 of the sensors appeared to be both difficult and time consuming, if not impossible. As a result, this method was abandoned for a more simple design. Contact Traces and Steel Wool The new design maintained the concept of detecting when the pins made contact with the moving board; however, the difference involved tracing a sensor circuit in copper onto the moving plate. Below each pin would be two copper pads (each connects to a separate line on the computer). When the metal pin contacts the two pads, the lines get shorted and the computer detects a pin contact. It was found that the sensors were most efficient when the bottoms of the pins were lined with steel wool. The steel wool provides a wide, cushioned contact bumper for the sensor, which allows for imprecise contacts and rough impacts. The steel wool was easily attached to the bottom of all the pins with heat-shrink tubing as shown in Figure 3-20.

PAGE 49

37 Figure 3-20. Steel wool attached to the bottom of a pin Trace Pattern and Board Creation Once there were definite designs for the clamp and sensor systems, it was possible to design and create the plates for the physical display. The five plates from the machine were designed in Protel 99SE and created through an out-of-house manufacturer. Each plate was made from a yellow 9” 9” 0.062” sheet of G10/FR-4 glass epoxy. While a greater thickness would have been preferable for the boards, anything thicker would have cost significantly more than the project budget would have allowed for. Figure 3-21 shows the Protel drawing for the mentioned bottom sensor. As described, this board will be used to raise and lower pins as well as detect their positions. Two 50-pin ribbon cables will attach to the board delivering thirty output currents to the board. Two other 50-pin ribbon cables will attach to the board, probing thirty input lines for pin contacts.

PAGE 50

38 Figure 3-21. Digital bottom plate Figure 3-22 shows the Protel design for the ground plate that is just above the sensor plate. This board allows the pins to pass through it freely. While the bottom of the SMA wires will be attached to it, the ground board will remain stable and supply grounds to the thirty rows of SMA wires. The grounds will come from the two 50-pin cable ribbon connections.

PAGE 51

39 Figure 3-22. Digital ground plate The power board, seen in Figure 3-23, supplies current to the attached wires which, in turn, pass the current on to the SMA wires. The power is once against supplied from two 50-pin ribbon cable connections to thirty rows of wires. The board is attached to eight push solenoids and can be moved up and down. Moving the plate up will stretch out the SMA wire and release the clamp’s grip on the pin. Moving the plate down will help maintain the clamp’s grip on the pins when there might be additional outside forces upon them.

PAGE 52

40 Figure 3-23. Digital power plate The support plates, shown in Figure 3-24 will be attached to the ground and power plates. These plates will be used to help keep the wires in place by pinching them in between the boards and the support boards will a dd strength to prevent deformation of the ground and power plates.

PAGE 53

41 Figure 3-24. Digital support plate The top board, displayed in Figure 3-25, will be what the user sees from above. It will be held stable by the surrounding skeletal structure and the clamps will be attached to the board. The pinheads will be wider than the holes in the top board and will not be able to move below it.

PAGE 54

42 Figure 3-25. Digital top plate Pin Movement System and Frame The pin movement system and frame had fewer restrictions on their creation and therefore, were basically designed by working around the other two systems. First, it had to be decided which type stepper motor and solenoid would be used. This was necessary because these pieces needed to be ordered from a manufacturer and the connecting pieces would need to be designed to fit around them. Motor Selection Because the plate moving the pins has to make strict, easily controlled movements, it was decided that three uniformly placed stepper motors would be best for shifting the plate up and down (Histand et al. 1999). The most economical motor that met all the desired specifications was the Thompson 55M048D1B Bipolar Stepper Motor, as shown in Figure 3-26.

PAGE 55

43 Figure 3-26. Thompson 55M048D1B Bipolar Stepper Motor It was not important what type of screws would be used to move the sensor board up and down; therefore, it was decided to use some 3/8” round screws that were conveniently already in-house. To decide on the stepper motor that would be used, it was necessary to find the torque required to turn the screws that so that they would smoothly move the sensor board with the pin resting upon it vertically. The necessary torque to be used by each screw, Tu, could be calculated by using Equation 1 and the free body diagram shown in Figure 3-27, where F is the downward force on the screw, f is the coefficient of friction for the metal screw against the plastic nut, L is the lead and Dp is the pitch diameter (Mott 1999). Figure 3-27. Free body diagram of forces on the screw

PAGE 56

44 L f D D f L D F Tp p p u 2 (1) The downward force, F, could be found using Equation 2 where mpin is the mass a single pin, n is the number of pins, and Wplate is the weight of the sensor board and the bracket it is attached to. plate pinW g m n F (2) The mass of an individual pin was measured at 2.92g on an electric scale, the weight of the sensor board and bracket was approximated at 5lb, and at the time of the calculation an n of 900-pins was used. An F of about 11lbs was calculated. Substituting that into Equation 1 along with a Dp, of 0.5”, an L of 0.25”, and a worst-case f of 0.5, Tu was calculated to be 1.68 lb-in. To find the desirable amount of torque that each stepper motor would provide, Equation 3 was used with a safety factor, SF, of 3, and a stepper motor total, nSM of 3. SM u desiredn T SF T (3) After converting the units the desired torque was found to be 31.5 oz-in per stepper motor. The Thompson 55M048D1B Bipolar Stepper Motor was chosen for use because it could provide 33.5 oz-in and was an economical purchase. Solenoid Selection In order to re-stretch the SMA after having been heated and condensed, the clamp sleeve (which is attached to the SMA) will be moved by one of the spinal plates. It was decided that the plate will be moved by a three Magnet-Schultz DC voltage cylindrical push solenoids, as displayed in Figure 3-28. Moving the plate will move the sleeve and in turn stretch the SMA.

PAGE 57

45 Figure 3-28. Magnet-Schultz DC Voltage Cylindrical Push Solenoid Because the motion required to move the ground board up and down was a discrete action, it was decided that a solenoid would perform this task best. In order to decide upon an exact solenoid it was necessary to decide how much force (Fmax) each solenoid was going to be required to apply. The majority of the force used from the solenoids would be used to push the freedom ring over the clamp, everything else could be considered negligible. The maximum force (F1) required to stretch out a single SMA was calculated using Equation 4, where mMAX is the maximum amount of mass required to pull the freedom ring over the clamp and g is the acceleration of mass due to earth's gravity. g m FMAX 1 (4) Tests with an electronic scale showed mMAX to be approximately 10 grams. Fmax was then calculated using Equation 5 where n is the number of freedom rings being pulled down and N is the number of solenoids being employed. N F n F1 max (5) It was decided that the operation would be most balanced if a solenoid were employed at each corner of the board, thus making N equal 4. At the time it was believed that there would be 900 pins, thus making n equal 900. Calculated out Fmax was found to be 353 oz. The Magnet-Schultz DC voltage cylindrical push solenoid was capable of supplying 515 oz of force and was most economical for its capabilities and therefore

PAGE 58

46 became the solenoid that was implemented. When n decreased as the design evolved, the force calculated was still sufficient because the additional force could only be beneficial. Shaft Selection The shaft system that moves the bottom sensor plate up and down was designed and drawn up in Mechanical Desktop, as seen in Figure 3-29. The stepper motor spins a 3/8” screw to move a nut up and down, and the bottom plate is attached to the nut. The stepper motor will be attached to the screw with a flexible coupling. The ends of the screw will be supported by single-row angular contact ball bearings which will protect the screw from axial and radials loads. Figure 3-29. Shaft / motor system The design of the shaft system began with an assessment of what was available, what was necessary, and what was desired. The following facts were known: Three 3/8” in diameter screws would serve as the shaft

PAGE 59

47 The 3 screws would need to be connected to three stepper motors having shafts of 0.25” diameter and 0.375” length 4 to 6 inches of plate movement were desired from the shaft Bearings would be necessary to hold the screws in place and prevent damage to the steppers (to isolate) Flexible couplings were bought to connect the screw shaft to the stepper motor shaft. It was decided that the couplings would be flexible so as to prevent damage for the likely situation that the two shafts would not be perfectly aligned. It was decided that the ends of the screws would be worked down to a quarter inch in diameter so as to match the stepper motor shaft and make it easier to find a functional coupling. Because bearings could not be found that would exactly fit the ends of the screws, bushings were created. In order to design the bushings, decide the exact screw length, and make certain that everything would fit together, all the pieces for the system were drawn twodimensionally in AutoCAD, as shown in Figure 3-30, and aligned properly.

PAGE 60

48 Figure 3-30. Two-dimensional drawing of the shaft/ motor system alignment With the shaft system and the plates designed, it was possible to create the surrounding structure, which is shown in Fi gure 3-31. The structure was designed in Mechanical Desktop and is approximately 24” 20” 27” in size. The openness of the design allows for easy viewing and accessibility to the machine’s many parts within the structure.

PAGE 61

49 Figure 3-31. Skeletal structure of 80/20 The frame was designed with the purpose of being capable of holding together all of the display’s working parts, bearing the we ight and forces from the display’s actions, and still being open for easy viewing and maintenance. It was made from extruded aluminum and connected using groove-fitted screws and nuts. Figure 3-32 shows the assembled frame with the cooling fan attached.

PAGE 62

50 Figure 3-32. Assembled 3-D display frame with fan Next, additional mounts had to be designed to connect all the pieces. The “motor mount,” seen in Figure 3-33, serves as an important hub between the shaft system and the machine structure. Made from delrin, this mount braces one of the single-row angular contact ball bearings attached to the shaft while also holding the stepper motor on the bottom. The flexible coupling is sheltered in the center of the mount, but still easily accessible for tightening through two side holes.

PAGE 63

51 Figure 3-33. Motor mount The “upper bearing mount” braces the contact ball bearing on the top of the shaft and connects it to the 80/20. Shown in Figure 3-34 and made from delrin, this bearing is aligned in the z-axis with the motor mount to complete the shaft system. The three shaft systems are used to move the bottom plate up and down.

PAGE 64

52 Figure 3-34. Upper bearing mount The “solenoid mount,” seen in Figure 3-35, has one push solenoid screwed into its top socket and another screwed into its bottom socket. The moving pins in the solenoids are pointing towards one another with the power plate in between them. There is a solenoid system attached to each corner of the power plate. The bottom solenoids are able to push the plate up with the purpose of stretching the SMA wires and releasing the clamps’ grips on the pins. The top solenoids push down on the power board in order to maintain the clamps’ grips on the pins. When neither solenoid is active, the plate is held in its desired place by springs resting around the pushing pins inside the solenoids.

PAGE 65

53 Figure 3-35. Push solenoid mount Figure 3-36 shows a preliminary drawing of the machine assembled without the pins, clamping systems, or plate attachments. The drawing shows the relative placements of the three shaft systems, four push solenoi d systems, and plates with respect to the aluminum structure.

PAGE 66

54 Figure 3-36. Preliminary assembly drawing Brackets were designed to attach the various plates to the to the machine frame. Each design took into consideration the fit of the parts around it, the actions it would be making, and the forces that it would have to withstand. It was decided that making the brackets from steel would be structurally s ound while still being economically efficient. Figure 3-37 shows a detailed dimensioned drawi ng of the bracket for the top plate. Two of the L-shaped brackets attached to opposing ends of the top plate. The top plate was screwed to one section of the “L” and the top plank of 80/20 extruded aluminum was screwed to the other section. These brackets hold the top board rigid while allowing for free movement of the solenoids (which move the ground plate) around the brackets.

PAGE 67

55 Figure 3-37. Top plate bracket Figure 3-38 shows the detailed dimensioned drawing of the brackets for the ground plate. These brackets were screwed to the opposing ends of the ground board for a rigid attachment. The other ends of the brackets were attached to four push solenoids each. Two solenoids were inserted into opposing ends of each hole in the brackets. Pushing from the solenoids on the brackets leads to a corresponding movement of the ground plate.

PAGE 68

56 Figure 3-38. Ground plate bracket Figure 3-39 shows the detailed dimensioned drawing of the brackets for the power plate. Opposing ends of the power plate were screwed to the inside on the longest section of these two C-shaped brackets. The two sm aller symmetric sections of the brackets are two planks of 80/20 on the level of the frame that is third from the top. These brackets hold the power plate rigid even when the attached SMA wires are pulling on it.

PAGE 69

57 Figure 3-39. Power plate bracket Figure 3-40 shows the detailed dimensioned drawing of the bracket for the sensor plate. The sensor plate is screwed to this singular piece for a rigid attachment. The three arms of the bracket are screwed to the nuts on the three motor / shaft assemblies. Turning the shafts in unison result in a symmetric movement of the nuts. This, in turn, moves the bracket and sensor plate up and down (lowering the sensor plate allows for movement of the pins).

PAGE 70

58 Figure 3-40. Sensor plate bracket All the parts that were designed specifically for the display device were manufactured. The following parts were all created in the UF machine shop: Two Top Plate Brackets Two Ground Plate Brackets Two Power Plate Brackets Three Upper Bearing Mounts Three Motor Mounts Four Solenoid Mounts Six Bushings Six Screw Shafts To make the top plate bracket, a precise milling machine was required; because this was not available in the UF shop, the piece was created by an out-of-house manufacturer. The pieces were press-fit together as with their associated parts. Figure 3-41 shows a picture of the finali zed physical display, having been fully assembled.

PAGE 71

59 Figure 3-41. The fully assembled physical display

PAGE 72

60 CHAPTER 4 CONTROL BOX Where the computer serves as the brains of the machine, the control box works as the circulatory system. Control signals are delivered from the computer to the box where they are processed, re-routed and the appropriate signals are sent to the display device. This system contains numerous terminal boards, power sources, controllers and thousands of wires that are vital to the transportation of electrical signal through the machine. This chapter will detail the processes used by the control box, how it is organized, and parts that make it up. Raster Scanning As a result of having hundreds of pins, the 3-D Visualization Device has hundreds of actuators and hundreds of sensors that need power. It was an important task that the amount of power being distributed be as mini mized as possible. The solution that was implemented was a process known as raster scanning. This process only works for grids of electrical devices where the number of columns is equal to the number rows, which was conveniently the case at hand. The benefit of raster scanning is that it decreases to the total number power lines needed to N2, where N is the total number of electrical devices. The negative is that only a single column of devices can be activated at any single moment. The simple solution to this problem is that each column is activated in a pattern of quick sequential bursts. The result is visually the same as activating everything at once. The setup for a raster scan is basically that each column of devices is receiving a common ground line and each row devices is receiving a common power line. This can

PAGE 73

61 be reversed, as the important part is that the columns are all receiving one type of power and the rows are receiving another type of pow er. Figure 4-1 diagrams an example setup for a three by three setup of actuators. Figure 4-1. An example 3x3 grid of actuators in a raster scanning sequence The diodes in the setup keep current from flowing in an undesired direction. For a of grid actuators, the columns would output power until it is desired to begin actuation. At that point, the first column will output a ground, and the rows will send power only to the rows that are to be actuated. Next, the first column outputs a power again. Then, the second column quickly outputs a ground and the row outputs change to give power to the actuators being activated in that column. This pattern is repeated for the remaining columns and continues for the remaining time of activation. To the human eye it appears that the actuators are all being activated once. For the sensors, such as in Figure 4-2, the process is a similar process with polling. The columns are only powered one at a time, while all the rows are read. The rows use a pull down resistor, trying to draw a current. If a current is read, then the sensor at that

PAGE 74

62 row and column has been contacted. The polling of columns can be done so quickly that the off time for each column is negligible. Figure 4-2. An example 33 grid of sensors in a raster scanning sequence Mapping The control box can initially be looked at as a black box, as show in Figure 4-3. Six 50-pin cable ribbons come out of the computer and connect to their own individual terminal boards. Each ribbon contains 24 actual signals, which are each broken down into an individual wire through the terminal board. Appendix D details the mapping of where each cable ribbon pin goes and what it does. After being processed, the various signals are processed and enter eight more terminal boards, which send information to the display device through eight more ribbon cables.

PAGE 75

63 Figure 4-3. Direction of signals through the control box Figure 4-4 shows a basic mapping of how the control box is connected within and to the surrounding components. Sixty-three of the signals are used to control sixty-three opto-isolators. These opto-isolators act as switches controlling the flow of current too high for the computer to handle. Fifteen optoisolators control the flow of power to the power board. Fifteen opto-isolators contro l the deliveries of ground flows from the ground board. Fifteen opto-isolators control the flow of current to the sensor board and fifteen opto-isolators control the reception of current from the sensor board back to the computer. One opto-isolator is used to contro l the flow of power to the four solenoids, which push the ground plate down, and another opto is used to control the flow of power to the four solenoids pushing the plate up. Another opto receives signals from the sensor on the display device that reads when the sensor plate is completely down. A stepper motor driver inside the box is used to control the timing, speed, and direction of the three stepper motors in the display device. Two power supplies are used to power all the components of the display device. A MAP130-1024 Power-One DC power supply sends 24 Volts to the eight solenoids on the display, and another 24 Volts that are used to power the stepper motor driver. An L&C Inc. switch power supply sends 12 Volts to the columns of the sensor board, and 5.25 Volts to the power board.

PAGE 76

64 Figure 4-4. Basic mapping of the control box

PAGE 77

65 CHAPTER 5 COMPUTER The computer serves as both the brain of the machine and a way for the user to communicate with the display devices. Its three main components are the Graphical User Interface (GUI), the path planner, and the Digital I/O. This chapter details how those components work and what they do. Overview The computer was setup for controlling the machine, as seen in Figure 5-1. A Linux operating system was installed and Motif and Open Inventor were installed for running the graphical user interface (Brain 1992, Wernecke 1995). Hardware programs were installed to run the digital I/Os and motor controllers. Figure 5-1. Computer setup

PAGE 78

66 Graphical User Interface In order for the user to be able to control the machine from a computer, a Graphical User Interface (GUI) was created using Motif and Open Inventor. This interface is used to both tell the physical display how to position itself and present a rendered representation of the current physical display. Figure 5-2a shows the GUI that was created to control the three-dimensional visualization machine. The user can read the current positions of the pins in the drawing by entering the row and column numbers of the desired pin and clicking on a “Get Height” button. It is important to note that the rows and columns are labeled “1” through “15”. Entering a row, column and height and pressing the “Set Height” button will change th e height of a pin. As shown in Figure 52b, an entire positioning set for the pins can be read from or saved to a file using the “Read” and “Write” buttons. To send the pin positions to the physical display, the “Send” button must be clicked. To display the current pin positions from the physical display, the “Receive” button must be pressed. Also, the display view can be adjusted by dragging the rotating wheels and using the icons alongside the display. (a) (b) Figure 5-2. Graphical user interface (a) Ex ample GUI display, (b) GUI reading a file

PAGE 79

67 The GUI also, has a series of buttons on the right side that can be used for testing the various sections of the physical display. The “Bottom Solenoids” button is used to make the solenoids push the power plate up and the “Top Solenoids” button has them push the plate down. The “Move Motor” button can be used to move the sensor plate up and down. The user places a desired height for the pins in the corresponding text box, then left-clicks the button and the pins are moved to that height. The pins and plate will remain at that position until another command is made. The “Activate Pin” button can be used to trigger a specific pin actuator. The user writes the row and then the column, respectively, of the pin to be actuated in the corresponding text boxes. After the button is clicked, the SMA will be heated and the pin will be clamped. The “Read Pin” button can be used to test the sensor board. The user enters the row and column of the pin to be tested in the text boxes used for the “Activate Pin” button and clicks the “Read Pin” button. Next, the test field next to the “Read Pin” button will say whether or not the pin at the specified position is engaged with the sensor. Path Planner In order for the user’s input to reach the pin display device, a “path planner” program was created in C++ that interprets the inputs from the user interface and executes them through the digital I/O. This program executes all major actions such as plate movement, pin sensing, pin clamping, a nd SMA stretching. It was decided that the program could use an open loop system to move the pins up and down because sensors will allow the program to rediscover the bottom plate’s location at the end of every display process and any location errors that could occur in between would be so negligible that the human eye would not be able to notice them.

PAGE 80

68 The path planner program utilizes the following algorithm for physical visualizations (The terms from the code can be interpreted by using Appendix E): 1. 'Send' button sendPhysical(), sendPhysical2() (MakeItReal.c++) -------sendCB() (main.c++) 2. If machine is not busy, lock machine and call sendPhysical(); else, stop. --------sendPhysical() (MakeItReal.c++) 3. Load all pins into a priority queue, sort by height (tallest first). 4. Set sensor board speed to fastest setting. 5. Move board to top. -------popup_routine() (MakeItReal.c++) 6. Popup dialog asking if user wants to unlock pins by activating solenoids. 7. Continue prompting until user answers, "No, continue." -------sendPhysical2() (MakeItReal.c++) 8. Set sensor board speed to slowest setting. 9. Pop front pin off of priority queue and descend to its height. 10. If that height is greater than zero, fire the pin. 11. Reset sensor board speed to fastest. 12. Unlock machine. In turn, for virtual visualization, the program uses the following algorithm: 1. 'Receive' button receivePhysical(), receivePhysical2() (CompDisplay.c++) -------receiveCB (main.c++) 2. If machine is not busy, lock machine and call receivePhysical(); else, stop. -------receivePhysical (CompDisplay.c++) 3. Set sensor board speed to fastest. 4. Move board to top. 5. Activate and hold solenoids to lock clamps. 6. Move board to bottom.

PAGE 81

69 7. Popup dialog allowing user to adjust pins. receivePhysical2() when clicked. -------receivePhysical2() (CompDisplay.c++) 8. Set sensor board speed to slowest. 9. Move board up and at each click check all pins to see if engaged. 10. Record the lowest height at which a pin was engaged (in 'pinsArea'). 11. Unlock the solenoids. 12. Reset sensor board speed to fastest. 13. Unlock machine. Digital I/O Two PCI-7296 digital I/O cards, seen in Figure 5-3, were installed into the computer. Each card has 96 entries capable of sending or receiving data signals. They are used for the following functions: 30 outputs for clamp control 15 outputs for running sensors 15 inputs for reading sensors 3 outputs for running motor controllers 2 outputs for controlling solenoids 1 input for sensing the bottom placement of the bottom plate Figure 5-3. PCI 7296 digital I/O card

PAGE 82

70 CHAPTER 6 ASSEMBLY The assembly process of this project was complicated, difficult, and greatly timeconsuming. While a majority of the design and testing occurred before the assembly process began, the design and testing processes became mixed into the assembly process as well. The physical display had to be restructured several times in order to be accessible for hand assembly as well as to solve unexpected problems. Multiple design questions could only be answered by performing tests as the device was being assembled. As this was a prototype machine, many decisions had to be made on the fly and the process took up a pattern of assemble, test, re-des ign, and repeat. This was the step of the project where a don’t-look-back approach had to be taken at times. If a problem occurred, it was corrected to the best degree possible without having to start over again. Now that the machine is completed, it would be recommended to perform some assembly processes differently with some redesign changes. Those recommendations are presented in Chapter 8. This chapter details exactly how this machine was assembled and how to recreate it. Control Box The control box was basically assembled by attaching the electrical controls to four perforated shelves. The shelves have rails on them, which were inserted into rails inside a large metal box. The seven opto-isolator boards, the two power sources, and the stepper motor driver were screwed into the perforations in the shelves. The terminal boards had no holes for such attachments, so they were connected to the shelves with

PAGE 83

71 snap ties and Velcro was added for additional support. The ports were then wired together following the previously mentioned mapping of Appendix D. Multiple colors of wire were used in order to distinguish separate sections of the mapping. Pre-Clamp Physical Display Because the pin clamp systems were delicate and permanent, it was necessary to have everything else on the physical display assembled and working correctly first. Frame The first part of the physical display to be assembled was the frame. It is composed from the following parts: Four 20” long extruded aluminum strips Four 16” long extruded aluminum strips Four 11.5” long extruded aluminum strips One 11” long extruded aluminum strips Four 10” long extruded aluminum strips Four 9” long extruded aluminum strips Four “T”-brackets Fifty-Six “L”-brackets Because the frame could not be assembled all at once, it had to be made in sections. Figure 6-1 shows what the assembled frame sections would look like without any other attachments.

PAGE 84

72 Figure 6-1. The sections of the frame where the red circle encompasses a waste section, the blue circle encompasses an arm section, and the green circle encompasses a leg section All the strips of aluminum were joined by placing T-nuts inside the aluminum extrusions and screwing the hinges to them. What follows are the three types of frame sections that were created: Two waists were made by joining the aluminum strips into the pattern shown in Figure 6-2. Figure 6-2. Waist section of the frame Two legs are used to support the entire display. Joining the aluminum strips in the pattern displayed by Figure 6-3 can create a leg.

PAGE 85

73 Figure 6-3. Legs section of the frame Two arms are used to hold the top two plat es. Figure 6-4 shows the aluminum strip assembly pattern Figure 6-4. Arms section of the frame The remainder of the assembly was matter of carrying out the following specific procedure:

PAGE 86

74 Motor/Shaft Systems 1. Press fit a bearing into each of the three motor mounts. 2. Press fit a bearing into each of the three upper bearing mounts. 3. Press a single bushing onto each of the three shafts so that the widest end of the bushing is flush against the widest end of the shaft. 4. Tighten the screw on the bushing so as to tighten their grip on the shafts. 5. Insert a flexible coupling into the center of each of the three motor mounts so that the couplings center hole is facing the top and bottom of the mount. 6. Insert a stepper motor shaft into the bottom of each motor mount and into center hole of the flexible coupling within. 7. Screw the three stepper motors on. 8. Insert the end of a screw shaft that hangs out of the bushing into the top of the motor mount and into the top end of the flexible coupling until the bushing is flush against the bearing. 9. Screw the nuts onto the screw-shaft so that the threads on the nut are at the top 10. Screw the three nuts onto the sensor board bracket and move the bracket to the lowest possible location. 11. Screw the three motor mounts onto one of the waste sections of the frame so that the corners of the bracket are symmetric with the section. 12. Screw the three upper bearing mounts to the other waste section of the frame (make certain to leave the screw loose enough that the mounts can slide on the frame section. 13. Place a bushing on the top of each of the screws so that the widest end of the bushing is flush against the widest end of the screw. 14. Tighten the screws on the bearings. 15. Place the second waste section of the frame on top of the other by inserting the ends of the screws coming out of the bushings into the bearings in the upper bearing mounts until the bushings are flush against the bearings. 16. Adjust the placement of the upper bearing mounts so that the screw-shafts are perfectly upright and the two waste sections of the frame are symmetric with one another. 17. Tighten the screws on the upper bearing mounts.

PAGE 87

75 18. Screw the leg sections of the frame onto the waste sections (the exact placement that they are attached isnt as important as making certain that the waste sections of the frame are level. Brackets 1. Screw a top plate bracket onto each arm section of the frame. 2. Screw the power plate brackets onto the top waste section of the frame. 3. Screw the ground plate brackets onto the ground plate. Plates 1. Screw the sensor plate onto the sensor plate bracket. 2. Screw the power plate onto the ground plate brackets. 3. Screw the top plate onto the top plate brackets. 4. Screw the arms loosely onto the top waste section of the frame. 5. Align the top plate with the power plate and tighten the arm attachments. Solenoids 1. Screw two solenoids into each of the four push solenoid mounts. 2. Screw the solenoid mounts loosely onto the four corners of the arms. 3. Place a solenoid shaft through each of the four top solenoids and tighten a collet around each shaft so that about 1/8 of the shaft hangs out of the collet. 4. Place a solenoid shaft through each of the four bottom solenoids, place a spring around each shaft, and tighten a collet around each shaft so that about 1/8 of the shaft hangs out of the collet. 5. Slide the ground plate into its proper position and insert the eight solenoid shafts into the four ground plate bracket holes so that each hole has a shaft entering it from each end. 6. Re-adjust the collets on the solenoid shafts so that they are flush against the ground plate bracket. Pin-Clamp Systems The pin-clamp systems were the most tedious and time-consuming work of the entire project. Most sections of the system had to be shaped, created, or assembled by hand. The time involved in creating the parts was a large reason that the number of pins was lowered from 900 to 225. The difficulty involved in the hand assembly was why the

PAGE 88

76 spacing of the pins was increased. Each section was created in mass before creating the next section and each unit assembly followed a carefully structured course of action. It is also important to note that the number of units created for each section outnumbered the number of units needed. This was important because many units were broken in during assembly and testing, and it saved time to have the additional units made. The following assembly procedures detail how to make each section of the system and how to install them into the machine: Mechanical Pencil Disassembling Each of the moving pins on the physical display device required a working clamp with a freedom ring to lock its position. Each clamp and freedom ring needed to be removed from a single Pentel Crayz 1.3 mm mechanical pencil. When it was still planned to have 900 pins, 996 pencils were ordered and stripped down to the clamp, spring and freedom ring. Because of the large number pencils, the stripping process was broken down into five specific steps. 1. Disassembly: All parts of the mechanical pencil that were capable of being removed by hand were detached and collected. The detached parts included the eraser cap, eraser, two sticks of lead, the pencil cap, and the cone cover. 2. Shell Removal: The clamp was attached to a long cylindrical, semi-rigid plastic sleeve. A much harder plastic shell surrounds all this, which is what the user grips when writing with the pencil. The hard shell could only be removed by being broken off. The pencil was placed in a vice and when the vice increased the pressure, the shell would fracture linearly and could be peeled off. 3. Sleeve Slicing: The semi-rigid plastic sleeve holds the clamp in place and an increase in the sleeve’s diameter holds the freedom ring and spring over the clamp. A dremel with circular blade was use to slice the sleeve through its smaller diameter where it starts to cover the clamp. This removed all of sleeve except for a thin ring, and, it allowed for the following step. 4. Spring Removal: Without the sleeve to hold it in place, the springs could be removed from around the clamp and placed and bin where they would be used later.

PAGE 89

77 5. Sleeve Dissection: In order to remove the remaining ring from the plastic sleeve around the clamp, the dremel with the circular blade was used to carefully slice through the ring on two sides. The ring was, in turn, dissected into two pieces that would just fall off. Because the clamp was made of a harder plastic than the sleeve, the dremel could be set at a low enough spinning rate that it would cut easily through the sleeve and cause a negligible amount of damage to the clamp. Once the sleeve was removed the freedom ring could be pulled off and the clamp and ring could be stored in their associated bins, as seen in Figure 6-5. Figure 6-5. Bins of separated materials Brass Cup and Wire Wrap Each brass cup needed to be attached to a wire-wrap wire. This process needed to be performed in a precise manner and units that did not meet the desired specifications had to be scrapped of everything except the brass ring. Because the wire-wrap can be delicate when bent, it was necessary to make an additional 75 functional units. What follows is the procedure used to create them. Step 1: Cut 300 0.125” lengths of 3/ 16” in diameter heat-shrink tubing.

PAGE 90

78 Step 2: Place the bottom of a drill bit in vice so that the bit stands vertical. Step 3: Place brass freedom cup around the top of the drill bit so that the rim of the cup is closest to the ground. Step 4: Hold a 3” long length of wire-wra p wire along side the freedom cup so that the wire is parallel to the bit and its center is against the cup. Step 5: Slide a length of heat-shrink tubing around the wire-wrap and freedom cup so that bottom of the tubing pushes against the rim of the cup. Step 6: Heat the tubing so that it shrinks completely around the cup and wire. Step 7: Let cool. Step 8: Repeat steps 3-7 another 299 times. Diode Curling The diodes, shown in Figure 6-6, were necessary to prevent electricity from flowing in the wrong direction. The outgoing wire needed to have a loop that the SMA wire could connect to and the incoming wire needed to have an arm that could be attached to the trace on the power board. Following is how each diode was structured: Step 1: Curl the outgoing end of a diode into a spiral. Step 2: Cut off the excess wire after the second curl. Step 3: Curl the ingoing end of the diode into an L-shape. Step 4: Cut off the excess wire so that both brackets of the L-shape are approximately 0.25”. Step 5: Repeat steps 1-4 another 249 times. Figure 6-6. A diode before assembly and a diode after assembly

PAGE 91

79 Guide Tubes The black guide tubes, seen in Figure 6-7, hold the pin clamps in a solid position and provide directional guidance for the pins. Following is the process for making them: Step 1: Cut out 275 1.25” long strips of bl ack 3/16” in diameter heat-shrink tubing. Step 2: Cut out 275 0.25” long strips of clear 3/16” in diameter plastic tubing. Step 3: Clamp one end of a steel rod into a horizontal position so that it is untouched everywhere else. Step 4: Place a strip of plastic tubing around the rod. Step 5: Place a strip of heat-shrink tubi ng around the plastic tubing so that one end overhangs the plastic tubing by about 0.25”. Step 6: Shrink the tubing with a hot air. Step 7: Let cool. Step 8: Repeat steps 4-7 another 274 times. Figure 6-7. On the top-left is the heat-shrink tubing, towards the center is the clear tubing, and on the right is the assembled guide tube Steel Wool Attachment Steel wool was attached to the bottom of each pin for the purpose of contact with the sensors. The amount of steel wool used for each pin was arbitrary and not very important as long as it was not so much that it would contact the next sensor over and not so little that it could not flew when it rested on the sensor board.

PAGE 92

80 Step 1: Cut out 250 0.5-inch lengths of 3/32” in diameter heat-shrink tubing. Step 2: Rip out a dime-sized amount of steel wool. Step 3: Wrap the steel wool around the end of a music wire. Roll the steel wool between fingers and squish the length down until the steel is about a half inch in length and has a football shape. Step 4: Place heat-shrink tubing around opposite end of wire and move down until most of steel wool is cover with a small ball of wool still hanging out (make certain the wire still goes to the end of the heat-shrink tubing. Step 5: Heat the tubing until it shrinks completely. Step 6: Let cool. Step 7: Repeat steps 2-6 249 more times. SMA System Assembly The connection of SMA wire to the diode and the wire-wrap wire needed to be done carefully for two reasons. Bending or over-stretching the SMA wire could render it useless without any visible problems until after being installed into the display device. Accidental bending of the wire-wrap wire coul d make it weak and lead to wire breaking during the later assembly. The pieces were attached to one another in the following manner: Step 1: Bend an SMA wire approximately 1” from its end. Step 2: Loop the bend and place the closest end through the loop, so as to create a loop that is approximately 0.25” in diameter. Step 3: Place the same end through the same hole in order to maintain the stability of the knot. Step 4: Bend the SMA 3.5” from the loop. Step 5: Cut the SMA 4.5” from the loop. Step 6: Repeat Steps 2&3 on the opposite end of the wire.

PAGE 93

81 Step 7: Bend the wire-wrap wire on a brass cup assembly piece so that when the SMA is completely straight and the wire-wrap wire is straight, the rim of the cup will be 4” from the furthest loop on the SMA wire. Step 8: Place the wire-wrap wire through the closest loop on the SMA wire and wrap around the loop so that the bend in the wire-wrap wire is around the very top of the SMA wire. Step 9: Wrap the excess wire-wrap wire around the SMA loop and down about another half-inch and back up again. Step 10: Swivel the unused SMA loop around the curled end of a shaped-diode and clamp into place with pliers. Step 11: Repeat steps 1-9 another 249 times. Final Attachment Process The installation of the clamping systems was easily the most difficult part of the assembly process. It required poise, patience, and awkward use of the hands. It is virtually impossible to perform this assembly by hand to a degree of precision that is necessary for perfect results. The following steps were taken to perform the attachment: Step 1: Place the threaded end of a 4-40 socket head cap screw through a hole in the ground board to the front-left of the hole where the Pin 1,1 will be placed so that the threaded end comes out on top. Step 2: Screw a 4-40 small pattern machine screw nut onto the screw and tighten it so that the screw can not move (if the screw is not secure, it may loosen and fall off later due to shaking of the board). Step 3: Repeat steps 1-2 for the remaining 224 pin holes, respectively. Step 4: Place a 4-40 socket head cap screw through the hole in the power board (so that the threaded end comes out from the bottom of the board) that is directly below the screw for Pin 1,1 in the ground board. Step 5: Screw a 4-40 small pattern machine screw nut loosely onto the screw. Step 6: Select an SMA system assembly and pinch the bottom wire of the “L” shaped diode wire between the head of the screw and the copper tracing beneath (it may help to curl the bottom stretch of wire with a pair of pliers so that it fits more snugly around the screw).

PAGE 94

82 Step 7: Tighten the nut on the screw so that it cannot move. Step 8: Loosen the screw directly above. Step 9: Wrap the wire-wrap wire around the loosened screw (between the nut and the copper tracing on the ground board) so that the SMA system assembly is taught. Step 10: Tighten the screw so that it cannot move. Step 11: Place a guide tube through the hole in the top board where Pin 1,1 will be placed. Step 12: Trim the guide tube. Step 13: Place a pencil spring around the guide tube and a pencil clamp through the freedom ring on the SMA and into the guide tube so that it is tight. Step 14: Thread the thin end of an assembled pin through the bottom of the power board, into the clamp and out the top of the guide tube. Step 15: Test the system to make certain it works (“C” shaped tubing can be wrapped around the guide tube to tighten the pencil spring and hot glue can be used to tighten the wire-wrap wire. Elimination of the Fan Originally, the final step of the assembly called for the attachment of a fan to be attached in line with the pin-clamp systems as shown in Figure 6-8. However, when the machine reached that stage, there was worry that the system was too fragile to handle the fan. Also, because only one SMA was being heated at a time, there really wasn’t much chance of the heat in that area being any hotter then the rest of the room; thus, making the fan unnecessary anyhow and therefore, the fan was never attached. Once the machine was run it was obvious that the fan was not necessary.

PAGE 95

83 Figure 6-8. Physical display with fan

PAGE 96

84 CHAPTER 7 TESTING AND RESULTS As discussed in the previous chapter, testing was a vital process in the progression of the design of the 3D virtualization device. Each concept needed to be implemented into a smaller scale prototype before it was assimilated into the final design. This process was used from the time the first design concepts were thought up until the final tinkering of the finished product. This chapter further details the tests used to prove or disprove concepts, their results, and the adjustments that they led to. Tests Clamp System Tests First SMA test When an initial clamp design was complete and a basic setup existed, tests were run in order to evolve the clamp system design. A one-pin model, shown in Figure 7-1, was created to test the clamping process that will be vital to the visualization machine. The model tested the method of using clamps from a mechanical pencil to hold a pin in place. Pulling a cylindrical collar over the clamp closes it and locks the pin in place. The collar is pulled down though use of SMA wires. These wires contract when heated with an electrical current. Several methods of attaching the wire to the collar had to be tested. The most successful tests came by drilling holes through a thin rim on the collar, threading the wire through the holes and attaching the two ends to the bottom of the model frame. The ends of the wire were where the power attachments were made. The frame was designed on a CAD program and then created in a rapid prototyping machine.

PAGE 97

85 The clamp was attached to the frame by being press-fit, and hot glue was used to hold it in position. Figure 7-1. Clamp actuation testing model This test officially proved that the SMA wire could be used to pull the freedom cup over the clamp and provide the necessary force to hold the pin in place. It also showed that the length of wire between the cylinder and electrical connection could be minimized to 3.5”. In the initial system, the wire was run to the cylinder and back again because there was no convenient way to place an electrical connection at the cylinder. It was found that the system remained functional when one of the sections of wire run from the cylinder to the electrical connection was replaced with a normal wire. This cut the amount of SMA wire needed in half. Micro helix tests Two Models were made to test the BioMetal Micro Helix in clamp actuation. Figure 7-2a shows the first model. The idea here was that the cup would be held stable in the plate and the micro helix would pull the clamp down into the cup, thus preventing

PAGE 98

86 movement of the pin. The second model shown in Figure 7-2b turned the clamp upsidedown and held it stable within the plate. This time the micro helix pulled the cup down to tighten the clamp and prevent movement of the pin. While both actuators accomplished the desired task, the second model proved to be more reliable. Leaving the clamp loose, as it was in the first model, made the pin shakier. Also, when the pin was moved, the clamp maintained a better grip when it was positioned as it was in the second model. (a) (b) Figure 7-2. Micro helix tests (a) Helix model 1, (b) Helix model 2 Unfortunately, the micro helix design proved to be too costly for this project and focus had to be returned to the SMA wire design that was originally in place.

PAGE 99

87 Electrical Tests Raster Scan Test A test was created in order to prove that the raster scan test would be capable of controlling the clamp and to make certain that some of the parts being used would fit the system properly. Figure 7-3 shows the setup for the test. Basically, a 33 grid of LEDs, each in series with a resistor, was used to represent the SMA’s. The LEDs were wired to a power source in the pre-described raster scan method. The power source was wired to the grid through a board of opto-isolators that would actually be used in the finally machine. The opto-isolators were controlled by PICmicro 16F627 micro-controller. A keypad that was wired to the micro-controller could be used to make a specific pattern of LEDs light up. The micro-controller would determine what button was pushed, the corresponding pattern that went with it, and would control the flow of current so as to create the desired pattern. Figure 7-3. The raster scan test setup

PAGE 100

88 The test provided to be a helpful lesson in using the parts system and setting it up. It proved that the system would work successf ully. It also showed that many of the available opto-isolators either did not work in th is type of system or just did not work at all. Sensor Tests Hall Effect With initial plans of detecting pin positions with hall-effect sensors, an effort was made to find the best way to use the sensors. Two methods of pin sensing, based on the fact that hall-effect sensors detect a change in the surrounding magnetic field, were devised and tested, as seen in Figure 7-4. Five sensors were arranged on a proto plate in a manner similar to how they would be arranged on the bottom plate of the machine. The first and most convenient detection method involved placing a magnet on each sensor. When a metal pin was lowered in front of the magnet, the magnetic field would change and the sensor output would show the change had occurred. The second method involved placing the magnets on the bottoms of the pins. When the magnet would get near the sensor, the field change would be detected. The tests monitored how all the sensors reacted to the pins being at various locations, and how different magnet shapes, thicknesses, and locations affected them. A method had to be found that would have the most reliable outputs from the detecting sensor while not interfering with the surrounding sensors.

PAGE 101

89 Figure 7-4. Sensor test setup For the first method, while the sensor outputs did relay the field change, they did so weakly. This made it difficult to distinguish pin appearances from common noise in the room. With the machine requiring 900 sensors, the inability to distinguish noise from actual pin contact could lead to many problems. The second method worked perfectly. The presence of a thin rubber magnet was strongly detected by the significant sensor from a short distance and no other sensors were noticeably affected. It was decided that the second method would be used and 1/32 thick magnetic tape would be used on the bottom of the pins. Unfortunately, difficulties with attachment to the board and detection by the digital I/O led to changing the type of sensor used. Contact Sensors The contact sensor method was easily tested by using the GUIs sensor test button. The board was hooked up to the digital I/O with a ribbon cable and the test was run for several seconds. Different types of metal were touched to the various sensor contact points on the sensor plate and the GUI would sa y which sensor was reading contact. This

PAGE 102

90 test showed that solid pieces of metal were very unreliable when used to contact the sensors. Something more flexible was necessary and the steel wool proved perfect. Testing the Steppers, Drivers, and Digital I/O In the process of interfacing the computer with the machine, several tests were run to make sure parts were running correctly and more efficiently. The stepper motor used to move the bottom plate up and down was wired to a stepper motor driver, as seen in Figure 7-5, which was wired to the digital I/O, which was plugged into the computer. The motor’s abilities (direction, speed, torque, step angle) were tested and modefied until it was running most efficiently. It was decided that the motor would be used with a step angle 7.5 degrees and would turn a shaft at 2/3 rev/s. A simple program was made for running the motor at its desired rates that would be called by the path planner. Figure 7-5. Stepper motor wired to a driver The part of the path planner program that tells the pin display device how to shape itself was tested using the setup shown in Fi gure 7-6. A display was chosen on the GUI

PAGE 103

91 and when the “send” button was clicked, the stepper motor was run when the bottom plate was to be moving and an array of LEDs were lit to represent individual clamps being activated. The same setup was also used for testing the method of detecting pin placement on the pin display device and making the computer show the image. The stepper motor was run when the bottom plate would be moving and switches could be thrown to mock pins making contact with the sensors. A set of LEDs was used in series with the switches in order verify the flowing currents. Figure 7-6. Path planner test setup Final Results During its assembly, the physical display, at best, could only be test in segments of the tasks that it would need to perform. It was not until the absolute completion of the device’s assembly that either of the two main tasks could be run in their entirety. As

PAGE 104

92 expected, both the physical virtualization and virtual visualization tasks did not perform exactly as desired on their first runs, but both had correctable problems that were dealt with in the most efficient manner possible. Physical Visualization Because of the delicacy of the pin systems and the difficulty of putting them together precisely, it was expected that some of the pins systems would not function when physical visualization was performed. When the first tests were run far more pins malfunctioned than expected. Several varied shapes were attempted for physical visualization and only about 20 to 30 pins would work properly on each run. Some of the malfunctions were that the pins got stuck in a position height above where they were supposed to be, but an overwhelming majority of the malfunctions were that the pins were slipping and falling down to their lowest positions. Often the pins would visibly hold to their desired positions for a few seconds before falling loose. There are several reasons for this slipping: Pencil Clamps: The main cause of the slipping is the pencil clamps. While the clamps perform ideally as far as holding the pins when closed and letting the pins move freely when open, they are weak at toggling between the two and staying toggled. There is no real snap to hold the freedom cup into place when it has been pulled down to close the clamp. In all preliminary tests the cup reliably held clamps shut, yet the combination of actual tasks going on in the final assembly of the physical display reliably allow the clamps to reopen. SMA Wires: The delay in clamp release was congruent with the amount of time it took the SMA wire to cool off. Many of the pins could be seen dropping right after the SMA wire returned to its original length. It’s possible that the minor stretching of the SMA wire caused just enough force to be pushed on the freedom rings that they came loose. It’s also possible that rings just came loose more easily once the SMA wires were no longer pulling them down. Stepper Motor/ Shaft Systems: Pulsing the stepper motors caused a vibration of the physical display. It is possible the intensity of the vibration may be a result of the shafts not all being perfectly aligned.

PAGE 105

93 Friction Prevention: The prevalent problem that occurred in all the preliminary tests for the clamping system was difficulty of movement by the pins because of friction. As a result, there was an over compensation in the assembly of the physical display making a strong attempt to prevent friction problems. Such adjustments included shortening of the pin guides and using thinner boards. The thinner boards were too flexible. Each clamp system attached pulled on a board above it and a board below it. As the systems were installed, the forces on the boards changed and because they were flexible, their shapes changed as well. This would explain why the systems that were installed last seemed to work best. Diodes: In order to compensate for the lack of precision involved in the hand assembly of the clamp systems, each diode was bent to a relative degree to make each system taught. Why this worked well at first, many diodes slowly were unbent with each successive pulling of the SMA on each run. Quick Fixes Assembling the physical display was both expensive and time consuming. Making a major change to correct the pin slipping problems would have taken an impractical amount or more time and money. However, a number of quick fixes were thought up and applied to the machine with the intention of making the physical visualization work as best as was possible. Heat: It was thought that if the room temperature of that the SMA wire was kept at was made a little hotter, the wire wouldn’t relax so much when it was deactivated and the freedom cup wouldn’t be pushed loose so easily. Bright spotlights that generated a good deal of heat were placed around the display and some tests were run. The lights were not seen to have any affect on the wires or the display. Surface Roughening: An attempt was made to increase the friction between the pins and clamps. First starch was sprayed on a pin from an aerosol can and a test was run. No changes were noticed. Next hairspray was sprayed on a couple pins and a test was run. This time there was a noticeable increase in friction. Unfortunately, there was too much friction and the pins would get randomly stuck. Also, it was not possible to spray the entire pins, which had already been installed, so some parts of the pin were too loose and others too tight. It was also noticed that occasional sparks occurred on the sensor board and it would be a bad idea to be using an aerosol around sparks. Therefore, this idea was not taken any further. Activation Time: The time that current was supplied to each SMA wire was increased from 2 seconds to 3 and 2.5 seconds for several tests. The hope was that the increased time would tighten the freedom cup on the clamp and prevent easy loosening. There appeared to be a very minor increase in functioning pins, which

PAGE 106

94 may or may not have been a result of the increase. Therefore, the activation time was left at 2.5 seconds. Stepper Motor Speed: When the test was first run, the stepper motors were run at their fastest speeds for the entire test. The process was changed so that the motors ran at their slowest speeds when the pins were capable of being knocked loose (when the table was going down). The decrease in step size decreased the vibration intensity and appeared to double the number of functioning pins. Fidgeting: The crudest fix used was to fidget with each of the diodes that could be reached so that those clamps systems remained taught. This appeared to have a minor. Figure 7-7 shows the result of a physical visualization test attempting to create a bowl shape with the pins. While a large percentage of the physical pins have fallen down by the end of the test, a slope towards the center of the grid is still noticeable. It is also noticeable that the number of functional pins per row increases with how recently that row was installed. (a) (b) Figure 7-7. Results of a physical visualizati on test where (a) is desired product that was formed on the GUI and (b) is the final product produced by the physical display Virtual Visualization When the virtual visualization tests were first performed, there were two obvious problems observed. The first problem was that about half the pins were not being held in place while display read the positions of the pins. This was easily explained. Using a

PAGE 107

95 thinner ground board had again caused a problem. The board was being pushed down by the solenoids on its four corners and was bending in an upside down bowl shape. The pins closest to the corners were being held best, but the pins closest to the center were the loosest. This problem was exaggerated by the fact that the solenoids were too weak to push the ground board as much as desired. The second problem was far more difficult to solve. The patterns displayed on the GUI were far different from the patterns created on the physical display device. This was made more inexplicable by the fact that the sensing process had worked in all previous tests. It had worked for tests involving individual pins and multiple pins; however, it turned out that there was a particular arrangement of pins that confused the sensors, which by chance had not been tested. When any three pins that are touching the sensor board form an “L-pattern” (one pin shares the same row of a second pin and the column of a third pin where the second and third pin have different rows and columns) the computer will sense a fourth pin at the column of the second pin and row of the third pin. This is a special case in the raster scanning process where the arrangement of pins allows a current to slip to a line where it should not be. It can occur in the four patterns displayed in Figure 7-8. The sensing process involves the pins maintaining contact with the sensors from first contact until the end of the process and therefore this special case arrangement is destined to occur for every single time this task is run.

PAGE 108

96 Figure 7-8. The four possible arrangements fo r incorrect sensor readings where the red line is the path of the current, the arrow heads are the direction of the current, the green dots are where the pins contact the sensor board, and the blue circle encompasses the surface that will be read as having pin contact when there is none The pin hold process could easily be fixed by placing a wedge in between the top board and the ground board so that the ground board is pushed down and the pins are held for the entire time the task is run. The stepping angle of the stepper motors was also decreased to its smallest possible amount in order to increase the strength of the motors and decrease the speed, which in turn decreased the vibration of the system. This kept loose pins from falling and allowed the sensing board to move the tightest pins with more ease. Unfortunately, the special case for the sensing process could not be completely fixed. Because the problem does not occur when only scanning a single row, it was

PAGE 109

97 decided that the test of virtual visualizati on would only test the front row in order to prove that everything else works. The sensors of all the rows except the front were covered with a sheet of paper and thereby rendered ineffective. Tests were then run for just the front row. These tests provided reliably identical images between the pattern of pins on the physical display and the GUI, as shown in Figure 7-9. Occasionally, one or two pins would go undetected due to those pins not lining up correctly with their assigned sensor on that run, but most tests were reliable. (a) (b) Figure 7-9. The results of a virtual visuali zation test where only the front row was tested; (a) shows the initial arrangement of the physical pins and (b) shows the virtual results (there is a difference in the pins behind the front row because the sensors for those pins were turned off and the computer assumed that those pins were at their highest positions)

PAGE 110

98 CHAPTER 8 FUTURE WORK This chapter presents possible solutions for previously mentioned problems, methods for improvements that could help the design, and general advice for future work on similar three-dimensional displays. General Changes If this machine was to be remade there are some very basic changes that would improve it a good deal. Use stronger stepper motors with smaller step angles. The smaller step angles would allow for smoother motion and decrease the vibration of the system. Increasing the strength would prevent jamming of the shaft system. Use stronger solenoids. While the solenoids that were used provided the desired force for minimal movement, the movement that was needed was greater than expected. Because the solenoids shafts get weaker as they get further from the shell, they provided insufficient force at times. Use thicker boards, especially for the ground and power boards. It is recommended that the current thickness of 0.062” be tripled or quadrupled. It was seen that the worries of pin friction were overcompensated for and the additional thickness would result in a significant increase in stiffness that would prevent bending. Lessen the wiring. The control box was initially wired to control 900 pins and 900 sensors, but is now only used for 225 pins and 225 sensors; however, because the wiring was still functional and there was no reason to fix something that wasn’t broken, nothing in the control box was changed. This means that there is more space and parts being used than is necessary. A second machine could use a smaller control box with less terminal boards, opto-isolator boards, ribbon cables and wires. Also, one of the digital I/Os could be eliminated. Use a pin made from a grittier material. The pins were too loose and a pin with a rougher surface likely would have had more surface friction with the clamp and held its position better. Another option would be to lightly spray all the pins with a lacquer or hair spray before installation.

PAGE 111

99 Major Changes While this project delivered a lot of important information in the creation of a 3D Visualization Display, it is unlikely that the current machine (even in perfect working order with all the general changes added) could properly serve the purposes that it has been created for. There are still improvements that need to be made within the physical display that will make the machine more exact, reliable, and efficient. These are some ways to upgrade the display and methods that may make those upgrades possible. Clamping The most necessary change for future work would be to evolve the clamps. Future devices need clamps that have a more definitive toggle. Some type of snap or bump-togroove setup would probably work well. The pre-designed clamps mentioned in Appendix A would probably be sufficient. Whatever clamp is put into use, it is necessary that it remain tightly in the position it has been toggled to until toggled again. Early in the assembly process, another possible clamp system was developed. At the time, it seemed far more complicated and was discarded; however, after the complications of assembly, it seems that it may be less complicated and has a more effective toggling action. The idea was that one could actuate an entire row at once and an entire column at once with pulleys. To lock a pin at position (i,j), the following is performed: 1. Actuate every pin in row i. 2. Actuate every pin in column j 3. De-actuate row i and column j Figure 8-1 exemplifies this procedure, where it is important to note that the left block moves with the pins’ rows and the right block moves with the pins’ columns. Before the procedure begins, all the pins look like (a) in the figure; there is slack on the

PAGE 112

100 pulley. After step one, all the pins in row i are taut like (b) while everything else still looks like (a). After step two, most of row i hasn’t changed and most of column j goes taut like (c); however, the pin at (i,j), which was already taut, pulls the pulley and collar down into the locked position to look like (d). After step three, all pins return to (a) except for pin (i,j); it goes slack, but it is locked until all the pins get reset. As for resetting, it could still be done in the same fashion currently being employed. Figure 8-1. Pulley system of clamp actuation where the left block moves with the pins’ rows and the right block moves with the pins’ columns Sensing A possible solution for the special case that prevents sensing the pin positions correctly involves rebuilding the system as follows. The physical display would only change in that the sensing board would basically be split into two boards, a board for pushing the pins and a board for sensing them. The pushing board would be identical to the current sensing board but would have no sensors (basically is would be just a plain flat plastic board). The new sensing board would have holes where the sensors are on the current sensing board so that the pins can pass through. The sensors on this board would be along the side of the holes or possibly in the form of brushes that would not impede the passing of the pins through the holes. Both boards would be able to move up and

PAGE 113

101 down the same distance that the current sensing board moves. The new sensing board would merely rest upon the pushing board any time that the pushing board would be in use. When sensing is to occur, the pushing board would remain at its lowest position and the sensing board could be raised up by itself. This would improve the system in that the physical display of pins would not be destr oyed during sensing as it is in the current device. The major change would be that each pin would have two points that would get read as opposed to the current design where each pin has only one. The read points could be made by adding a magnet, steel wool, or whatever material the sensor in use is capable of detecting. Each pin would get read at its lowest point and a specified higher point. In the current design, the sensing board reads one column at a time. The special case occurs when a pin is sensed at an incorrect row in that column. Therefore, the new design would have the second read points placed at equal distances from the first read point on the columns but different distances on the rows, as shown in Figure 8-2. This setup will allow contact to occur at specific points in time rather than for the remaining amount of time. The computer will know that there need to be a specific amount of time between contacts and should it experience more than two contacts, the correct two had a specific amount of time between them. The pin would be place at the first of those two. This will make it far less likely that the special case will occur and on the rare occasion when it does, it will not be able to occur at the second reading because of the different heights.

PAGE 114

102 Figure 8-2. A two-level sensor reading system in a raster scan arrangement where the red ovals represent sensor holes, the blue squares represent the single-level first read points, and the green dots represent the multi-level set of second read points Conclusions The following conclusions can be made as a result of this development project: Shape memory alloy wires were capable of working a clamp system for a grid of closely spaces pins The speed of the physical visualization process was most dependent on the time it took to activate the SMA wires which was about two seconds per wire The speed of the virtual visualization process was most dependent on how fast the stepper motors could move the sensor plate from its lowest position to its highest position The clamps used were capable of holding the pins with the necessary amount of friction The toggling system used by the clamps was completely inefficient as the brass freedom cup slipped far too easily on the clamp

PAGE 115

103 A single level of contact sensors could not be used to reliably detect a grid of objects when wired in a raster scan arrangement (while a second level of sensors might correct that problem) The sensor arrangement was capable of detecting a single row of pins with at least 80% accuracy at a reliable rate The computer driven by the GUI and path planner programs was capable of directing all the parts being controlled (motors, solenoids, sensors, opto-isolators) as fast as was necessary Opinions and Personal Advice The clamping system requires far too much precision to be assembled by hand and provide reliable results. Finding a more mechanical method of assembly (perhaps from another machine) would not only provide a more effective device, but also allow for a tighter packing of the pins. In fact, the mini ature scale of parts being dealt with in this project is so small that an argument could be made that the best route of design would be through use of microfabrication. The greatest possible improvement that could be made to this device would be to make the clamping systems modular. The machine would simply be far more efficient in assembly, repair, and maintenance if the clamping system could be plucked in and out of the display at any given time without affecting the surrounding parts.

PAGE 116

104 APPENDIX A DEVELOPMENT OF THE FIRST PIN CLAMP This section of the appendix describes the creation of a clamp from design to actual creation. The design was eventually scrapped for the use of a pre-made mechanical pencil clamp. Home design With a basic structure for the actuation process established it was necessary to create a more detailed design of the clamp. The top cylindrical section of the clamp design, pictured in Figure A-1, known as the “head”, serves as the attachment that will ground the clamp to the top table. This piece w ill be pressed into the table and will be the only part of the clamp that does not move. The pin will go through the hole in the center of the head. The hole will be large enough that the pin will be able to move freely through it.

PAGE 117

105 Figure A-1. Initial clamp design Extending from the head are the clamp’s “arms.” These perform the actual clamping action and are where all the actual bending will be occurring. In deciding the number of arms necessary, it was realized that having two arms would allow for slipping of the pin in some directions tangent to the axis while having three or more arms would keep the pin on a single path. In order to keep manufacturing time and complexity to a minimum, it was decided to have three arms. Each of the arms has two bumps extruding toward the center (where Nub 1 is the top bump, and Nub 2 is the bottom bump). At steady state the diameter of the free space in-between the three Nub 1’s is smaller than th e diameter of the pin. Therefore, when the pin is inserted into the clamp, Nub 1 on each of the three arms will be pressing against the pin, preventing it from moving. When the collar is pressed in between the circle of Nub 2’s, the arms will be extended further in an outward direction. The second set of nubs will hold the collar in place so that the arms remain open. The first set of nubs will have been pulled far

PAGE 118

106 enough from the center that the pin will move freely. Heating a SMA that is attached to the collar will cause the collar to be pulled from the clamp and return the clamp to its previous state. Therefore, pressing the collar into the clamp allows for free motion of the pin, and pulling the collar out of the clamp prevents motion of the pin. Finite Element Analysis The initial design was given dimensions that were convenient for use in the parent design while still realistic for manufacturing. With the basic shape of the pin clamps decided, an FEA analysis could be used to parameterize the dimensions. The results would be used to derive the final design di mensions. The dimensions analyzed revolved around the initial dimensions. Test 1 The first decision that needed to be made was what material the clamps would be made from. For manufacturing purposes, it seemed most practical to make the clamps either out of a plastic or a metal. Therefore, an analysis using the initial clamp dimension was performed on a clamp made from each material where delrin was the plastic used and general isotropic steel, shown in Figure A-2, was the metal used. The analysis determined the amount of force that the clamp would return when the nubs were displaced a common amount (0.025”). The analysis showed that the plastic returned a force unacceptably small force while the steel return a more suitable force value that was 80 times greater than the plastic. This made steel the definitive choice for material.

PAGE 119

107 Figure A-2. FEA analysis for material choice Test 2 The next analysis was performed in order to determine the size and placement location of Nub 1 on each clamp arm. This test used a constant beam cross-section while varying the location of the nub. A beam analysis was performed on each of eight possible locations for Nub 1. The tested locations, or nodes, were spaced 0.125” apart, as presented in Figure A-3, where Nub 1 is at the head of the clamp. The analysis was performed by varying the force input to each nub location, then, in turn the maximum stress and displacement of the nub was output. The best location would have the greatest nub displacement, while the max stress was under 29,000psi (the stress level when fatigue begins to set in) and the force from each arm was about 1/3 to 1/9 lb. The best placement proved to be 0.75” from Node 1 at the start of the arm.

PAGE 120

108 Figure A-3. Node placement on beam analysis for Nub 1 placement Test 3 Thirdly, a test was performed to see the effects of increasing the cross-section. While increasing the cross-section would be detrimental to the parent design (because the pins would have to be spaced further apart from one another), this analysis would show if the benefits of increasing the cross-sectional area would outweigh the negatives. The cross-sectional area was enlarged by increasing the outer diameter of the clamp arm only. Like the previous analyses, this analysis was performed by inputting different forces at Nub 1 and outputting the maximum stress and the nub displacement. Using the previous analysis as a starting point, nub locations tested were placed around 0.75”. The analysis showed that increasing the outer diameter actually had negative results and increased the max stress. This made sense as the extra material would make the arms stiffer and thus create greater difficulty in reaching the desired displacements. Test 4 When it was realized that the cross-sectional shape would be inconvenient for the manufacturing process, the cross sectional shape was changed from a tube with triangular slices removed to a tube with rectangular sections removed. This new cross-section could easily be achieved by slotting the tube axially with a saw blade. Test 2 was

PAGE 121

109 repeated on the new cross-section, however, the previous results allowed the testing points to be narrowed to the most successful points of the previous analysis. The results provided the same peak points while showing the design would be more helpful to the overall needs of the design. Test 5 When it was realized that the steel rod was most commonly found in fractional units rather than decimal the outer diameter was changed from 0.1” to 3/32”. Test 4 was repeated and surprisingly, the results were almost identical to those of Test 2; therefore, Nub 1 was placed at Node 7 (0.75” from the head of the clamp). Test 6 The final decision was the placement of Nub 2. An analysis was performed using the same process followed in the previous five tests on a beam with the same crosssection as that of Test 5 where a fifteen-degree outward bend occurs at Node 7. Points were tested at four nodes placed 0.125”, 0.25”, 0.5”, and 0.75” from Node 7 as shown in Figure 16. It was decided to place Nub 2 a distance of 0.5” from Node 7 based on four concerns: 1. Make certain the max stress is under 29,000psi 2. Try to have force of about 1/3lb or more on the cylinder 3. Make certain the displacement at Node 7 (location of first nub) can be greater than 1.25E-2” 4. Location of the second nub will provide adequate room for the cylinder to fit between while still not touching the pin

PAGE 122

110 Figure A-4. Node placement on beam analysis for Nub 2 placement The Clamp Manufacturing Process The initial idea for the manufacturing of the clamps was to make it out of a steel rod. While it was preferable to start with an already hollow tube, no tube mills could be found that made steel tubes with the exact de sired dimensions (OD=3/32" and ID=0.05"). In fact, it was difficult to find a steel rod with an outer diameter as small as 3/32". The softest most manufacturable steel rod that could found at the desired size was drill rod made of type A-1 steel. The drill rod was hollowed out with a drill bit (OD = 0.0492") in a lathe. Next the three arms could be cut out of the hollow drill rod by placing three slots along the centerline of the rod in 120-degree increments. The hollow rod would be slotted through use of a computer numerically controlled (CNC) Mini-mill. Attached to the mill's moving x-y plate was an indexable fixture with a collet in it. The collet would hold one end of the hollow rod and the fixture would allow for rotation of the rod after each slot. The other end of the rod would be held in a metal block that was also attached to the moving x-y plate. The end of the rod would be placed in a hole in the metal block. The placement would not prevent movement in the direction of the cut feed (that was

PAGE 123

111 done by the collet), but it would prevent motion in the directions that would allow for bending of the rod. However, it was discovered that the bit used to hollow out the rod could only clear out a depth of about 3/4" on each end of the rod (1.5" total). The slotting process required a hollow rod with 1.25" for cutting and 1/2" at each end for clamp-to-blade clearance. This meant that the drill rod could not be used for slotting. The best solution was to use stainless steel tubing for the slotting. The inner diameter was larger than would have been preferred (ID=0.538") and the stainless steel (90 Brunell hardness) was much harder than the drill rod. In case this method did not work, another method would be using the drill rod. Rather than hollowing out the rod, a center hole would result from the intersection of the three slots placed on the mini mill. The method is less preferable because the three arms would have flat inner faces opposed to the curved faces resulting in the previous method. Having curved faces gives the clamps a greater contact area with the pins. Fortunately, the stainless steel tubing could be slotted by the CNC machine; however, a slow federate had to be used. Three-inch slots were placed along the sides of 3.5” long tubes. The slotted tubes could then be cut in half for the creation of two clamps. A mandrill was designed so that the slotted tubes could be given the desired bends. The mandrill was rapid prototyped using a FDM prototyping process. It functioned as a plastic shell where the slotted tube was placed inside the mandrill and when pressed in a vice, the tubing was bend. The dipping process was performed successfully. This involved dipping part of the clamp in a primer, and once dry, dipping the primed portion in a plastic. It was found that to get the desired thickness of plastic, it

PAGE 124

112 is best to dip the clamp in plastic multiple times. Figure A-5 shows the first completed clamp prototypes. Figure A-6 shows the evol ution of the manufacturing process in the creation of the clamp Figure A-5. Completed clamps Figure A-6. Evolution of clamp creation Sample clamps were made in order verify this process. The first step involved slotting the tubes so that they had arms. The arms were then stamped over a mandrill, which bent the arms into the desired shapes. Finally the arms were dipped in a coat of plastic so that they had the desired grip.

PAGE 125

113 For the stamping process, a mandrill was designed and rapid prototyped using a FDM prototyping process. While the plastic mandrill was proven to be capable of performing the desired bends on the slotted tube, it began falling apart after four uses. The dipping process involved dipping part of the clamp in a primer, and once dry, dipping the primed portion in a plastic. It was found that to get the desired thickness of plastic, it is best to dip the clamp in plastic multiple times. Figure A-7a shows the first completed clamp prototypes. Figure A-7b s hows the evolution of the stainless steel tubing into a completed clamp. (a) (b) Figure A-7. Manufactured clamps (a) Completed clamps, (b) Clamp changes through production

PAGE 126

114 APPENDIX B THE RASTER SCAN TEST REPORT It was decided that the number of controller outputs could be minimized by wiring the actuators with a raster scanning technique, which will be explained in further detail later in the report. This form of wiring would require the use of opto-isolators. Combining the fact that the relay-boards that were on hand held a maximum of sixteen isolators with the desired size of the pin array, it was decided that the next machine would have a 30x30 grid of pins with the remaining isolators left open for other necessary connections. Therefore, it would be necessary to have a circuit that would control a 30 x 30 array of Shape Memory Alloy (SMA) wires. This decision was followed by a project where a scaled down model (3 x 3) was created in order to verify the circuit design for the actual machine was acceptable. The system was modeled with a 3 x 3 array of LED’s where LED’s acted as the SMAs. A keypad was used to tell which lights to light up. Figure B-1 shows the position of the lights and their labels. The keypad contained keys with the numbers one through nine. Pressing Button One would result in L1 lighting up. Pressing Button Two would cause L1 and L2 to light up, and so on, until pressing Button Nine lights up L1 through L9 (all the LED’s).

PAGE 127

115 Figure B-1. Configuration of LED’s The raster scanning technique makes it so that only 2n controller outputs are needed for any nxn array of actuators. In this case, nine LED’s were capable of being lit by six controller outputs. The system was created by configuring the circuit so that each LED was in series with all the other LED’s in the same row and column. This configuration meant that only 6 outputs were needed for the model (3 grounds for the columns and 3 voltage sources for the rows). In order to make certain the desired LED display occurred, only one column could be lit up at one time; however, the lighting of each column would occur so quickly that the human eye would perceive all the columns as being lit up at once. Only one ground would be given to any column at any time, thus allowing for only one column to be lit up at any time. The process would begin by the first column of lights being given a ground and cu rrent being sent to the appropriate rows. The ground would then be shifted to the second column and its appropriate rows would be given current. Finally, the third column would be given the ground and once again the desired rows would be given power. The process would occur so quickly that it would need to be repeated hundreds of times before a human eye would notice any light, and at

PAGE 128

116 that point it would appear that all the lights were being lit simultaneously. Figure B-2 shows the wiring diagram. Figure B-2. The Wiring Diagram For the LED Display The System, displayed in Figure B-3, was powered by an APX Technologies Inc. class 2 power supply and composed of the following items.

PAGE 129

117 Figure B-3. The Assembled Electrical System The supply sent a line of five volts and a line of ground to a JAMECO JE24 breadboard The breadboard contained the LED display, the microprocessor, and the keypad. The Velleman Matrix Type Keypad had six lines: a line to each of the 3 columns, and a line to each of the three rows. Pressing a button created a short circuit between the wire connected to that row and the wire connected to that column. Finding which button has been pressed was accomplished by performing a polling technique. The column lines of the keypad were attached to output pins of the microprocessor and the three row lines were attached to three input pins of the microprocessor. When a current was sent to the column of a button when it was pressed, the input pin attached to the button’s row read a high value; otherwise, the pin read a low value because a pull-down resistor was attached to all the input/row lines. The pull-down resistor prevented floating and kept the input/row lines low when no button was pushed. The processor performed a button-by-

PAGE 130

118 button search in order to find which had been pushed. First a high output would be sent to the column being checked. Then row inputs were read. If a high voltage were found entering a pin, the processor would know that the button at that row and column had been pressed and the LED display was configured appropriately. Then the processor would continue polling the columns until a button is pushed again. The Potter and Brumfield relay board along with six CRYDOM ODC5 optoisolators were used in order to keep the system as close to the 30 x 30 system as possible. Attaching the items together created a six-relay system. The system ran on five volts drawn from the breadboard. Six output pins from the microprocessor were attached to six control pins on the relay board. The relay board had six outputs; each attached to an input to the LED display. Three of the relay board’s outputs were in series with grounds and the other three outputs were in series with a voltage source (the ground and five volt source were drawn from the breadboard, but this will be different for the case of the 30 x 30 array of SMAs). However, a current could only flow between the relay board output and the attached ground or voltage source when the attached opto-isolator had been activated. When the microprocessor wanted to activate an output on the relay board, it would send a low output to a particular control. This would in turn activate the attached opto-isolator, which in turn would allow curre nt to flow between the attached output and ground/voltage source. Ground outputs were connected to the columns in the LED display and voltage outputs were attached to rows of the LED display. When the microprocessor triggered a ground output and voltage output on the relay board, a light on the LED display would light up at the corresponding row and column.

PAGE 131

119 The PICmicro 16F627 was used to control all actions in the system. Six outputs from Port B were connected to the relay board and used to control the LED display. Three outputs and three inputs from Port A were used to poll the keypad. The microprocessor found the button pushed and created the LED display as desired on each test as shown in Figure B4 and Figure B-5. The system design and coding algorithms used on the model allow for easy conversion to an n x n system of actuators, where n is the number of rows in the array (n is also equal to the number of columns). Figure B-4. Five LEDs Being Lit Figure B-5. Nine LEDs Being Lit

PAGE 132

120 One major concern that arose in the testing of the system model was that current was going through LEDs that should not have been receiving any. It could be seen that any column or row that had at least one LED lit up would result in all the LEDs in that column or row emitting some amount of light. The light from the “misfiring” LEDs could be decreased to nearly undetectable levels by adding delays to the code when a column is lit up and/or turning off all the LEDs before displaying the following column configuration. Adding a delay to the time that the lights remained off between column displays also decreased the amount of undesired emitted light. A pull-up resistor was also added to the inputs of the LED columns, but this failed to have any effect on the system. The reason for this undesired light was not clear but the following are possible: 1. The opto-isolators could not switch as quickly as desired. 2. The quick change of outputs created a leak from one display to another on the column outputs. 3. Some item in the breadboards allowed for a slight amount of grounding to occur in the columns at all times. Some other important occurrences and observances noted for future work include the following: 1. Some opto-isolators never worked. 2. The opto-isolators only allow current to flow in one direction through the output lines. 3. Continuity never occurs between the opto-isolator outputs

PAGE 133

121 APPENDIX C ODDS AND ENDS Hole Pattern The design of the hole pattern began while the clamp design still involved using a V-shaped strand of SMA wire that would be attached on opposing ends of the pin. In order to form an idea of the machine’s shape and the placement of its parts, an initial decision needed to be made on the hole patterns for the pins that would pass through the spinal plates. The patterns had to be made based on the following desires: 1. There can be no overlapping parts 2. Have pins as densely packed as possible, 3. Use a symmetric hole pattern, 4. Allow as much open space as possible for convenience during assembly and maintenance, and 5. Remove as little as possible from the plates to prevent bending, fracturing, and breaking. Three hole patterns were created and evaluated. In looking at the parts inserted into the plates, preventing overlapping between the pins and the heads of the vented screws would prevent all possible overlapping. Therefore, the pin patterns were created with the placement of each pin and the two, vented screws that were assigned to each pin in mind. The first and densest design pattern, shown in Figure 2.1b-17, was so close to overlapping that it seemed certain the addition of any form of tolerance error or incorrect dimension would lead to assembly problems. The second design, seen in Figure 2.1b-18 eliminated all overlapping worries but there was still worry about the structural integrity

PAGE 134

122 of the plates with the holes so close. The third and most space consuming design, displayed in Figure C-1, provided the greatest f actor of safety while not sacrificing much in the way of pin separation. (a) (b) (c) Figure C-1. The three hole patterns that were analyzed It was decided to use the third design, but an extra level of four plates would be added to the machine so that the number of necessary holes for vented screws would be cut in half on each plate. With the changing of the design, the screw holes were eliminated and it was decided to do a simple row offset pattern in a 30x30 set. When it was found that additional space was necessary for assembly, every other hole was skipped and the hole pattern became square.

PAGE 135

123 Wiring Scheme With the aim of keeping the physical display aesthetically pleasing and safely arranged, its wiring was weaved through the extruded aluminum frame. Wires moving vertical were threaded down the center of the extruded aluminum bars. Wires moving horizontal were held in the side grooves of the bars with specially made plugs. SMA Attachments A method for attaching the ends of the Shape Memory Alloy (SMA) wires to electrical connections was designed. Figure C-2 shows the basic machine setup as far as plate arrangement and pin module design. The method has each end of the SMA pass through the center of a vented screw as seen in the close-up cross-section view in Figure C-3. The vented screws have been inserted into the bottom of a nonconductive stable board. The bottom of the board will be attached to a pc board with wire traces etched into it. The end of the SMA will come out through the head of the screw and be tucked under a washer against the tightened screw head. The wire will be held in place between the washer and a part of the board containing the end of a wire trace. One board will allow the current to enter the SMA and the other will allow the current to leave the SMA. Figure C-2. Table Arrangement

PAGE 136

124 Figure C-3. Close-up of electrical connection It was realized that the miniature vented screws to be used for connecting the SMA wires to the Electrical board would be difficult to attach through common means. The screws were intended to be attached by inserting an Allen wrench into the head cap; however, because the SMA wire would be passing through the head cap, the insertion is not possible. Also, because the head cap is cylindrical, it could not be gripped by a hexwrench. Two new screw designs were created and sent to small parts manufacturers for price quotes. The first and most useful desi gn is shown in Figures C-4 and C-5. The screw is identical to the original design except that it has a hole just below the head cap. With this design, the SMA wire could be pulled out of the hole before passing through the head cap. This would allow for insertion of the Allen wrench into the head cap and the bonus result would have the SMA wire already under the head cap, so the task of looping the SMA would be eliminated.

PAGE 137

125 Figure C-4. Design 1 concept drawing Figure C-5. Design 1 detail drawing The more basic second design, displayed in Figure C-6 and C-7, differs from the original screw design in that the head cap has been changed from a circular shape to a hex shape and the counter bore for the Allen wrench insertion has been eliminated. The result would make the head cap more easily gripped from the outside by a wrench.

PAGE 138

126 Figure C-6. Design 2 concept drawing Figure C-7. Design 2 detail drawing The manufacturing of both designs proved to be far more expensive than would have been desired, so they were scrapped. In an effort to prevent from losing any more time on this problem, it was decided that the screws would have to be turned by inserting an Allen wrench that was a couple sizes smaller than the insert hole into the head cap. If this method proved to be too harmful to the SMA wires, then the screws would be turned by gripping the head caps with a pair of pliers. If this method proved to be ineffective as well, then a special tool would have to be created

PAGE 139

127 SMA Spring Designs Discovery of a spring-shaped SMA opened the way for new, more efficient designs (Huettl et al. 2000). This material, depending on the type, can either be stretched or compressed, and when heated will return to its original shape. The SMA spring can replace the SMA wire in the original design as shown in Figure C-8. An inelastic band will be threaded around the brass cup, as the SMA wire would have been. Instead of being connected to the board, the ends will be connected to a ring that loosely encompasses the pin. The ring will then be attached to the board with the SMA spring. As in the original design, heating the SMA will pull the brass cup down and clamp the pin. Also like the previous design, all the clamps will be opened and the SMA will be stretched by pushing the brass cup upwards with a moving plate. This method cuts the number of boards with electrical traces in half and lowers the number of holes in each board to one hole per pin.

PAGE 140

128 Figure C-8. SMA spring actuation with inelastic band The most ideal design comes from the arrangement displayed in Figure C-9. This design utilizes an SMA that expands when heated. The SMA is attatched to the bottom of the top board and wraps around the pin. The other end is attatched to a ring that holds the brass cup. When the SMA receives a current from the top board, it expands and pushes the cup over the clamp and the pin is gripped. As with the original design, a moving plate lifts the cup and resets the SMA. This design has the following benefits: 1. Less Perforated Boards: The design goe s from needing eight perforated boards with four containing electrical traces to two perforated boards with one needing electrical traces. 2. Fewer holes: Each bored only needs one hole per pin. 3. Eliminates need for miniature vented screws 4. Easier assembly: There are fewer parts and attachments arent as complicated. The boards are no longer connected 5. Easier maintenance: Having less parts and taking up less space, there will be less possibilities for parts breaking and it will be easier to make adjustments. 6. Eliminates size limitations: In the original design, the spacing of the pin separation was governed by the bulkiness of the actuation setup. With SMA springs having coil diameters as small as 0.20 mm, the size limitations are now governed by how small the clamps and pins can be made.

PAGE 141

129 Figure C-9. Ideal spring SMA actuation design Because SMA springs that compress when heated are more common than those that expand, a design was created using the more common SMA spring, as shown by Figure C-10. Once again the SMA is wrapped around the pin and connected to the top board, which supplies the necessary current. However, in this case, the other end of the SMA is attached to the clamp instead of the brass cup. Heating the SMA pulls the clamp upwards into a stable perforated plated, causing the clamp to close and the pin to be gripped. A moving plate resets the SMA by pushing the clamp downward and out of the stable perforated board. For Virtualization, a moving plate can be moved upward to hold the clamp inside the stable perforated board in order to keep the pins gripped when they are being adjusted. This design maintains all the benefits of the previous design, except to a lesser degree in some cases.

PAGE 142

130 Figure C-10. Compressing SMA spring actuation design The only SMA springs that could be found in mass production were either significantly larger in diameter than the pins or significantly smaller. This led to the design shown in Figure C-11. The only differen ce in this design from the previous one is that the SMA location has been moved from being around the pin to being beside it. This may increase the space between pins from that of the previous design, but because the SMA is so thin, it is considered a negligible change.

PAGE 143

131 Figure C-11. Miniature compre ssing SMA spring actuation design

PAGE 144

132 APPENDIX D SPECIFICS OF WIRING Pin Assignment Table Table D-1 maps each pin on the digital I/Os. It details the specific pin names, the connected ribbon cables, and which board inputs and outputs are controlled. Table D-1. Pin assignments of the digital I/Os Digital I/O to Ribbon Cable Labels Power BoardGround Board Sensor Board Output Ou tput Input Input Variable # n = 1 2 3 4 x = a b c b = 0-7 RibbonPower Ground Power (Read) Program Call Name Card Ribbon # on Card PortPinRibbonPins Rows 2ndCols 2nd Rows 2ndCols 2nd out_row_power 0 2 c 7 2 1 1 out_row_power 0 2 c 6 2 3 2 out_row_power 0 2 c 5 2 5 3 out_row_power 0 2 c 4 2 7 4 out_row_power 0 2 c 3 2 9 5 out_row_power 0 2 c 2 2 11 6 out_row_power 0 2 c 1 2 13 7 out_row_power 0 2 c 0 2 15 8 out_row_power 0 2 b 7 2 17 9 out_row_power 0 2 b 6 2 19 10 out_row_power 0 2 b 5 2 21 11 out_row_power 0 2 b 4 2 23 12 out_row_power 0 2 b 3 2 25 13 out_row_power 0 2 b 2 2 27 14 out_row_power 0 2 b 1 2 29 15 out_row_power 0 2 b 0 2 31 16 out_row_power 0 2 a 7 2 33 17 out_row_power 0 2 a 6 2 35 18 out_row_power 0 2 a 5 2 37 19 out_row_power 0 2 a 4 2 39 20 out_row_power 0 2 a 3 2 41 21 out_row_power 0 2 a 2 2 43 22 out_row_power 0 2 a 1 2 45 23 out_row_power 0 2 a 0 2 47 24 out_col_ground 0 3 c 7 3 1 30 out_col_ground 0 3 c 6 3 3 29 out_col_ground 0 3 c 5 3 5 28 out_col_ground 0 3 c 4 3 7 27

PAGE 145

133 out_col_ground 0 3 c 3 3 9 26 out_col_ground 0 3 c 2 3 11 25 out_col_ground 0 3 c 1 3 13 24 out_col_ground 0 3 c 0 3 15 23 out_col_ground 0 3 b 7 3 17 22 out_col_ground 0 3 b 6 3 19 21 out_col_ground 0 3 b 5 3 21 20 out_col_ground 0 3 b 4 3 23 19 out_col_ground 0 3 b 3 3 25 18 out_col_ground 0 3 b 2 3 27 17 out_col_ground 0 3 b 1 3 29 16 out_col_ground 0 3 b 0 3 31 15 out_col_ground 0 3 a 7 3 33 14 out_col_ground 0 3 a 6 3 35 13 out_col_ground 0 3 a 5 3 37 12 out_col_ground 0 3 a 4 3 39 11 out_col_ground 0 3 a 3 3 41 10 out_col_ground 0 3 a 2 3 43 9 out_col_ground 0 3 a 1 3 45 8 out_col_ground 0 3 a 0 3 47 7 out_row_power 0 4 c 7 4 1 25 out_row_power 0 4 c 6 4 3 26 out_row_power 0 4 c 5 4 5 27 out_row_power 0 4 c 4 4 7 28 out_row_power 0 4 c 3 4 9 29 out_row_power 0 4 c 2 4 11 30 out_col_ground 0 4 c 1 4 13 6 out_col_ground 0 4 c 0 4 15 5 out_col_ground 0 4 b 7 4 17 4 out_col_ground 0 4 b 6 4 19 3 out_col_ground 0 4 b 5 4 21 2 out_col_ground 0 4 b 4 4 23 1 in_row_power 0 4 b 3 4 25 1 in_row_power 0 4 b 2 4 27 2 in_row_power 0 4 b 1 4 29 3 in_row_power 0 4 b 0 4 31 4 in_row_power 0 4 a 7 4 33 5 in_row_power 0 4 a 6 4 35 6 in_row_power 0 4 a 5 4 37 7 in_row_power 0 4 a 4 4 39 8 in_row_power 0 4 a 3 4 41 9 in_row_power 0 4 a 2 4 43 10 in_row_power 0 4 a 1 4 45 11 in_row_power 0 4 a 0 4 47 12 in_row_power 1 1 c 7 5 1 13 in_row_power 1 1 c 6 5 3 14 in_row_power 1 1 c 5 5 5 15 in_row_power 1 1 c 4 5 7 16

PAGE 146

134 in_row_power 1 1 c 3 5 9 17 in_row_power 1 1 c 2 5 11 18 in_row_power 1 1 c 1 5 13 19 in_row_power 1 1 c 0 5 15 20 in_row_power 1 1 b 7 5 17 21 in_row_power 1 1 b 6 5 19 22 in_row_power 1 1 b 5 5 21 23 in_row_power 1 1 b 4 5 23 24 in_row_power 1 1 b 3 5 25 25 in_row_power 1 1 b 2 5 27 26 in_row_power 1 1 b 1 5 29 27 in_row_power 1 1 b 0 5 31 28 in_row_power 1 1 a 7 5 33 29 in_row_power 1 1 a 6 5 35 30 EMPTY 1 1 a 5 5 37 EMPTY 1 1 a 4 5 39 EMPTY 1 1 a 3 5 41 EMPTY 1 1 a 2 5 43 EMPTY 1 1 a 1 5 45 EMPTY 1 1 a 0 5 47 in_col_read 1 2 c 7 6 1 1 in_col_read 1 2 c 6 6 3 2 in_col_read 1 2 c 5 6 5 3 in_col_read 1 2 c 4 6 7 4 in_col_read 1 2 c 3 6 9 5 in_col_read 1 2 c 2 6 11 6 in_col_read 1 2 c 1 6 13 7 in_col_read 1 2 c 0 6 15 8 in_col_read 1 2 b 7 6 17 9 in_col_read 1 2 b 6 6 19 10 in_col_read 1 2 b 5 6 21 11 in_col_read 1 2 b 4 6 23 12 in_col_read 1 2 b 3 6 25 13 in_col_read 1 2 b 2 6 27 14 in_col_read 1 2 b 1 6 29 15 in_col_read 1 2 b 0 6 31 16 in_col_read 1 2 a 7 6 33 17 in_col_read 1 2 a 6 6 35 18 in_col_read 1 2 a 5 6 37 19 in_col_read 1 2 a 4 6 39 20 in_col_read 1 2 a 3 6 41 21 in_col_read 1 2 a 2 6 43 22 in_col_read 1 2 a 1 6 45 23 in_col_read 1 2 a 0 6 47 24 in_col_read 0 1 c 7 1 1 25 in_col_read 0 1 c 6 1 3 26 in_col_read 0 1 c 5 1 5 27 in_col_read 0 1 c 4 1 7 28

PAGE 147

135 in_col_read 0 1 c 3 1 9 29 in_col_read 0 1 c 2 1 11 30 board_switch 0 1 c 1 1 13 EMPTY 0 1 c 0 1 15 step 0 1 b 7 1 17 direction 0 1 b 6 1 19 EMPTY 0 1 b 5 1 21 EMPTY 0 1 b 4 1 23 EMPTY 0 1 b 3 1 25 step_res_con_1 0 1 b 2 1 27 step_res_con_2 0 1 b 1 1 29 EMPTY 0 1 b 0 1 31 solenoid_up 0 1 a 7 1 33 solenoid_dn 0 1 a 6 1 35 dumb 0 1 a 5 1 37 dumb 0 1 a 4 1 39 dumb 0 1 a 3 1 41 dumb 0 1 a 2 1 43 fan_switch 0 1 a 1 1 45 EMPTY 0 1 a 0 1 47

PAGE 148

136 APPENDIX E PROGRAM CODE Guide to Code clicks.c++ Global variable describing the length of a pin in inches ('pinLength') and utilities to convert between inches and 'clicks' of the stepper motor. o double pinLength = 4.1; // In inches. o int step_fraction_inv = 1; // In clicks. The stepper motor can operate at several different speeds. At the fastest speed, one step corresponds to one click of which there are 48 per revolution. At the slowest speed one click corresponds to eight steps. This variable contains the number of steps per click. o double clicksToGUI(int clicks); This function converts the number of clicks to the real, physical inches. It depends on the speed of the motor. o int GUIToClicks(double inches); Inverse of above. o int MAX_CLICKS = GUIToClicks(pinLength); Total number of clicks to go from bottom to top, to traverse `pinLength` inches. CompDisplay.c++ Function to read pin heights from the machine. o long delay_nsec = 6000000; Time in nanoseconds to pause when changing which row gets power. We need this because it takes time for the power on a row to set up and to power down. If this is set too low, multiple rows will receive power at once and when we see that a column also has power, we won't know from which row it received its power. o void wait(long nsec); This will sleep for 'nsec' nanoseconds. We cannot use the system call 'usleep' because it lacks the necessary resolution. E.g., if we ask it to sleep for one milisecond, it might sleep for as much as 20 milliseconds. Please keep nsec < 1e9.

PAGE 149

137 o bool isPinEngaged(int row, int col); Returns 'true' if the pin at (row,col) is engaged. 'row' and 'col' indexed from zero. o void receivePhysical(); Operate machine and read in the pin heights. Update the display with the new heights. This function moves the board to the bottom, then puts up a prompt to allow the user to adjust the heights of the pins. Clicking off the prompt calls 'receivePhysical2.' o void receivePhysical2(); This function moves the board up and polls the pins as it moves, recording the height at which a pin first engages the board. main.c++ GUI and main. o PinViewer *pinsArea; Main data structure holding pin heights. o Widget myBB; Main GUI widget. o int machine_busy_flag = 0; o void machine_lock(); { machine_busy_flag = 1; } o void machine_unlock(); { machine_busy_flag = 0; } o bool is_machine_busy(); { return machine_busy_flag; } This variable and these functions prevent multiple requests from going to the machine at once. When the user begins an operation, the operation locks the machine and if the user should try to initiate another operation before his first terminates (and releases its lock), the second operation will quit upon seeing that the machine is locked. o void motorCB(Widget w, XtPointer client_data, void *call_data); o void acpinCB(Widget w, XtPointer client_data, void *call_data); o void readpCB(Widget w, XtPointer client_data, void *call_data); o void solupCB(Widget w, XtPointer client_data, void *call_data); o void soldnCB(Widget w, XtPointer client_data, void *call_data); When the user presses the 'Debugging ops' buttons on the right side of the GUI, he calls these functions. They respectively move the motor, activate a pin, read the status of a pin, activate the upper solenoids, and activate the lower solenoids. o void sendCB(Widget w, XtPointer client_data, void *call_data); o void receiveCB(Widget w, XtPointer client_data, void *call_data);

PAGE 150

138 These callbacks call sendPhysical and receivePhysical respectively. Pressing the 'Send' or 'Receive' button calls them. o float getMotorInputField(); o int getIntFromInputField(Widget text_field); o int getRowInputField(); o int getColInputField(); o float getHeightInputField(); o void setRowInput(int row); o void setColInput(int col); o void setHeightInput(float pHeight); These functions access/modify the GUI text fields. o void setPinHeight(Widget w, XtPointer client_data, void *call_data); o void getPinHeight(Widget w, XtPointer client_data, void *call_data); Called by clicking 'set/get height' buttons. o Widget makeLabel(Widget parent, char *s); o Widget makeButton(Widget parent, char *s, XtCallbackProc cb); o void setOnBoardPos(Widget w, int x, int y); o void setOnBoardPos(SoXtComponent *c, int x, int y); o void setWidgetSize(Widget w, int x, int y); Convenience functions. o void main(int argc, char **argv); Main function. Initializes GUI, jumps into GUI event loop. MakeItReal.c++ Writes pin heights to the physical system. -------------o int pin_burn_time_usec = 2500000; Time (in microseconds) for which to activate each SMA. o void wait_until_height(height_t desiredHeight); // Move sensor board down to 'desiredHeight' o void activate_pin(const Pin &p); // interfaces to digital I/O in order to activate pin. // Fire SMAs for time in pin_burn_time_usec. o void sendPhysical(); // Called from main. // Move the sensor board and activate the pins to lock them in place. Move board to the top; then popup a dialog, allowing user to unlock pins. Continuing through dialog calls sendPhysical2.

PAGE 151

139 o void sendPhysical2(); Lowers board, pausing to lock each pin at appropriate height. mapping.c++ Reads file 'pin_map.table' maps logical addresses to I/O card. This is the lowest level interface with the digital I/O cards. Must call read_mapping() before using any of the functions. read_mapping() insures it will only be called once. read_mapping() calls init_cards() and initializes the instance 'board' of the class 'SensorBoard.' PinViewer.c++ Draws 3D representation of pin array. SensorBoard.c++ Controls movement of the sensor board up and down. StepperMotor.c++ Used inside SensorBoard.c++ to activate the clicks of the motor. waitDialog.c++ Creates the pause that occurs between executing sendPhysical() and sendPhysical2(), as well as receivePhysical() and receivePhysical2().

PAGE 152

140 LIST OF REFERENCES Brain, M. (1992) Motif Programming: The Esse ntials and More. Digital, Burlington, Massachusetts. Gallagher, J. Refreshable Braille Displays. A-Z to Deafblindness. < http://www.deafblind.com/display.html > (2002, June 13). Gilbertson, R. G. (1994) Muscle Wires Project Book. Mondotronics, San Anselmo, California. Histand, M. B., and D. G. Alciatore. (1999) Introduction to Mechatronics and Measurement Systems. WCB/McGraw-Hill, Boston, Massachusetts. Huettl, B., and C. E. Willey. (2000) Design and development of miniature mechanisms for small spacecraft. Proceedings of the 14th AIAA/USU Conference on Small Satellites Logan, Utah, 14:1-13. Iwata, H., H. Yano, F. Nakaizumi, and R. Kawamura. (2001) Project FEELEX: adding haptic surface to graphics. Proceedings of the SIGGRAPH2001 Los Angeles, California, 18:1-7. Kelliher, A., and G. Pangaro. (2001) Le viTABLE: Electromagnetic Topographical Display. MIT Media Library. < http://web.media.mit.edu/~anjchang/ti01/ levitable.pdf > (2002, June 10). Mayer, J. (2001) Three dimensional surface di splay device. Master of Science Thesis. University of Florida, Gainesville, Florida. Mott, R. L. (1999) Machine Elements In Mechanical Design, Th ird Edition. Prentice Hall, Upper Saddle River, New Jersey. Pawluk, D. T. V., C. P. van Buskirk, J. H. Killebrew, S. S. Hsiao, and K. O. Johnson. (1998) Control and pattern specification for a high density tactile display. Proceedings of the ASME Dynamic Systems and Control Division New York, New York, 64:1-7. Wagner, C. R., S.J. Lederman, and R. D. Howe. (2002) A tactile shape display using RC servomotors. Proceedings of the Tenth Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems, Orlando, Florida, 10:1-2.

PAGE 153

141 Wernecke, J. (1995) The Invetor Mentor: Programming Object-Ori ented 3D Graphics with Open Inventor, Release 2. Addi son-Wesley, Reading, Massachusetts.

PAGE 154

142 BIOGRAPHICAL SKETCH David Lewis was born in Hollywood, Florid a, on Valentine’s Day of 1979. He followed in his parents’ footst eps by attending the University of Florida and followed in his grandfather’s footsteps by receiving a B achelor of Science de gree in Mechanical Engineering. He decided to stay at UF for his Master of Science degree.


Permanent Link: http://ufdc.ufl.edu/UFE0007064/00001

Material Information

Title: Three-Dimensional Visualization Display Device
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: UFE0007064:00001

Permanent Link: http://ufdc.ufl.edu/UFE0007064/00001

Material Information

Title: Three-Dimensional Visualization Display Device
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: UFE0007064:00001


This item has the following downloads:


Full Text












THREE-DIMENSIONAL VISUALIZATION DISPLAY DEVICE


By

DAVID LEWIS

















A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2004

































Copyright 2004

by

David Lewis















ACKNOWLEDGMENTS

I would like to thank Dr. Carl Crane for the opportunity to work at CIMAR, DOE

for their sponsorship, Shannon Ridgeway and Jose Santiago for their advice, Kayla

Spellman and Matt Borondy for their superb craftsmanship, my parents for their

encouragement, Lora Gay for her moral and technical support, and most especially Eric

Spellman for always having the time to help with any problem I had.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iii

LIST OF FIGURES ............................. .. .......... .................................. vii

ABSTRACT .................................................. .................... xii

CHAPTER

1 IN TR O D U CTIO N ..................................... ....................... ... .1

M ayer's 3-D D display D evice ..................................... ........................ ...............2......
S im ilar P rojects...................................................................................................... .. 4

2 PROJECT APPROACH AND OUTLINE .............................................................11

O b j e c tiv e ..................................................................................................................... 1 1
G o a ls ........................................................................................................ . .............. 1 1
E v solution .............. ........................................................................ . 11
R e ev alu atio n ........................................................................................................ 12
R resulting M machine ...................................................... ....... ...... ........ ........ ..... 13
System Architecture and Overview of 3 Systems.........................................13
H ow It W o rk s ...................................................................................................... 16

3 PH Y SIC A L D ISPL A Y ....................................................................... ............... 19

S k eletal D e sig n ........................................................................................................... 19
Pin Motion Inhibitors...................................................................... 20
P possible A ctuators ......... .. ..................... ................................................ 20
M miniature rotational solenoid .................................................. ................ 20
M agnetorheological dam opening .............................................. .................. 21
Shape m em ory alloy tubes ...................................................... ................ 2 1
Shape m em ory alloy w ires ...................................................... ................ 23
C lam p C reaction .................................................................................................... 2 4
H om e design ......................................................................................... 24
P encil clam ps ................................................................... ............... 26
E v solution of D esign ........................................................................ ...............2 8
B ioM etal M icro H elixes ................. ............................................................... 31
H o le P pattern ................................................................................................................ 3 5









P in Sensors .................................................................................. ....................... 35
H all-E effect Sensors .... .. ... .................. .. ....................... ....................... 35
C contact T races and Steel W ool ........................................................ ................ 36
Trace Pattern and B oard C reation............................................................ ................ 37
Pin M ovem ent System and Fram e........................................................... ................ 42
M otor Selection ............................ .... .......... ......... ............... 42
Solenoid Selection .. ................................................................................. 44
S h a ft S e le ctio n .....................................................................................................4 6

4 CON TROL BOX .......................................................................................... . 60

R aster Scanning .............. ......................... ........................ ............... .. .. .... 60
M apping ............................................................................. 62

5 C O M PU TER ................................................................................. ...................... 65

O overview ............................................................................................. 65
G raphical U ser Interface ...........................................................................................66
P ath P planner ................................................................................. ...................... 67
D digital I/O ......................................... ............. .................. 69

6 A S S E M B L Y ................................................................................................................ 7 0

C control B ox ............................................................................................................. 70
Pre-C lam p Physical D display .....................................................................................71
F ra m e ...................................................................................................................7 1
M otor/Shaft System s .........................................................................................74
B rackets .............. ...................................................................... ......75
Plates ............................................. ................................. 75
Solenoids .............................................................................................................75
Pin-C lam p System s ...................................................................................................75
M echanical Pencil Disassembling .....................................................................76
B rass Cup and W ire W rap.................................................................................77
D io d e C u rlin g ......................................................................................................7 8
G uide Tubes ........................................................................... ......................... 79
Steel W ool A ttachm ent ......................................................................................79
SM A System A ssem bly .....................................................................................80
Final A ttachm ent Process .................................................................................. 81
E lim nation of the F an ..............................................................................................82

7 TESTIN G AN D RESULTS .......................................................................................84

T ests ............................................................................. ..................... 84
Clamp System Tests ...................................................................84
F irst SM A test ............................................................................................84
M icro helix tests .........................................................................................85
Electrical Tests ............................................................................................. . 87


v









Raster Scan Test ............................ ............................87
Sensor T ests ................................................................................. ...................... 88
H all E effect .................................................................................................... 88
C contact Sensors .................................................... ....................... 89
Testing the Steppers, Drivers, and Digital I/O .................................................90
F final R results .................................................................................. ...................... 9 1
Physical V isualization ......................................................................................92
Q uick Fixes .................................................................................................... 93
V irtual V isualization ........................................................................................94

8 FU TU RE W O RK ................................................................................................... 98

G general C changes ................................................................................................... 98
M major Changes ............................................................................................. . 99
C lam ping .............................................................................................................99
Sensing .............. ..................................................................... ......100
C o n c lu sio n s...............................................................................................................1 0 2
Opinions and Personal A dvice............................................................................103

APPENDIX

A DEVELOPMENT OF THE FIRST PIN CLAMP...............................................104

B THE RASTER SCAN TEST REPORT..................................114

C O D D S A N D E N D S ....................................................................................................12 1

D SPECIFICS OF W IRIN G ....................................................................................132

E P R O G R A M C O D E ....................................................................................................136

LIST OF REFEREN CE S ...........................................................................................140

BIOGRAPH ICAL SKETCH .....................................................................................141
















LIST OF FIGURES

Figure page

1-1 Jeremy Mayer's 3-D display device (translating pins not shown)........................3...

1-2 Mayer's 3-D display device had eight physical pins and eight virtual pins that
could m irror one another.......................................... ........................ ...............4...

1-3 The LeviTABLE ...................... ............ .............................5

1-4 Tactile shape display using RC servom otors...................................... ...............6...

1-5 Servom otor positioning ........................................... ......................... ...............7...

1-6 T h e F E E L E X 1 .................................................. .............................................. 8

1-7 T h e F E E L E X 2 .................................................. ..............................................9

1-8 H igh D ensity T actile D display .............................................................. ...............9...

1-9 R efreshable B raille D display ..................................... ...................... ................ 10

2-1 The 3-D visualization device ........................................................ 14

2-2 System architecture ... .................................................................. ... ......... ... 14

2-3 C control B ox ............................................................................ ........ ...................... 15

2-4 Physical display ............................... ......... ....................... 16

2-5 G raphical U ser Interface ........................................ ........................ ............... 17

3-1 T he initial skeletal design ....................................... ....................... ................ 20

3-2 A m miniature rotational solenoid ........................................................ ................ 21

3-3 The SMA tubing is bent and then returns to its original shape...........................22

3-4 Attempts at cutting SMA tubing resulted in irreversible damage.......................23

3-5 The clam ping process ................................................................... ................ 24









3-6 Initial clam p shape .............................. ............................................ 25

3-7 FEA analysis for m material choice...................................................... ................ 25

3-8 M manufacturing of a clam p.................................... ....................... ................ 26

3-9 Example clamps from mechanical pencils ................ ....................................27

3-10 T w o clam p sy stem s. ....................................................................... ................ 27

3-11 Clam p actuation testing m odel.................. .................................................... 28

3-12 T he fi rst hole pattern ...................................................................... ................ 29

3-13 Close up of electrical connection...................................................... ................ 29

3-14 P late arrangem ent ... .. ........................................... ........................ ... .......... 30

3-15 B ioM etal m icro helixes.......................................... ........................ .................. 3 1

3-16 H elix m odel 2 ................................................................................................... 32

3-17 M multiple pin clam ping m odel............................................................ ................ 33

3-18 Final pin-clam p system design.......................................................... ................ 34

3-19 H all-effect sensors ............... ................ .............................................. 35

3-20 Steel wool attached to the bottom of a pin ................ .................................... 37

3-21 D igital bottom plate ............. ................ ............................................... 38

3-22 D igital ground plate ............. ................ ............................................. 39

3-23 D igital pow er plate .. ..................................................................... ............. 40

3-24 D igital support plate. ................................................................... ................ 41

3-25 Digital top plate........................... ........... .........................42

3-26 Thompson 55M048D1B Bipolar Stepper Motor..............................................43

3-27 Free body diagram of forces on the screw ................ ....................................43

3-28 Magnet-Schultz DC Voltage Cylindrical Push Solenoid..................................45

3-29 Shaft / m otor system ..................................................................... ................ 46

3-30 Two-dimensional drawing of the shaft/ motor system alignment ......................48









3-3 1 Skeletal structure of 80/20 ....................................... ...................... ................ 49

3-32 Assembled 3-D display frame with fan ............... ....................................50

3 -3 3 M otor m ou nt .......................................................................................................... 5 1

3-34 U pper bearing m ount .................................................................... ................ 52

3-35 Push solenoid m ount ...................................................................... ............... 53

3-36 Prelim inary assem bly draw ing.......................................................... ............... 54

3-37 Top plate bracket........................... .......... ........................ 55

3-38 G round plate bracket ...................................................................... ................ 56

3-39 Pow er plate bracket ............... ................ .................................................. 57

3-40 Sensor plate bracket ................................................................... 58

3-41 The fully assem bled physical display ............................................... ................ 59

4-1 An example 3x3 grid of actuators in a raster scanning sequence .......................61

4-2 An example 3 x3 grid of sensors in a raster scanning sequence..........................62

4-3 Direction of signals through the control box .................................... ................ 63

4-4 B asic m apping of the control box ..................................................... ................ 64

5-1 Computer setup ......................... ............ ........................... 65

5-2 Graphical user interface (a) Example GUI display, (b) GUI reading a file...........66

5-3 P C I 7296 digital I/O card ........................................ ....................... ................ 69

6-1 The sections of the frame where the red circle encompasses a waste section,
the blue circle encompasses an arm section, and the green circle encompasses
a leg se ctio n ............................................................................................................ 7 2

6-2 W aist section of the fram e ...................................... ....................... ................ 72

6-3 L egs section of the fram e ........................................ ....................... ................ 73

6-4 A rm s section of the fram e.................................... ....................... ................ 73

6-5 B ins of separated m materials ...................................... ...................... ................ 77

6-6 A diode before assembly and a diode after assembly ......................................78









6-7 On the top-left is the heat-shrink tubing, towards the center is the clear tubing,
and on the right is the assembled guide tube .................................... ................ 79

6-8 P hy sical display w ith fan ........................................ ....................... ................ 83

7-1 C lam p actuation testing m odel.......................................................... ................ 85

7-2 Micro helix tests (a) Helix model 1, (b) Helix model 2 .........................................86

7-3 T he raster scan test setup .................................................................. ................ 87

7-4 Sensor test setup .. ............... ......... .... ... ............... 89

7-5 Stepper m otor w ired to a driver ........................................................ ................ 90

7-6 P ath planner test setup .......................................... ......................... ................ 9 1

7-7 Results of a physical visualization test where (a) is desired product that was
formed on the GUI and (b) is the final product produced by the physical
d isp la y ................................................................................................................ .. 9 4

7-8 The four possible arrangements for incorrect sensor readings ..............................96

7-9 The results of a virtual visualization test where only the front row was tested;
(a) shows the initial arrangement of the physical pins and (b) shows the virtual
r e su lts ................................................................................................................. .... 9 7

8-1 Pulley system of clamp actuation where the left block moves with the pins'
rows and the right block moves with the pins' columns.................................100

8-2 A two-level sensor reading system in a raster scan arrangement where the red
ovals represent sensor holes, the blue squares represent the single-level first
read points, and the green dots represent the multi-level set of second read
p o in ts ................................................................................................................ ... 1 0 2

A -i Initial clam p design ................................................................................. 105

A-2 FEA analysis for material choice.............. ..................... 107

A-3 Node placement on beam analysis for Nub 1 placement................................108

A-4. Node placement on beam analysis for Nub 2 placement................................110

A -5 C om pleted clam ps... ................................................................... .............. 112

A -6 Evolution of clam p creation...... ........... ........... ..................... 112

A-7 Manufactured clamps (a) Completed clamps, (b) Clamp changes through
p ro d u ctio n ............................................................................................................ 1 13









B -i Configuration of LED 's ................. .......................................................... 115

B-2 The Wiring Diagram For the LED Display ..............................................116

B-3 The Assembled Electrical System .............. ......................117

B -4 Five L E D s B being L it........................................ ......................... ............... 119

B -5 N ine LED s B being Lit ......................................................... 119

C-i The three hole patterns that were analyzed...... .... ................................... 122

C -2 Table A rrangem ent .................... .............................................................. 123

C-3 Close-up of electrical connection...... .... ...... ...................... 124

C-4 D design 1 concept draw ing................................... ...................... ................ 125

C -5 D esign 1 detail draw ing ........................................ ....................... ............... 125

C-6 D design 2 concept draw ing................................... ...................... ............... 126

C-7 D esign 2 detail draw ing ................. .......................................................... 126

C-8 SMA spring actuation with inelastic band...... ........................................128

C-9 Ideal spring SMA actuation design...........................................129

C-10 Compressing SMA spring actuation design...... .......................................130

C-11 Miniature compressing SMA spring actuation design...................................131
















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

THREE-DIMENSIONAL VISUALIZATION DISPLAY DEVICE

By

David Lewis

August 2004

Chair: Carl D. Crane, III
Major Department: Mechanical and Aerospace Engineering

This research focuses on the design of a device used for presenting three-

dimensional images. The machine was built with the purpose of performing two

particular tasks. The first task was to be able to adjust an image on a computer screen

and have a physical display mirror that image. The second task was to be able to adjust a

physical display and have the computer screen mirror the image. Such a device would be

useful for rapid prototyping, virtual modeling, topography, measurement, and touch-

screens.

The device was built with the hope of escaping the limitations typically involved

with virtually creating an object and the time typically involved in physically creating an

object. This device utilizes three main components: a computer, a control box, and a

physical display. The physical display uses a grid of parallel pins that can be moved in a

common direction to create a "moving wall." Some of the key issues in designing such

apparatus include minimizing the power consumption, packing the pins as tightly as

possible, overall use of space, and the accuracy of images.














CHAPTER 1
INTRODUCTION

In this age of information, society has placed great value upon methods for

transferring large amounts of information as quickly and tightly packed as possible. One

of the best ways to instantly transmit a lot of information is to deliver an image. A three-

dimensional visualization is an ideal way of presenting every aspect of an object in a

single, complete image. One of the most popular methods of 3-D visualization is through

computer graphics displays. Many of these machines can lead to problems where the user

is required to wear bulky headgear and only one user can view the image at a time. In

2001, Jeremy Mayer attempted to solve this problem by creating a machine that

completed the following three objectives (Mayer 2001):

1. Presents a three-dimensional representation of the object being displayed

2. Does not require the user to wear any special glasses or head mounted tracking
system

3. Allows multiple users to simultaneously view the same object. Such a device will
greatly aid an operator (or team of operators) in tele-operation and tele-supervision
operations.

Mayer's project aimed to create three-dimensional images out of a "moving wall"

where the "moving wall" was composed of thin parallel pins that could be moved up and

down individually. He successfully created a prototype display device that moved eight

parallel pins up and down. It was the desire of this project to create a machine that would

improve the capabilities of the prototype design while being more efficient with its use of

size, power, and cost.









Such a machine could be of use in the following areas:

* Rapid Prototyping: An instant three-dimensional model could be made for an
object designed on a CAD program. This would allow a designer to quickly
provide others with a model of a product that they can view and walk around.

* Measurement: An object could be pressed into the pins and its dimensions would
be recorded in the process of making a virtual image. These dimensions would be
easily accessible for the user.

* Virtual Modeling: Virtual models could be made of real life objects. Awkwardly
shaped containers and objects could be virtually rendered after being pressed
against the pins. Then a computer could be used to solve premium packing
situations.

* Topography: The display could be used to show large areas of land. Such a device
could be useful to anyone from engineers deciding where to place a bridge to
military officers deciding where to position their men. To get such a tool currently,
these people need to have a model sculpted from hand. This device would pair
well with a new technology that allows an airplane to reflect radar over a section of
land and a computer program uses the reading to render a CAD image of the land.
This image would then be easily transferred to the physical display.

* Picture: This would be a new way for friends and families to send one another
images. A grandmother could see the three-dimensional hand print of her newborn
grandchild that is hundreds of miles away.

* Touch-Screen: The display could be adjusted into a grid of buttons and when the
user presses a button, the device would be able to detect the adjustment. An image
could be shown down upon the display to help aid in the visual, as well. This
would allow for various interactions between user and machine that would be
useful in such devices as ATMs or Information Kiosks.

Mayer's 3-D Display Device

Jeremy Mayer's machine, seen in Figure 1-1, moved eight pins that were aligned in

a row. Depending upon the user's desire, the pins could be moved manually and a

computer image could imitate the pattern, or the computer image could be altered and the

pins would imitate the image's pattern. This is represented in Figure 1-2. The device's

actions were controlled by a graphical user interface (GUI) where the user could adjust

the virtual image of pins and tell physical display which task to run. The physical pins










rested upon push-pull cables, which were allowed and prevented from moving by

solenoid-controlled swivels. The push-pull cables were composed of a metal wire held

within a plastic sleeve by friction. The top end of each sleeve was attached to a stable

plate, which the physical pins rose out of. The bottom of each sleeve was attached to a

plate, which could be moved vertically, where the metal wire hung out loosely. To create

a physical pattern of pins, the pins began at their bottom position while all the swivels

were activated and pinched the metal wires into a stable position. The moving table

could then be lowered by a stepper motor, causing the sleeve to lower, but not the metal

wire; in turn, the pins would be raised. When a pin reaches its desired height, the

corresponding solenoid is deactivated. This rotates the swivel and the pin is held in place

by the friction of the push-pull cable.




1-1.q Upper Plote





son lenoids


Figure 1-1. Jeremy Mayer's 3-D display device (translating pins not shown)


















Figure 1-2. Mayer's 3-D display device had eight physical pins and eight virtual pins that
could mirror one another

To create a virtual pattern of pins to mirror a physical arrangement, the pins are

moved to their highest positions and the swivels are opened. The user can physically

adjust the pins and the friction of the push-pull cables will hold them in place. Each

hanging metal wire is then gently pressed against an individual switch by the spring-

loaded solenoid. The moving plate is translated upward by the stepper motor and the

hanging metal wires are pulled up. Eventually, each hanging metal wire will be pulled

above the swivel and the spring-loaded solenoid will push the swivel into the switch. The

switch will be activated and the relative location of moving plate will tell the controlling

computer where the pin was located. The GUI will do the conversions and render a

mirror image of pins.

Similar Projects

Research was invested into finding similar projects and inventions in existence.

Understanding the designs, successes, and failures from these works could help set an

efficient path for the next generation display and prevent previous errors from being

repeated. The following are some of the most similar and practical related machines that

were found:









The LeviTABLE, seen in Figure 1-3, is being worked on at MIT (Kelliher et al.

2001). This apparatus is an "electromagnetic topographical display" which basically

means that it uses electromagnets to create a physical, three-dimensional feature. The

device contains a 6.5"x6.5" actuation platform with an array of electromagnets placed in

an 8x8 formation. Above the actuation platform is a table system containing 64 drinking

straws, also arranged in an 8x8 pattern. The straws, eight inches from top to bottom,

move freely in the vertical direction and contain neodymium magnets within them.

Varying the strength and polarity of each of the electromagnets in the actuation platform

allows one to conform the grid of straws into desirable topographical displays. The

display can be changed as quickly as the desired magnetic states can be relayed to the

electromagnets and the footprint remains equivalent to the space taken by the grid of

straws. However, the straws are limited to 1.5" of movement. The project's most blatant

flaw is that although a theoretical movement tolerance of 1.0 mm is claimed, the creators

have not yet been able to implement any more than a movement of the straws from one

extreme position to the other. Furthermore, the machine requires a continuous supply of

energy for all times of use, and the thickness of the straws along with the spacing

distances do not allow for a very "tight" image.


Figure 1-3. The LeviTABLE









A Harvard project resulted in the creation of a tactile shape display using RC

servomotors, seen in Figure 1-4, with the purpose of "conveying small-scale force and

shape information to the tip of the finger" (Wagner et al. 2002). The device is a 6x6 pin

display, with 2mm maximum displacement and is an "attempt to realistically simulate

skin deformations that occur when interacting with real objects by transmitting small-

scale shape information to the fingertip." RC servomotors (small, high-performance ball

bearing servos) actuate the mechanical pins as seen in Figure 1-5. Positive features

include a low parts cost, simple construction, and the ability to move in time to a specific

height. However, the machine also has limited movement and scales in size poorly when

an increased number of pins are desired. Also, servos rotate so the pin is being moved in

two directions, but since the movement is so small in comparison to the pin size the

developers might be thinking the pin crookedness is negligible; however, this will be

detrimental if the pin movement is made greater.


Figure 1-4. Tactile shape display using RC servomotors














I11.9 cm







Figure 1-5. Servomotor positioning

The University of Tsukuba's FEELEX Project aims to combine haptic sensation

with computer graphics (Iwata et al. 2001). Basically, the shape and image of an object

will change as a result of outside forces on the object. The basic system is comprised of

an array of actuators beneath a flexible screen. A projector (720x420) shines a two-

dimensional image down onto the screen, which in turn becomes three-dimensional when

the actuators deform the screen. Sensors on the actuators can determine the force applied

against the screen by an outside source. The readings can be used to determine a

response in the shape of screen. Two prototypes for this system structure were created.

FEELEX 1, shown in Figure 1-6, was "designed to enable double-handed

interaction using the whole of the palms." The flexible screen (composed of a 3 mm

thick rubber plate and a white nylon cloth) is 24cm x 24cm in size with a 6x6 linear

actuator array beneath it. Each actuator is made of a screw mechanism driven by a DC

motor. The machine is PC controlled with two strain gauges atop each actuator acting as

force sensors. The apparatus moves in time, is structurally durable, and has a linear ratio

between grid size and footprint size. Unfortunately, the machine has a high power









requirement, poor resolution of shape, and there are shape limitations, as sharp edges

cannot be created.














Figure 1-6. The FEELEX 1

FEELEX 2, shown in Figure 1-7, was created with the intention of improving the

resolution of the previous design so that interaction can occur with the fingers. This

system has a 50mm x 50mm screen with a desired resolution of 8mm. Because the

servomotors used are much larger than the desired resolution a piston-crank system was

used to move an array of rods (six inches in diameter) that change the shape of the screen.

Force is measured by reading the electric current going into each servomotor. This

machines best feature is its tight resolution. Its drawbacks are that the piston-crank

system creates size of footprint difficulties for pin expansion, it requires a high-energy

intake, and (like the FEELEX 1) there are shape limitations, as sharp edges cannot be

created.

The "high density tactile display," created at John Hopkins University and

presented in Figure 1-8, is intended for use in neurophysiological and psychophysical

experimentation (Pawluk et al. 1998). It contains a 20 x 20 square array of pins. There is

0.5 mm between each pin. The pins are moved by linear motors stacked in layers of 100




























Figure 1-7. The FEELEX 2

motors above the array. Each 100 motor layer controls a 10 x 10 section of the pins. The

pins can be moved 2.5 mm lengthwise. The machine has the ability to move the pins in

time to any desired position as well as maintain a desired pressure from the pins. This

design makes it spatially difficult to scale to a larger number of pins (as shown in Figure

1-8), is limiting on the available length of pin movement, requires large supply of power,

and is considerably expensive.

















Figure 1-8. High Density Tactile Display









Refreshable Braille Displays, pictured in Figure 1-9, were the last similar device

that was discovered during the literature search (Gallagher). These devices can be

purchased commercially from various vendors. These machines are commonly used by

the blind for the purpose of reading text that would be displayed by a computer monitor.

The Refreshable Braille Display has a row of "soft cells." Each cell has 8 holes filled

with linearly actuated pins. The pins can be moved up or down (using either miniature

solenoids or piezo-electric outputs) so that each cell contains a Braille character. The

user can scroll through lines of text changing the character displays instantly. The pin

movement is limited as they can only be located at one of two positions (up or down).

The machines are also expensive, as an average, commercially sold displays run in the

tens of thousands of dollars.


Figure 1-9. Refreshable Braille Display














CHAPTER 2
PROJECT APPROACH AND OUTLINE

The knowledge of other work performed on similar three-dimensional displays

helped create guidelines to steer this project. It was important that past errors not be

repeated and as much improvement be made as possible. This chapter reviews the

project's initial aims to its final product.

Objective

This project began with and was always governed by the three main goals that

were created and used by Mayer in his project:

1. Presents a three-dimensional representation of the object being displayed

2. Does not require the user to wear any special glasses or head mounted tracking
system

3. Allows multiple users to simultaneously view the same object. Such a device will
greatly aid an operator (or team of operators) in tele-operation and tele-supervision
operations.

Goals

Evolution

For the next generation display, a new set of subsidiary goals had to be created

that would make the display more effective and more efficient while still maintaining the

original three objectives as the top priorities. A set of secondary objectives inspired by

the experiences and achievements of the previous machine were created. These goals

were not set in stone, but merely guidelines meant to help provide direction and an

ambitious aim for the project. What follows are the original secondary goals listed in

order of importance:









* Pin Pattern: The most vital modification to the new machine is to go from a linear
pattern of pins to a 2-D array of pins

* Actions: The apparatus will be capable of displaying the shape of the pins upon the
computer screen as well as displaying a computer image physically with the pins

* Pin Grid Size vs. Total Machine Footprint: The device will be designed so that
adding additional pins will increase the pin grid size the same amount that it will
increase the overall machine footprint.

* Number of Pins: The minimum number of pins will be twenty-five; however, the
pin count will ideally be in the thousands

* Spacing of Pins: The pins have a diameter of 0.045". They will be kept within 2
mm (center to center) of one another.

* Power Allotment: No more than 1.5 Watts will be allotted to the actuation of each
pin.

* Height: The top of the machine will be less than 3 feet high so that a person of
average height could look down upon the display.

* Footprint: The machine will be within a footprint of 2' x2' so that it can be easily
moved through average size doorways.

Reevaluation

Next it was important to evaluate the existing 8-pin machine. An analysis was

performed to decide which of the current features should be altered in order to achieve

the desired goals for the next machine and alleviate the difficulties from the first one.

The following is a list and description of those decisions:

* Actuation System: Renovating this feature is the most vital action of the project.
While the moving table system worked well in the 8-pin display, the push-pull
cables caused a number of problems. It wasn't always reliable as changes in
friction occasionally occurred with movement, resulting in display inaccuracies.
The rotational solenoids (which supplied far more pressure then was necessary)
used for clamping were too large and required too much power to be practical for a
machine with an expanded grid of pins. These actuators will have to be replaced
with something smaller, cheaper, and less energy draining.

* Power Supply and Control: Having hundreds of pins where each pin needed to be
actuated individually would require a more conservative system that requires less
power per pin. A new wiring system will also have to be implemented where the









outputs from the controllers will be activating multiple actuators, so as to minimize
the number of necessary outputs.

* Sensors: The current sensors are contact sensors that provide acceptable values;
however, using something more precise, such as optical sensors, could provide
more exact readings of the pins and reduce measurement errors.

* Interface: The interface can be improved with the addition of location axis and data
tables giving the exact location of each pin. The tables could be used for both
reading where the pin is located and writing where the pin's next desired location
is. It would be convenient to give the display an option so that the heads of the pins
are seen as a single skin and a section of a CAD drawing could be transposed onto
the physical pin display.

Resulting Machine

System Architecture and Overview of 3 Systems

Figure 2-1 shows a picture of the resulting machine. It is composed of three main

systems, which can be broken down into the system architecture that is shown in Figure

2-2. The three main components of the setup are the computer, the control box, and the

display device. The Computer contains the Graphical User Interface (GUI), the path

planner, and the digital input/outputs. The GUI is a display that will appear on the

computer monitor and allow the user to interact with the rest of the machine. The data

distributor is a program code that translates data instructions between the GUI and the

digital I/O. The digital I/O is a PCI card that sends and receives multiple electric signals

from the computer to the control box.












*-iK P. I.CAL
B te 6 f-s! a _
^^^HHfl'iii"ii:::.\.**


Figure 2-1. The 3-D visualization device


USER


COMPUTER


CONTROL
BOX -




PIN
DISPLAY
DEVICE





USER


Figure 2-2. System architecture









The control box, shown in Figure 2-3, is made up of opto-isolators, power sources

and motor controllers, as well as the numerous wires, cables and terminal boards tying

everything together. The motor controllers will regulate the speed and position of the

stepper motors. The opto-isolators will regulate when specific sections of the display

device receive power.

















Figure 2-3. Control Box

The display device, pictured in Figure 2-4, uses parallel pins to form a three-

dimensional image that is visible to the user. The pins are moved and monitored using

hard plastic plates, clamps, and sensors. The plates are moved by solenoids and stepper

motors. The clamps are toggled by shape memory alloy (SMA) wires. The following

three chapters will go into the details of the research and development for each of these

sections, as well as how they work.

























Figure 2-4. Physical display

How It Works

The machine performs two main tasks. One is that it can be provided a virtual

arrangement of pins through the computer and then arrange the pins symmetrically on the

physical display. The other task is that the pins on the physical display can be arranged

into a particular pattern and the computer will render a mirrored pattern of pins. The first

task of physical visualization is performed in the following manner.

Employing the GUI shown in Figure 2-5, the user will arrange the pins to their

desired locations. This can be done by clicking on the rendered pins with the mouse and

dragging them, by typing in the exact height of each pin, or by loading a pre-written file

with all the pin heights in it. Once happy with the arrangement, the user will click on the

"Send" button of the GUI. At this point the pins will be raised to their top positions by the

sensor table and the solenoids will open and release clamps in order to loosen their grips on

the pins. Next a new window will appear on the GUI asking if the clamps need to be

opened again. If any clamps are still preventing movement of any pins, then the user will

continue to click the "yes" button until all the pins are loose. Once all the pins are loose,







17


the user clicks the "no" button. Next, the table that the pins are resting on will lower all the

pins to the location of the height of the highest pin. At this location, the table will pause

while actuators are activated to hold all of the pins that have that desired height into place.

Once activated, the table will lower the pins to the next highest pin location. The pins with

activated actuators will be held at their previous locations and new actuators will be

activated to hold new pins at the new location. This pattern of lowering, pausing, and

activating actuators will be repeated until all the pins have been moved to their desired

positions and the table is at its lowest location. The user can then leave the pins at the

current configuration or perform another task.



Debugging Operations:

Move Motor: [


ReadPin:I --

Top SolenoidsI
Bottom Solenoidsj





RotX Roty Tx Dolly
Set desired Pin heights.
Row Column
Set Height

Send Receive Read We | ui


Figure 2-5. Graphical User Interface

The second task of virtual visualization is performed in the following manner:

The user will click the "Receive" button on the GUI. The sensor table will move

the pins to their highest possible position, the clamps will be toggled into a locking position

by the solenoids, and the table will move back to its lowest possible position. The pins will









remain held at their highest position and a window will appear on the GUI asking the user

to adjust the pins to their desired positions. At this point, the user will physically move the

pins into the arrangement he or she wants displayed. Next, the user will click the "OK"

button on the new window. This will trigger the sensor board to start moving upward

again. The board will move to its top position in a single slow and smooth motion, moving

each pin it comes in contact with upward with it. When the board stops, all the pins will be

back at their top positions and the GUI will show a rendered image of the user's

arrangement of the pins. The user then has the choice of leaving the machine in its current

configuration, recording the pin arrangement into a *.pins file, or performing another task.

Further details on how to use the GUI with detailed flowcharts of the task

processes are presented in Chapter 5.














CHAPTER 3
PHYSICAL DISPLAY

The Physical Display is the most complicated of the three main components found

within the System architecture. This system is a display device that accomplishes the

three main goals of the project. The display contains a grid of movable pins that can be

adjusted to create a desired image. The pins can reach their desired positions as a result

of two methods: the display will receive position commands from the computer and

adjust the pins itself or the user will manually adjust the pins so that the display can

detect the positions of the pins and an image of the pin positions can be rendered on the

computer screen. This chapter details the components of the physical display, how they

were designed, and how they work.

Skeletal Design

The design of the physical display essentially began after a thorough evaluation of

Mayer's project and related work. With a guiding set of goals in place, it was decided

that the display device would have the basic design that is shown in Figure 3-1. The pins

would pass through a device that would be capable of preventing their motion and rest

upon a movable plate. Both tasks would begin with the pins at their highest position. For

physical visualization, the plate would lower and the pins would follow until the motion

of each pin was prevented. For virtual visualization, the motion would be prevented, the

table would lower out of the way, the user would adjust the pins positions, and a separate

system would detect the positions of the pins. This meant that three systems had to be









designed: a system for preventing motion of the pins, a system for moving the table, and

a system for sensing the position of the pins.


PINS



PIN MOTION
INHIBITOR
SENSOR





MOVING TABLE


Figure 3-1. The initial skeletal design

Pin Motion Inhibitors

Possible Actuators

The entire design of the physical display revolved around the fact that some form

of actuator was going to be necessary to either push, pull, or pinch the movable pins in

the display. The type of actuator used would have a direct outcome on the display

device's pin-to-footprint-size ratio and its power drainage. The following methods of

actuation were initially analyzed and compared in order to find the best method of

actuation.

Miniature rotational solenoid

A miniature rotational solenoid, such as the one in Figure 3-2, could have been

used to pinch the pins in the same fashion as the actuators in Mayer's design except less

space would be necessary for the actuators and only 0.3 Watts would be used per actuator









as opposed to the 5 Watts per actuator used by his machine. The worries with this

actuation were that as the number of pins was increased, they might be too expensive and

still consume too much space. This was a tempting option because the solenoid greatly

decreased necessary power and was proven to work quite effectively with Mayer's

design.













Figure 3-2. A miniature rotational solenoid

Magnetorheological dampening

This method would have had the cables moving through a magnetorheological

fluid. When a magnetic field is applied to the fluid, the fluid hardens and grips the pin

and when the field is removed, the pin is released. This method could not only have been

messy, but it could have lead to corrosion and electrical problems. A system using this

method could not be thought up where either the movement of pins or the physical well

being of the machine would not be harmed.

Shape memory alloy tubes

This method would have had the pins passing through a shape memory alloy tube.

A shape memory alloy (SMA) is a metallic material that holds a specific shape but can be

stretched or bent into another shape. As seen in Figure 3-3, when the temperature of the

material is changed above or below a specific temperature (this temperature can be hot or









cold and depends on how the SMA was manufactured), the material returns to its original

shape. In this case, when the tube is heated with an electrical current, it shrinks and grips

the cables. When the current is stopped the tube returns to its original state, allowing free

movement of the cables. It uses little power (1.2Watts) and takes up little space. The

worries were that the SMAs may take a while to react to the current as well as to reset

themselves. There was also the possibility that heat from an SMA with current running

through it might affect the SMAs nearby.













Figure 3-3. The SMA tubing is bent and then returns to its original shape

Because of its economical cost and structural simplicity, the SMA tubing actuation

initially appeared to be the best method available; therefore, work was begun to design

and create model that would implement it. The design that was created used three stable

plates stacked on one another. The top and bottom plates would have grids of holes just

big enough for the pins to move through freely; however, the middle plate would have

thicker holes to hold miniature SMA tubes. A current could be passed from the top plate

to the bottom plate to heat the tubes. So basically, the pins would be going through the

triple-plate stack, able to move freely, but resting on a fourth plate. The fourth plate

would begin lowering and each time a pin reached its desired location, its SMA tube









would be heated, thus gripping it and stopping the pin's motion. This would continue

until all pins could no longer be lowered.

This plan had to quickly be abandoned, when an attempt to build a model proved

that the SMA tubes were very difficult to cut. Each attempt to cut the tubes reliably

resulted in chipping, cracking, and irreversible bending, as shown in Figure 3-4. Because

pre-cut tubing was too expensive, a new method had to be sought out.














Figure 3-4. Attempts at cutting SMA tubing resulted in irreversible damage

Shape memory alloy wires

While the SMA tubing was a disappointment, the best route still appeared to be to

use SMA material. Shape memory alloy wires have the ability to be stretched to a certain

length and then heating or cooling makes them return to their original length (Gilbertson

1994). An initial method of actuation was devised using this material. Figure 3-5

presents that method which would be the beginnings of the final method. Basically, the

pins would pass through a clamp and cylinder. When left alone, the clamp would prevent

any movement of the pin; however, when the cylinder, which allows the pin to pass

through freely, is pressed into the clamp, the clamp opens and the pin can move freely.

The cylinder could be removed from the clamp when an SMA, attached to the cylinder on

one end and a stable plate on the other, is heated; thus, contracting the SMA and pulling











on the cylinder. All the cylinders could be pushed back into their respective clamps when

a moving plate, with holes allowing the pins and SMAs to pass through, pushes them into

the clamps and re-stretches the SMAs.


SITUATION 1:
Clamp w/out Pin
= No Displacement


SITUATION 2: SITUATION 3:
Clamp w/ Pin Inserted Clamp w/ Pin and
= 1st Displacement Cylinder Inserted
+ Stable Pin = 2nd Displacement
+ Free Moving Pin
Pill





|/


Figure 3-5. The clamping process

Clamp Creation

Home design

With a basic structure for the actuation process established by the SMA wire

method, it was necessary to create a more detailed design of the clamp. It was initially

decided that the clamps would be designed from scratch and manufactured in-house.

Figure 3-6 shows the prototype shape that was created for the clamp.




























Figure 3-6. Initial clamp shape

A finite element analysis (FEA) was performed using the Ideas software to find

ideal clamp dimensions that would create the most effective and long-lasting clamp. This

analysis helped determine the clamps dimensions, shape, and material. Figure 3-7 shows

an example of one of the analyses.















Figure 3-7. FEA analysis for material choice

Next, a complicated manufacturing process was developed for actually creating

physical clamps and several prototypes were made, as seen in Figure 3-8. Appendix A









has the full detailed report on the shape design, the FEA analysis and the manufacturing

process.









Bent Clamp

Coated/Complete Cla rp -:

Figure 3-8. Manufacturing of a clamp

Pencil clamps

After refining the manufacturing process for clamp production, it was determined

that if clamps were being produced every single day at eight hours per day, it would still

take anywhere from five months to a year to produce the desired one thousand clamps.

Because that was a longer time investment than was desired, the next best option was to

use an out-of-house machinist who possessed more powerful tools. When no machinists

were willing to take the job, it was decided to explore further areas for clamp production.

It was observed that a majority of commercially sold mechanical pencils are composed of

clamps, presented in Figure 3-9, with similar size and shape.




















Figure 3-9. Example clamps from mechanical pencils
Pencil clamps that hold 1.3 mm lead were found to provide the results desired from

the original clamp; however, the pencil clamps use a system mechanically reversed to the
original design. The pencil clamps place the hollowed cylinder on the outside of the
clamp while the original design had the cylinder moving inside the clamp (as shown in
Figure 3-10). It was decided to proceed by replacing the original design with the pencil
clamp system.


First Clamp System
Clamped Free Movement
Pm.



Cylinder I


Pencil Clamp System
Free Movement Clamped


II
2K


Figure 3-10. Two clamp systems









Evolution of Design

Now that a clamp design was complete and a basic setup existed, tests were begun

in order to evolve the clamp system design. A one-pin model, shown in Figure 3-11, was

created to test the method of using clamps from a mechanical pencil to hold a pin in

place. This test allowed for design concepts to be confirmed, tinkered with, and refined.

It proved that the 3.5" SMA wire could comfortably pull the cylinder the desired distance

while the keeping the clamp closed with the desired pressure. It also provided helpful

guidance related to spacing issues.




















Figure 3-11. Clamp actuation testing model

The next step would be a complicated design process to decide upon a hole pattern.

The goal was to maximize the density of pins without sacrificing the structural integrity

of the plates and corresponding parts. Figure 3-12 shows the pattern that was chosen and

Appendix C contains details of the analysis.
















Pin Holes

Screw
S) Holes




Figure 3-12. The first hole pattern

It was desired to use vented screws to attach the SMA wires to the board as shown

in Figure 3-13. Because off-the-shelf vented screws were too expensive, two alternate

vented screws were designed that better met the needs of the system. The thought behind

this was that it would be cheaper to have a manufacturer produce a large quantity of one

of the two designs. Appendix C recounts the design of these screws.

SMA Wire
SStable
Nonconductive
/ Board




Conductive PC Board
Washer
Vented Screw Conductive Paste


Figure 3-13. Close up of electrical connection

An extra level of four plates would be added to the machine so that the number of

necessary holes for vented screws would be cut in half on each plate. Figure 3-14 shows

how the refinements evolved the system's design.











Elecica Pa I

S i- I i Mou Board



T eofplatesian t e n w a ollst

pren adoliw Bfoaice
--- I -- -Io 1 Negative
comlplectron Board


1 ectiBoard B d

P Pl Moveeig


Figure 3-14. Plate arrangement

The five types of plates in the design were as follows:

1. Clamp-holders: These plates have the clamps press-fit into their holes. These will
prevent and allow for motion of the pins.

2. Clamp-openers: These plates will open all the clamps and allow the pins and the
SMA to pass through without contact, but not the cylinders around the clamps.
When the solenoids beneath these plates push them up, the cylinders will be pushed
up, the clamps will open, and the pins will be able to move freely.

3. Power-suppliers: These plates will contain the vented screws that supply power to
the SMA wires.

4. Grounds-providers: These plates will contain the vented screws that provide a
ground to the SMA wires.

5. Pin-mover: This plate will have no holes, but will have sensors attached. The pins
will rest on it and be moved by it.

At this point, trying to move forward became very difficult as multiple problems

started surfacing. It was realized that creating the various hole configurations necessary

for the multiple plates of the machine would be highly difficult if not impossible to create

with the University facilities (especially those that required electrical traces). Therefore,

it was decided to seek help from professional machine shops. Several redesigns of the

plates were created to match the capabilities of the different machine shops. It was found









that the plate design was going to have to be compromised or the plate creation would

cost significantly higher than would have been desired. Shortly after it was found that

making the vented screws would be too expensive as well.

BioMetal Micro Helixes

The build up of difficulty led to the system's design taking a step back in order to

be capable of moving forward. Problems with board and screw creation led to a

reevaluation of the pin clamping actuation design. Discovery of a spring-shaped SMA

opened the way for new, more efficient designs. Several new designs using this material

were brainstormed and compared in detail. These analyses can be read in Appendix C.

Some of the better designs were prototyped and tested using BioMetal Micro

Helixes, shown in Figure 3-15. The micro helixes are an improvement on the SMA wire

because they require less power, take up less space, move a greater distance than wire

strands of the same length, and are easier to attach.


Figure 3-15. BioMetal micro helixes









The most effective of the designs using the micro helix models is seen in Figure 3-

3-16. The idea here was similar to the initial design in that the clamp was still held by a

stable plate and there was a stable power giving plate and another ground plate.

However, the micro helix was so small that it could be attached to the power and ground

plates with a normal wire which could be soldered or wound for any attachment type

desired. Also, the ground plate could double as a moving plate for opening the clamp.

This model performed the gripping actions better than the previous design, was easier to

assemble and created far less clutter.


PIN

WIRE


CLAMP
& CUP


"' CRl MP


.MICRO
HELIX


"VIRE


Figure 3-16. Helix model 2

Unfortunately, the micro helix design proved to be too costly for this project and

focus had to be returned to the SMA wire design that was originally in place. Even so,









the micro helix design was very helpful in evolving the SMA wire design. Figure 3-17

shows the three different designs together. With the new design, one end of a strand of

SMA wire is attached to a plate below the clamp and the other is attached to the cup and

a regular wire. The other end of the regular wire is attached to a plate above the clamp.

Power and ground are supplied from the two plates and when heated the cup is pulled

down and the clamp is tightened. The upper plate can move upward and pull on the

regular wire, which in turn lifts the cup and releases the clamps grip on the pin.


Figure 3-17. Multiple pin clamping model








While the model proved this system to be reliably functional for a single pin,

prototyping a system with multiple pins and clamps within an accurate environment lead

to several new problems. There were numerous locations where short circuits occurred,

hand assembly was difficult, and several connections were structurally unreliable.

Updating to the design shown in Figure 3-18 reduced the severities of the

problems. This final design allowed for the system to be assembled in a more convenient

step-by-step process and minimized the chance of short circuits.



TOP PLATE

J GROUND PLATE

SHRINK-WRAP TUBING MINIATURE SCREW
FREEDOM RING> WIRE-WRAP WIRE

SMA WIRE-. CLAMP
PIN
DIODE-
POWER PLATE




SENSOR PLATE



Figure 3-18. Final pin-clamp system design

The use of the shrink-wrap tubing eliminated most current shorting difficulties and

actually allowed for a smoother flow of the pin. Using looped and curled wire ends









eliminated the need for crimps and allowed for a simpler assembly and a better situation

for maintenance.

Hole Pattern

With the changing of the design the screw holes were eliminated and it was decided

to do a simple row offset pattern in a 30x30 array. When it was found that additional

space was necessary for assembly, every other hole was skipped and the hole pattern

became square.

Pin Sensors

In order to read the position of the pins, the bottom plate needed to be lined with

sensors. The sensors needed to require a minimal amount of power and take up a

minimal amount of space. Most sensors looked at were either too big or not precise

enough. The first sensor seriously considered was a hall-effect sensor.

Hall-Effect Sensors

Miniature Hall effect sensors, shown in Figure 3-19, were ordered and tested for

use in the machine. The idea was that, during virtual visualization, actual pins would be

arranged in a specific manner above the bottom plate of sensors. The plate would be

raised and the sensors would detect when each pin comes in contact with the plate.

Knowing the contact time and the location of the table at all times will result in the

knowledge of all the pin placements.


Figure 3-19. Hall-effect sensors









Extensive testing was performed in trying to find the best manner to use the sensors

to detect the pins. A successful setup was found where miniature magnets would be

placed on the bottoms of the pins, and their presence when within a desired distance

would change the magnetical field enough to alter the output of the sensor.

Unfortunately, while tests showed that the miniature hall-effect sensors were

capable of sensing when individual pins came into contact with the moving plate, several

other difficulties were discovered. The detection signals, which the sensors output, were

so small that a complicated interpretation system would have been necessary for the

computer to notice them. More importantly, the process of actually attaching 900 of the

sensors appeared to be both difficult and time consuming, if not impossible. As a result,

this method was abandoned for a more simple design.

Contact Traces and Steel Wool

The new design maintained the concept of detecting when the pins made contact

with the moving board; however, the difference involved tracing a sensor circuit in

copper onto the moving plate. Below each pin would be two copper pads (each connects

to a separate line on the computer). When the metal pin contacts the two pads, the lines

get shorted and the computer detects a pin contact. It was found that the sensors were

most efficient when the bottoms of the pins were lined with steel wool. The steel wool

provides a wide, cushioned contact bumper for the sensor, which allows for imprecise

contacts and rough impacts. The steel wool was easily attached to the bottom of all the

pins with heat-shrink tubing as shown in Figure 3-20.
















J' STEEL
^ WOOL


Figure 3-20. Steel wool attached to the bottom of a pin

Trace Pattern and Board Creation

Once there were definite designs for the clamp and sensor systems, it was possible

to design and create the plates for the physical display. The five plates from the machine

were designed in Protel 99SE and created through an out-of-house manufacturer. Each

plate was made from a yellow 9" x 9" x 0.062" sheet of G10/FR-4 glass epoxy. While a

greater thickness would have been preferable for the boards, anything thicker would have

cost significantly more than the project budget would have allowed for. Figure 3-21

shows the Protel drawing for the mentioned bottom sensor. As described, this board will

be used to raise and lower pins as well as detect their positions. Two 50-pin ribbon

cables will attach to the board delivering thirty output currents to the board. Two other

50-pin ribbon cables will attach to the board, probing thirty input lines for pin contacts.































Figure 3-21. Digital bottom plate

Figure 3-22 shows the Protel design for the ground plate that is just above the

sensor plate. This board allows the pins to pass through it freely. While the bottom of

the SMA wires will be attached to it, the ground board will remain stable and supply

grounds to the thirty rows of SMA wires. The grounds will come from the two 50-pin

cable ribbon connections.
































Figure 3-22. Digital ground plate

The power board, seen in Figure 3-23, supplies current to the attached wires which,

in turn, pass the current on to the SMA wires. The power is once against supplied from

two 50-pin ribbon cable connections to thirty rows of wires. The board is attached to

eight push solenoids and can be moved up and down. Moving the plate up will stretch

out the SMA wire and release the clamp's grip on the pin. Moving the plate down will

help maintain the clamp's grip on the pins when there might be additional outside forces

upon them.

































Figure 3-23. Digital power plate

The support plates, shown in Figure 3-24 will be attached to the ground and power

plates. These plates will be used to help keep the wires in place by pinching them in

between the boards and the support boards will add strength to prevent deformation of the

ground and power plates.


































Figure 3-24. Digital support plate

The top board, displayed in Figure 3-25, will be what the user sees from above. It

will be held stable by the surrounding skeletal structure and the clamps will be attached

to the board. The pinheads will be wider than the holes in the top board and will not be

able to move below it.


P P-































Figure 3-25. Digital top plate

Pin Movement System and Frame

The pin movement system and frame had fewer restrictions on their creation and

therefore, were basically designed by working around the other two systems. First, it had

to be decided which type stepper motor and solenoid would be used. This was necessary

because these pieces needed to be ordered from a manufacturer and the connecting pieces

would need to be designed to fit around them.

Motor Selection

Because the plate moving the pins has to make strict, easily controlled movements,

it was decided that three uniformly placed stepper motors would be best for shifting the

plate up and down (Histand et al. 1999). The most economical motor that met all the

desired specifications was the Thompson 55M048D1B Bipolar Stepper Motor, as shown

in Figure 3-26.



























Figure 3-26. Thompson 55M048D1B Bipolar Stepper Motor

It was not important what type of screws would be used to move the sensor board

up and down; therefore, it was decided to use some 3/8" round screws that were

conveniently already in-house. To decide on the stepper motor that would be used, it was

necessary to find the torque required to turn the screws that so that they would smoothly

move the sensor board with the pin resting upon it vertically. The necessary torque to be

used by each screw, Tu, could be calculated by using Equation 1 and the free body

diagram shown in Figure 3-27, where F is the downward force on the screw, f is the

coefficient of friction for the metal screw against the plastic nut, L is the lead and Dp is

the pitch diameter (Mott 1999).


F = LOAD
TO BE
MOVED Ff

P
L = LEAD
THREADED
SURFACE N


iD


Figure 3-27. Free body diagram of forces on the screw










SF-D L+-.D f.-D
2 TT-D -f-L

The downward force, F, could be found using Equation 2 where mpin is the mass a

single pin, n is the number of pins, and Wpiate is the weight of the sensor board and the

bracket it is attached to.

F = n m,,, g + Wpte (2)

The mass of an individual pin was measured at 2.92g on an electric scale, the

weight of the sensor board and bracket was approximated at 51b, and at the time of the

calculation an n of 900-pins was used. An F of about 1 llbs was calculated. Substituting

that into Equation 1 along with a Dp, of 0.5", an L of 0.25", and a worst-case f of 0.5, Tu

was calculated to be 1.68 lb-in. To find the desirable amount of torque that each stepper

motor would provide, Equation 3 was used with a safety factor, SF, of 3, and a stepper

motor total, nsM of 3.


Tdesired -SFT- (3)


After converting the units the desired torque was found to be 31.5 oz-in per stepper

motor. The Thompson 55M048D1B Bipolar Stepper Motor was chosen for use because

it could provide 33.5 oz-in and was an economical purchase.

Solenoid Selection

In order to re-stretch the SMA after having been heated and condensed, the clamp

sleeve (which is attached to the SMA) will be moved by one of the spinal plates. It was

decided that the plate will be moved by a three Magnet-Schultz DC voltage cylindrical

push solenoids, as displayed in Figure 3-28. Moving the plate will move the sleeve and

in turn stretch the SMA.
















Figure 3-28. Magnet-Schultz DC Voltage Cylindrical Push Solenoid

Because the motion required to move the ground board up and down was a discrete

action, it was decided that a solenoid would perform this task best. In order to decide

upon an exact solenoid it was necessary to decide how much force (Fmax) each solenoid

was going to be required to apply. The majority of the force used from the solenoids

would be used to push the freedom ring over the clamp, everything else could be

considered negligible. The maximum force (F1) required to stretch out a single SMA was

calculated using Equation 4, where mMax is the maximum amount of mass required to

pull the freedom ring over the clamp and g is the acceleration of mass due to earth's

gravity.

F, = mMA, g (4)

Tests with an electronic scale showed mMax to be approximately 10 grams. Fmax

was then calculated using Equation 5 where n is the number of freedom rings being

pulled down and N is the number of solenoids being employed.

n.F
Fmax = (5)
N

It was decided that the operation would be most balanced if a solenoid were

employed at each corner of the board, thus making N equal 4. At the time it was believed

that there would be 900 pins, thus making n equal 900. Calculated out Fmax was found to

be 353 oz. The Magnet-Schultz DC voltage cylindrical push solenoid was capable of

supplying 515 oz of force and was most economical for its capabilities and therefore









became the solenoid that was implemented. When n decreased as the design evolved, the

force calculated was still sufficient because the additional force could only be beneficial.

Shaft Selection

The shaft system that moves the bottom sensor plate up and down was designed

and drawn up in Mechanical Desktop, as seen in Figure 3-29. The stepper motor spins a

3/8" screw to move a nut up and down, and the bottom plate is attached to the nut. The

stepper motor will be attached to the screw with a flexible coupling. The ends of the

screw will be supported by single-row angular contact ball bearings which will protect

the screw from axial and radials loads.



SCREW






BUSHING
BEARING
FLEXIBLE
COUPLING
STEPPER
t >J MOTOR




Figure 3-29. Shaft / motor system

The design of the shaft system began with an assessment of what was available,

what was necessary, and what was desired. The following facts were known:

* Three 3/8" in diameter screws would serve as the shaft









* The 3 screws would need to be connected to three stepper motors having shafts of
0.25" diameter and 0.375" length

* 4 to 6 inches of plate movement were desired from the shaft

* Bearings would be necessary to hold the screws in place and prevent damage to the
steppers (to isolate)

Flexible couplings were bought to connect the screw shaft to the stepper motor

shaft. It was decided that the couplings would be flexible so as to prevent damage for the

likely situation that the two shafts would not be perfectly aligned. It was decided that the

ends of the screws would be worked down to a quarter inch in diameter so as to match the

stepper motor shaft and make it easier to find a functional coupling. Because bearings

could not be found that would exactly fit the ends of the screws, bushings were created.

In order to design the bushings, decide the exact screw length, and make certain that

everything would fit together, all the pieces for the system were drawn two-

dimensionally in AutoCAD, as shown in Figure 3-30, and aligned properly.







































Figure 3-30. Two-dimensional drawing of the shaft/ motor system alignment

With the shaft system and the plates designed, it was possible to create the

surrounding structure, which is shown in Figure 3-31. The structure was designed in

Mechanical Desktop and is approximately 24" x 20" x 27" in size. The openness of the

design allows for easy viewing and accessibility to the machine's many parts within the

structure.


































Figure 3-31. Skeletal structure of 80/20

The frame was designed with the purpose of being capable of holding together all

of the display's working parts, bearing the weight and forces from the display's actions,

and still being open for easy viewing and maintenance. It was made from extruded

aluminum and connected using groove-fitted screws and nuts. Figure 3-32 shows the

assembled frame with the cooling fan attached.































Figure 3-32. Assembled 3-D display frame with fan

Next, additional mounts had to be designed to connect all the pieces. The "motor

mount," seen in Figure 3-33, serves as an important hub between the shaft system and the

machine structure. Made from delrin, this mount braces one of the single-row angular

contact ball bearings attached to the shaft while also holding the stepper motor on the

bottom. The flexible coupling is sheltered in the center of the mount, but still easily

accessible for tightening through two side holes.






























Figure 3-33. Motor mount

The "upper bearing mount" braces the contact ball bearing on the top of the shaft

and connects it to the 80/20. Shown in Figure 3-34 and made from delrin, this bearing is

aligned in the z-axis with the motor mount to complete the shaft system. The three shaft

systems are used to move the bottom plate up and down.























Figure 3-34. Upper bearing mount
The "solenoid mount," seen in Figure 3-35, has one push solenoid screwed into its
top socket and another screwed into its bottom socket. The moving pins in the solenoids
are pointing towards one another with the power plate in between them. There is a
solenoid system attached to each corner of the power plate. The bottom solenoids are
able to push the plate up with the purpose of stretching the SMA wires and releasing the
clamps' grips on the pins. The top solenoids push down on the power board in order to
maintain the clamps' grips on the pins. When neither solenoid is active, the plate is held
in its desired place by springs resting around the pushing pins inside the solenoids.


. eAft*S






























Figure 3-35. Push solenoid mount

Figure 3-36 shows a preliminary drawing of the machine assembled without the

pins, clamping systems, or plate attachments. The drawing shows the relative placements

of the three shaft systems, four push solenoid systems, and plates with respect to the

aluminum structure.










PLATE
FAN


SOLENOID
MOUNT

80/20
SKELETAL
STRUCTURE






SHAFT


Figure 3-36. Preliminary assembly drawing

Brackets were designed to attach the various plates to the to the machine frame.

Each design took into consideration the fit of the parts around it, the actions it would be

making, and the forces that it would have to withstand. It was decided that making the

brackets from steel would be structurally sound while still being economically efficient.

Figure 3-37 shows a detailed dimensioned drawing of the bracket for the top plate. Two

of the L-shaped brackets attached to opposing ends of the top plate. The top plate was

screwed to one section of the "L" and the top plank of 80/20 extruded aluminum was

screwed to the other section. These brackets hold the top board rigid while allowing for

free movement of the solenoids (which move the ground plate) around the brackets.


































a-rt, I- [Qe~ Brackel
MoaerIla, 6061 ALunlnur
Tolerance, :0.002'
Units Inches
S-l~e, 2,1O flo+e, 10/2/023


Figure 3-37. Top plate bracket

Figure 3-38 shows the detailed dimensioned drawing of the brackets for the ground

plate. These brackets were screwed to the opposing ends of the ground board for a rigid

attachment. The other ends of the brackets were attached to four push solenoids each.

Two solenoids were inserted into opposing ends of each hole in the brackets. Pushing

from the solenoids on the brackets leads to a corresponding movement of the ground

plate.







56



























Part Power Plate 1rrcket
Matgerlat: 6061_ Alumlnum
ToLer&ncel 0.002'
Uni-ts Inches
Scole, 310 JDate 10/2/03


Figure 3-38. Ground plate bracket

Figure 3-39 shows the detailed dimensioned drawing of the brackets for the power


plate. Opposing ends of the power plate were screwed to the inside on the longest section


of these two C-shaped brackets. The two smaller symmetric sections of the brackets are


two planks of 80/20 on the level of the frame that is third from the top. These brackets


hold the power plate rigid even when the attached SMA wires are pulling on it.












































Figure 3-39. Power plate bracket

Figure 3-40 shows the detailed dimensioned drawing of the bracket for the sensor


plate. The sensor plate is screwed to this singular piece for a rigid attachment. The three


arms of the bracket are screwed to the nuts on the three motor / shaft assemblies. Turning


the shafts in unison result in a symmetric movement of the nuts. This, in turn, moves the


bracket and sensor plate up and down (lowering the sensor plate allows for movement of


the pins).


[_J




















Par Ground PR-te lrackei
Drawn By: Davd Lewls
MaterlaUi 6061 Akumlnum
Tolerance: -t0.002"
Units, Inches
ScoL'l 2'10 Ilnte: 1011/03

































Figure 3-40. Sensor plate bracket

All the parts that were designed specifically for the display device were

manufactured. The following parts were all created in the UF machine shop:

Two Top Plate Brackets
Two Ground Plate Brackets
Two Power Plate Brackets
Three Upper Bearing Mounts
Three Motor Mounts
Four Solenoid Mounts
Six Bushings
Six Screw Shafts

To make the top plate bracket, a precise milling machine was required; because this

was not available in the UF shop, the piece was created by an out-of-house manufacturer.

The pieces were press-fit together as with their associated parts.

Figure 3-41 shows a picture of the finalized physical display, having been fully

assembled.



























Figure 3-41. The fully assembled physical display














CHAPTER 4
CONTROL BOX

Where the computer serves as the brains of the machine, the control box works as

the circulatory system. Control signals are delivered from the computer to the box where

they are processed, re-routed and the appropriate signals are sent to the display device.

This system contains numerous terminal boards, power sources, controllers and

thousands of wires that are vital to the transportation of electrical signal through the

machine. This chapter will detail the processes used by the control box, how it is

organized, and parts that make it up.

Raster Scanning

As a result of having hundreds of pins, the 3-D Visualization Device has hundreds

of actuators and hundreds of sensors that need power. It was an important task that the

amount of power being distributed be as minimized as possible. The solution that was

implemented was a process known as raster scanning. This process only works for grids

of electrical devices where the number of columns is equal to the number rows, which

was conveniently the case at hand. The benefit of raster scanning is that it decreases to

the total number power lines needed to 2 N, where N is the total number of electrical

devices. The negative is that only a single column of devices can be activated at any

single moment. The simple solution to this problem is that each column is activated in a

pattern of quick sequential bursts. The result is visually the same as activating everything

at once. The setup for a raster scan is basically that each column of devices is receiving a

common ground line and each row devices is receiving a common power line. This can






61

be reversed, as the important part is that the columns are all receiving one type of power

and the rows are receiving another type of power. Figure 4-1 diagrams an example setup

for a three by three setup of actuators.


P









GROUND


Figure 4-1. An example 3x3 grid of actuators in a raster scanning sequence

The diodes in the setup keep current from flowing in an undesired direction. For a

of grid actuators, the columns would output power until it is desired to begin actuation.

At that point, the first column will output a ground, and the rows will send power only to

the rows that are to be actuated. Next, the first column outputs a power again. Then, the

second column quickly outputs a ground and the row outputs change to give power to the

actuators being activated in that column. This pattern is repeated for the remaining

columns and continues for the remaining time of activation. To the human eye it appears

that the actuators are all being activated once.

For the sensors, such as in Figure 4-2, the process is a similar process with polling.

The columns are only powered one at a time, while all the rows are read. The rows use a

pull down resistor, trying to draw a current. If a current is read, then the sensor at that






62


row and column has been contacted. The polling of columns can be done so quickly that

the off time for each column is negligible.














POWER


Figure 4-2. An example 3x3 grid of sensors in a raster scanning sequence

Mapping

The control box can initially be looked at as a black box, as show in Figure 4-3.

Six 50-pin cable ribbons come out of the computer and connect to their own individual

terminal boards. Each ribbon contains 24 actual signals, which are each broken down

into an individual wire through the terminal board. Appendix D details the mapping of

where each cable ribbon pin goes and what it does. After being processed, the various

signals are processed and enter eight more terminal boards, which send information to the

display device through eight more ribbon cables.






63



C ACTUATORS D
O o SENSORS I
M WI S
P DIGITAL CONTROL E SOLENOIDS
R P
U SIGNALS BOX MOTORS L
T PIN A
T q N IN CONTACT A
E I SENSORS Y
R PLATE



Figure 4-3. Direction of signals through the control box

Figure 4-4 shows a basic mapping of how the control box is connected within and

to the surrounding components. Sixty-three of the signals are used to control sixty-three

opto-isolators. These opto-isolators act as switches controlling the flow of current too

high for the computer to handle. Fifteen opto-isolators control the flow of power to the

power board. Fifteen opto-isolators control the deliveries of ground flows from the

ground board. Fifteen opto-isolators control the flow of current to the sensor board and

fifteen opto-isolators control the reception of current from the sensor board back to the

computer. One opto-isolator is used to control the flow of power to the four solenoids,

which push the ground plate down, and another opto is used to control the flow of power

to the four solenoids pushing the plate up. Another opto receives signals from the sensor

on the display device that reads when the sensor plate is completely down. A stepper

motor driver inside the box is used to control the timing, speed, and direction of the three

stepper motors in the display device. Two power supplies are used to power all the

components of the display device. A MAP130-1024 Power-One DC power supply sends

24 Volts to the eight solenoids on the display, and another 24 Volts that are used to power

the stepper motor driver. An L&C Inc. switch power supply sends 12 Volts to the

columns of the sensor board, and 5.25 Volts to the power board.










CONTROL BOX


Figure 4-4. Basic mapping of the control box














CHAPTER 5
COMPUTER

The computer serves as both the brain of the machine and a way for the user to

communicate with the display devices. Its three main components are the Graphical User

Interface (GUI), the path planner, and the Digital I/O. This chapter details how those

components work and what they do.

Overview

The computer was setup for controlling the machine, as seen in Figure 5-1. A

Linux operating system was installed and Motif and Open Inventor were installed for

running the graphical user interface (Brain 1992, Wernecke 1995). Hardware programs

were installed to run the digital I/Os and motor controllers.


















Figure 5-1. Computer setup






66


Graphical User Interface

In order for the user to be able to control the machine from a computer, a

Graphical User Interface (GUI) was created using Motif and Open Inventor. This

interface is used to both tell the physical display how to position itself and present a

rendered representation of the current physical display. Figure 5-2a shows the GUI that

was created to control the three-dimensional visualization machine. The user can read

the current positions of the pins in the drawing by entering the row and column numbers

of the desired pin and clicking on a "Get Height" button. It is important to note that the

rows and columns are labeled "1" through "15". Entering a row, column and height and

pressing the "Set Height" button will change the height of a pin. As shown in Figure 5-

2b, an entire positioning set for the pins can be read from or saved to a file using the

"Read" and "Write" buttons. To send the pin positions to the physical display, the

"Send" button must be clicked. To display the current pin positions from the physical

display, the "Receive" button must be pressed. Also, the display view can be adjusted by

dragging the rotating wheels and using the icons alongside the display.








I
S- f_





(a)l gur (b5 Grp ic user itr c (al G I

Figure 5-2. Graphical user interface (a) Example GUI display, (b) GUI reading a file









The GUI also, has a series of buttons on the right side that can be used for testing

the various sections of the physical display. The "Bottom Solenoids" button is used to

make the solenoids push the power plate up and the "Top Solenoids" button has them

push the plate down. The "Move Motor" button can be used to move the sensor plate up

and down. The user places a desired height for the pins in the corresponding text box,

then left-clicks the button and the pins are moved to that height. The pins and plate will

remain at that position until another command is made. The "Activate Pin" button can be

used to trigger a specific pin actuator. The user writes the row and then the column,

respectively, of the pin to be actuated in the corresponding text boxes. After the button is

clicked, the SMA will be heated and the pin will be clamped. The "Read Pin" button can

be used to test the sensor board. The user enters the row and column of the pin to be

tested in the text boxes used for the "Activate Pin" button and clicks the "Read Pin"

button. Next, the test field next to the "Read Pin" button will say whether or not the pin

at the specified position is engaged with the sensor.

Path Planner

In order for the user's input to reach the pin display device, a "path planner"

program was created in C++ that interprets the inputs from the user interface and

executes them through the digital I/O. This program executes all major actions such as

plate movement, pin sensing, pin clamping, and SMA stretching. It was decided that the

program could use an open loop system to move the pins up and down because sensors

will allow the program to rediscover the bottom plate's location at the end of every

display process and any location errors that could occur in between would be so

negligible that the human eye would not be able to notice them.









The path planner program utilizes the following algorithm for physical

visualizations (The terms from the code can be interpreted by using Appendix E):

1. 'Send' button sendPhysical(), sendPhysical2() (MakeltReal.c++)
-------- sendCB() (main.c++)

2. If machine is not busy, lock machine and call sendPhysical(); else, stop.
--------- sendPhysical( (MakeltReal.c++)

3. Load all pins into a priority queue, sort by height (tallest first).

4. Set sensor board speed to fastest setting.

5. Move board to top.
-------- popup_routineo (MakeltReal.c++)

6. Popup dialog asking if user wants to unlock pins by activating solenoids.

7. Continue prompting until user answers, "No, continue."
-------- sendPhysical2() (MakeltReal.c++)

8. Set sensor board speed to slowest setting.

9. Pop front pin off of priority queue and descend to its height.

10. If that height is greater than zero, fire the pin.

11. Reset sensor board speed to fastest.

12. Unlock machine.

In turn, for virtual visualization, the program uses the following algorithm:

1. 'Receive' button receivePhysical(), receivePhysical2() (CompDisplay.c++)
-------- receiveCB (main.c++)

2. If machine is not busy, lock machine and call receivePhysical(); else, stop.
-------- receivePhysical (CompDisplay.c++)

3. Set sensor board speed to fastest.

4. Move board to top.

5. Activate and hold solenoids to lock clamps.

6. Move board to bottom.









7. Popup dialog allowing user to adjust pins. receivePhysical2() when clicked.
-------- receivePhysical2() (CompDisplay.c++)

8. Set sensor board speed to slowest.

9. Move board up and at each click check all pins to see if engaged.

10. Record the lowest height at which a pin was engaged (in 'pinsArea').

11. Unlock the solenoids.

12. Reset sensor board speed to fastest.

13. Unlock machine.

Digital I/O

Two PCI-7296 digital I/O cards, seen in Figure 5-3, were installed into the

computer. Each card has 96 entries capable of sending or receiving data signals. They

are used for the following functions:

* 30 outputs for clamp control

* 15 outputs for running sensors

* 15 inputs for reading sensors

* 3 outputs for running motor controllers

* 2 outputs for controlling solenoids

* 1 input for sensing the bottom placement of the bottom plate


Figure 5-3. PCI 7296 digital I/O card














CHAPTER 6
ASSEMBLY

The assembly process of this project was complicated, difficult, and greatly time-

consuming. While a majority of the design and testing occurred before the assembly

process began, the design and testing processes became mixed into the assembly process

as well. The physical display had to be restructured several times in order to be

accessible for hand assembly as well as to solve unexpected problems. Multiple design

questions could only be answered by performing tests as the device was being assembled.

As this was a prototype machine, many decisions had to be made on the fly and the

process took up a pattern of assemble, test, re-design, and repeat. This was the step of the

project where a don't-look-back approach had to be taken at times. If a problem

occurred, it was corrected to the best degree possible without having to start over again.

Now that the machine is completed, it would be recommended to perform some assembly

processes differently with some redesign changes. Those recommendations are presented

in Chapter 8. This chapter details exactly how this machine was assembled and how to

recreate it.

Control Box

The control box was basically assembled by attaching the electrical controls to four

perforated shelves. The shelves have rails on them, which were inserted into rails inside

a large metal box. The seven opto-isolator boards, the two power sources, and the

stepper motor driver were screwed into the perforations in the shelves. The terminal

boards had no holes for such attachments, so they were connected to the shelves with









snap ties and Velcro was added for additional support. The ports were then wired

together following the previously mentioned mapping of Appendix D. Multiple colors of

wire were used in order to distinguish separate sections of the mapping.

Pre-Clamp Physical Display

Because the pin clamp systems were delicate and permanent, it was necessary to

have everything else on the physical display assembled and working correctly first.

Frame

The first part of the physical display to be assembled was the frame. It is composed

from the following parts:

* Four 20" long extruded aluminum strips
* Four 16" long extruded aluminum strips
* Four 11.5" long extruded aluminum strips
* One 11" long extruded aluminum strips
* Four 10" long extruded aluminum strips
* Four 9" long extruded aluminum strips
* Four "T"-brackets
* Fifty-Six "L"-brackets

Because the frame could not be assembled all at once, it had to be made in sections.

Figure 6-1 shows what the assembled frame sections would look like without any other

attachments.


ARM









Figure 6-1. The sections of the frame where the red circle encompasses a waste section,
the blue circle encompasses an arm section, and the green circle encompasses
a leg section

All the strips of aluminum were joined by placing T-nuts inside the aluminum

extrusions and screwing the hinges to them. What follows are the three types of frame

sections that were created:

* Two waists were made by joining the aluminum strips into the pattern shown in
Figure 6-2.


Figure 6-2. Waist section of the frame

* Two legs are used to support the entire display. Joining the aluminum strips in the
pattern displayed by Figure 6-3 can create a leg.


















11.5




2011


Figure 6-3. Legs section of the frame

* Two arms are used to hold the top two plates. Figure 6-4 shows the aluminum strip
assembly pattern


Figure 6-4. Arms section of the frame

The remainder of the assembly was matter of carrying out the following specific


procedure:









Motor/Shaft Systems

1. Press fit a bearing into each of the three motor mounts.

2. Press fit a bearing into each of the three upper bearing mounts.

3. Press a single bushing onto each of the three shafts so that the widest end of the
bushing is flush against the widest end of the shaft.

4. Tighten the screw on the bushing so as to tighten their grip on the shafts.

5. Insert a flexible coupling into the center of each of the three motor mounts so that
the coupling's center hole is facing the top and bottom of the mount.

6. Insert a stepper motor shaft into the bottom of each motor mount and into center
hole of the flexible coupling within.

7. Screw the three stepper motors on.

8. Insert the end of a screw shaft that hangs out of the bushing into the top of the
motor mount and into the top end of the flexible coupling until the bushing is flush
against the bearing.

9. Screw the nuts onto the screw-shaft so that the threads on the nut are at the top

10. Screw the three nuts onto the sensor board bracket and move the bracket to the
lowest possible location.

11. Screw the three motor mounts onto one of the waste sections of the frame so that
the corners of the bracket are symmetric with the section.

12. Screw the three upper bearing mounts to the other waste section of the frame (make
certain to leave the screw loose enough that the mounts can slide on the frame
section.

13. Place a bushing on the top of each of the screws so that the widest end of the
bushing is flush against the widest end of the screw.

14. Tighten the screws on the bearings.

15. Place the second waste section of the frame on top of the other by inserting the
ends of the screws coming out of the bushings into the bearings in the upper
bearing mounts until the bushings are flush against the bearings.

16. Adjust the placement of the upper bearing mounts so that the screw-shafts are
perfectly upright and the two waste sections of the frame are symmetric with one
another.

17. Tighten the screws on the upper bearing mounts.









18. Screw the leg sections of the frame onto the waste sections (the exact placement
that they are attached isn't as important as making certain that the waste sections of
the frame are level.

Brackets

1. Screw a top plate bracket onto each arm section of the frame.
2. Screw the power plate brackets onto the top waste section of the frame.
3. Screw the ground plate brackets onto the ground plate.

Plates

1. Screw the sensor plate onto the sensor plate bracket.
2. Screw the power plate onto the ground plate brackets.
3. Screw the top plate onto the top plate brackets.
4. Screw the arms loosely onto the top waste section of the frame.
5. Align the top plate with the power plate and tighten the arm attachments.

Solenoids

1. Screw two solenoids into each of the four push solenoid mounts.

2. Screw the solenoid mounts loosely onto the four corners of the arms.

3. Place a solenoid shaft through each of the four top solenoids and tighten a collet
around each shaft so that about 1/8" of the shaft hangs out of the collet.

4. Place a solenoid shaft through each of the four bottom solenoids, place a spring
around each shaft, and tighten a collet around each shaft so that about 1/8" of the
shaft hangs out of the collet.

5. Slide the ground plate into its proper position and insert the eight solenoid shafts
into the four ground plate bracket holes so that each hole has a shaft entering it
from each end.

6. Re-adjust the collets on the solenoid shafts so that they are flush against the ground
plate bracket.

Pin-Clamp Systems

The pin-clamp systems were the most tedious and time-consuming work of the

entire project. Most sections of the system had to be shaped, created, or assembled by

hand. The time involved in creating the parts was a large reason that the number of pins

was lowered from 900 to 225. The difficulty involved in the hand assembly was why the









spacing of the pins was increased. Each section was created in mass before creating the

next section and each unit assembly followed a carefully structured course of action. It is

also important to note that the number of units created for each section outnumbered the

number of units needed. This was important because many units were broken in during

assembly and testing, and it saved time to have the additional units made. The following

assembly procedures detail how to make each section of the system and how to install

them into the machine:

Mechanical Pencil Disassembling

Each of the moving pins on the physical display device required a working clamp

with a freedom ring to lock its position. Each clamp and freedom ring needed to be

removed from a single Pentel Crayz 1.3 mm mechanical pencil. When it was still

planned to have 900 pins, 996 pencils were ordered and stripped down to the clamp,

spring and freedom ring. Because of the large number pencils, the stripping process was

broken down into five specific steps.

1. Disassembly: All parts of the mechanical pencil that were capable of being
removed by hand were detached and collected. The detached parts included the
eraser cap, eraser, two sticks of lead, the pencil cap, and the cone cover.

2. Shell Removal: The clamp was attached to a long cylindrical, semi-rigid plastic
sleeve. A much harder plastic shell surrounds all this, which is what the user grips
when writing with the pencil. The hard shell could only be removed by being
broken off. The pencil was placed in a vice and when the vice increased the
pressure, the shell would fracture linearly and could be peeled off.

3. Sleeve Slicing: The semi-rigid plastic sleeve holds the clamp in place and an
increase in the sleeve's diameter holds the freedom ring and spring over the clamp.
A dremel with circular blade was use to slice the sleeve through its smaller
diameter where it starts to cover the clamp. This removed all of sleeve except for a
thin ring, and, it allowed for the following step.

4. Spring Removal: Without the sleeve to hold it in place, the springs could be
removed from around the clamp and placed and bin where they would be used
later.









5. Sleeve Dissection: In order to remove the remaining ring from the plastic sleeve
around the clamp, the dremel with the circular blade was used to carefully slice
through the ring on two sides. The ring was, in turn, dissected into two pieces that
would just fall off. Because the clamp was made of a harder plastic than the sleeve,
the dremel could be set at a low enough spinning rate that it would cut easily
through the sleeve and cause a negligible amount of damage to the clamp. Once
the sleeve was removed the freedom ring could be pulled off and the clamp and
ring could be stored in their associated bins, as seen in Figure 6-5.
























Figure 6-5. Bins of separated materials

Brass Cup and Wire Wrap

Each brass cup needed to be attached to a wire-wrap wire. This process needed to

be performed in a precise manner and units that did not meet the desired specifications

had to be scrapped of everything except the brass ring. Because the wire-wrap can be

delicate when bent, it was necessary to make an additional 75 functional units. What

follows is the procedure used to create them.

* Step 1: Cut 300 0.125" lengths of 3/16" in diameter heat-shrink tubing.









* Step 2: Place the bottom of a drill bit in vice so that the bit stands vertical.

* Step 3: Place brass freedom cup around the top of the drill bit so that the rim of the
cup is closest to the ground.

* Step 4: Hold a 3" long length of wire-wrap wire along side the freedom cup so that
the wire is parallel to the bit and its center is against the cup.

* Step 5: Slide a length of heat-shrink tubing around the wire-wrap and freedom cup
so that bottom of the tubing pushes against the rim of the cup.

* Step 6: Heat the tubing so that it shrinks completely around the cup and wire.

* Step 7: Let cool.

* Step 8: Repeat steps 3-7 another 299 times.

Diode Curling

The diodes, shown in Figure 6-6, were necessary to prevent electricity from

flowing in the wrong direction. The outgoing wire needed to have a loop that the SMA

wire could connect to and the incoming wire needed to have an arm that could be

attached to the trace on the power board. Following is how each diode was structured:

* Step 1: Curl the outgoing end of a diode into a spiral.

* Step 2: Cut off the excess wire after the second curl.

* Step 3: Curl the ingoing end of the diode into an L-shape.

* Step 4: Cut off the excess wire so that both brackets of the L-shape are
approximately 0.25".

* Step 5: Repeat steps 1-4 another 249 times.


., "


Figure 6-6. A diode before assembly and a diode after assembly









Guide Tubes

The black guide tubes, seen in Figure 6-7, hold the pin clamps in a solid position

and provide directional guidance for the pins. Following is the process for making them:

* Step 1: Cut out 275 1.25" long strips of black 3/16" in diameter heat-shrink tubing.

* Step 2: Cut out 275 0.25" long strips of clear 3/16" in diameter plastic tubing.

* Step 3: Clamp one end of a steel rod into a horizontal position so that it is
untouched everywhere else.

* Step 4: Place a strip of plastic tubing around the rod.

* Step 5: Place a strip of heat-shrink tubing around the plastic tubing so that one end
overhangs the plastic tubing by about 0.25".

* Step 6: Shrink the tubing with a hot air.

* Step 7: Let cool.

* Step 8: Repeat steps 4-7 another 274 times.













Figure 6-7. On the top-left is the heat-shrink tubing, towards the center is the clear
tubing, and on the right is the assembled guide tube

Steel Wool Attachment

Steel wool was attached to the bottom of each pin for the purpose of contact with

the sensors. The amount of steel wool used for each pin was arbitrary and not very

important as long as it was not so much that it would contact the next sensor over and not

so little that it could not flew when it rested on the sensor board.









* Step 1: Cut out 250 0.5-inch lengths of 3/32" in diameter heat-shrink tubing.

* Step 2: Rip out a dime-sized amount of steel wool.

* Step 3: Wrap the steel wool around the end of a music wire. Roll the steel wool
between fingers and squish the length down until the steel is about a half inch in
length and has a football shape.

* Step 4: Place heat-shrink tubing around opposite end of wire and move down until
most of steel wool is cover with a small ball of wool still hanging out (make certain
the wire still goes to the end of the heat-shrink tubing.

* Step 5: Heat the tubing until it shrinks completely.

* Step 6: Let cool.

* Step 7: Repeat steps 2-6 249 more times.

SMA System Assembly

The connection of SMA wire to the diode and the wire-wrap wire needed to be

done carefully for two reasons. Bending or over-stretching the SMA wire could render it

useless without any visible problems until after being installed into the display device.

Accidental bending of the wire-wrap wire could make it weak and lead to wire breaking

during the later assembly. The pieces were attached to one another in the following

manner:

* Step 1: Bend an SMA wire approximately 1" from its end.

* Step 2: Loop the bend and place the closest end through the loop, so as to create a
loop that is approximately 0.25" in diameter.

* Step 3: Place the same end through the same hole in order to maintain the stability
of the knot.

* Step 4: Bend the SMA 3.5" from the loop.

* Step 5: Cut the SMA 4.5" from the loop.

* Step 6: Repeat Steps 2&3 on the opposite end of the wire.









* Step 7: Bend the wire-wrap wire on a brass cup assembly piece so that when the
SMA is completely straight and the wire-wrap wire is straight, the rim of the cup
will be 4" from the furthest loop on the SMA wire.

* Step 8: Place the wire-wrap wire through the closest loop on the SMA wire and
wrap around the loop so that the bend in the wire-wrap wire is around the very top
of the SMA wire.

* Step 9: Wrap the excess wire-wrap wire around the SMA loop and down about
another half-inch and back up again.

* Step 10: Swivel the unused SMA loop around the curled end of a shaped-diode and
clamp into place with pliers.

* Step 11: Repeat steps 1-9 another 249 times.

Final Attachment Process

The installation of the clamping systems was easily the most difficult part of the

assembly process. It required poise, patience, and awkward use of the hands. It is

virtually impossible to perform this assembly by hand to a degree of precision that is

necessary for perfect results. The following steps were taken to perform the attachment:

* Step 1: Place the threaded end of a 4-40 socket head cap screw through a hole in
the ground board to the front-left of the hole where the Pin 1,1 will be placed so
that the threaded end comes out on top.

* Step 2: Screw a 4-40 small pattern machine screw nut onto the screw and tighten it
so that the screw can not move (if the screw is not secure, it may loosen and fall off
later due to shaking of the board).

* Step 3: Repeat steps 1-2 for the remaining 224 pin holes, respectively.

* Step 4: Place a 4-40 socket head cap screw through the hole in the power board (so
that the threaded end comes out from the bottom of the board) that is directly below
the screw for Pin 1,1 in the ground board.

* Step 5: Screw a 4-40 small pattern machine screw nut loosely onto the screw.

* Step 6: Select an SMA system assembly and pinch the bottom wire of the "L"
shaped diode wire between the head of the screw and the copper tracing beneath (it
may help to curl the bottom stretch of wire with a pair of pliers so that it fits more
snugly around the screw).









* Step 7: Tighten the nut on the screw so that it cannot move.

* Step 8: Loosen the screw directly above.

* Step 9: Wrap the wire-wrap wire around the loosened screw (between the nut and
the copper tracing on the ground board) so that the SMA system assembly is taught.

* Step 10: Tighten the screw so that it cannot move.

* Step 11: Place a guide tube through the hole in the top board where Pin 1,1 will be
placed.

* Step 12: Trim the guide tube.

* Step 13: Place a pencil spring around the guide tube and a pencil clamp through
the freedom ring on the SMA and into the guide tube so that it is tight.

* Step 14: Thread the thin end of an assembled pin through the bottom of the power
board, into the clamp and out the top of the guide tube.

* Step 15: Test the system to make certain it works ("C" shaped tubing can be
wrapped around the guide tube to tighten the pencil spring and hot glue can be used
to tighten the wire-wrap wire.

Elimination of the Fan

Originally, the final step of the assembly called for the attachment of a fan to be

attached in line with the pin-clamp systems as shown in Figure 6-8. However, when the

machine reached that stage, there was worry that the system was too fragile to handle the

fan. Also, because only one SMA was being heated at a time, there really wasn't much

chance of the heat in that area being any hotter then the rest of the room; thus, making the

fan unnecessary anyhow and therefore, the fan was never attached. Once the machine

was run it was obvious that the fan was not necessary.







83


PLATE
SOLENOID
FAN MOUNT


Figure 6-8. Physical display with fan














CHAPTER 7
TESTING AND RESULTS

As discussed in the previous chapter, testing was a vital process in the progression

of the design of the 3D virtualization device. Each concept needed to be implemented

into a smaller scale prototype before it was assimilated into the final design. This process

was used from the time the first design concepts were thought up until the final tinkering

of the finished product. This chapter further details the tests used to prove or disprove

concepts, their results, and the adjustments that they led to.

Tests

Clamp System Tests

First SMA test

When an initial clamp design was complete and a basic setup existed, tests were

run in order to evolve the clamp system design. A one-pin model, shown in Figure 7-1,

was created to test the clamping process that will be vital to the visualization machine.

The model tested the method of using clamps from a mechanical pencil to hold a pin in

place. Pulling a cylindrical collar over the clamp closes it and locks the pin in place. The

collar is pulled down though use of SMA wires. These wires contract when heated with

an electrical current. Several methods of attaching the wire to the collar had to be tested.

The most successful tests came by drilling holes through a thin rim on the collar,

threading the wire through the holes and attaching the two ends to the bottom of the

model frame. The ends of the wire were where the power attachments were made. The

frame was designed on a CAD program and then created in a rapid prototyping machine.









The clamp was attached to the frame by being press-fit, and hot glue was used to hold it

in position.




















Figure 7-1. Clamp actuation testing model

This test officially proved that the SMA wire could be used to pull the freedom cup

over the clamp and provide the necessary force to hold the pin in place. It also showed

that the length of wire between the cylinder and electrical connection could be minimized

to 3.5". In the initial system, the wire was run to the cylinder and back again because

there was no convenient way to place an electrical connection at the cylinder. It was

found that the system remained functional when one of the sections of wire run from the

cylinder to the electrical connection was replaced with a normal wire. This cut the

amount of SMA wire needed in half.

Micro helix tests

Two Models were made to test the BioMetal Micro Helix in clamp actuation.

Figure 7-2a shows the first model. The idea here was that the cup would be held stable in

the plate and the micro helix would pull the clamp down into the cup, thus preventing









movement of the pin. The second model shown in Figure 7-2b turned the clamp upside-

down and held it stable within the plate. This time the micro helix pulled the cup down

to tighten the clamp and prevent movement of the pin. While both actuators

accomplished the desired task, the second model proved to be more reliable. Leaving the

clamp loose, as it was in the first model, made the pin shakier. Also, when the pin was

moved, the clamp maintained a better grip when it was positioned as it was in the second

model.








.PIN

WIRE





CLAMPCLAMP
a & CUP





MICRO
HELIX



WIRE
(a) (b)

Figure 7-2. Micro helix tests (a) Helix model 1, (b) Helix model 2

Unfortunately, the micro helix design proved to be too costly for this project and

focus had to be returned to the SMA wire design that was originally in place.









Electrical Tests

Raster Scan Test

A test was created in order to prove that the raster scan test would be capable of

controlling the clamp and to make certain that some of the parts being used would fit the

system properly. Figure 7-3 shows the setup for the test. Basically, a 3x3 grid of LEDs,

each in series with a resistor, was used to represent the SMA's. The LEDs were wired to

a power source in the pre-described raster scan method. The power source was wired to

the grid through a board of opto-isolators that would actually be used in the finally

machine. The opto-isolators were controlled by PICmicro 16F627 micro-controller. A

keypad that was wired to the micro-controller could be used to make a specific pattern of

LEDs light up. The micro-controller would determine what button was pushed, the

corresponding pattern that went with it, and would control the flow of current so as to

create the desired pattern.


Figure 7-3. The raster scan test setup









The test provided to be a helpful lesson in using the parts system and setting it up.

It proved that the system would work successfully. It also showed that many of the

available opto-isolators either did not work in this type of system or just did not work at

all.

Sensor Tests

Hall Effect

With initial plans of detecting pin positions with hall-effect sensors, an effort was

made to find the best way to use the sensors. Two methods of pin sensing, based on the

fact that hall-effect sensors detect a change in the surrounding magnetic field, were

devised and tested, as seen in Figure 7-4. Five sensors were arranged on a proto plate in

a manner similar to how they would be arranged on the bottom plate of the machine.

The first and most convenient detection method involved placing a magnet on each

sensor. When a metal pin was lowered in front of the magnet, the magnetic field would

change and the sensor output would show the change had occurred. The second method

involved placing the magnets on the bottoms of the pins. When the magnet would get

near the sensor, the field change would be detected. The tests monitored how all the

sensors reacted to the pins being at various locations, and how different magnet shapes,

thicknesses, and locations affected them. A method had to be found that would have the

most reliable outputs from the detecting sensor while not interfering with the surrounding

sensors.