<%BANNER%>

Residual Strain Measurement of Plain Weave Composites Using the Cure Reference Method

Permanent Link: http://ufdc.ufl.edu/UFE0021829/00001

Material Information

Title: Residual Strain Measurement of Plain Weave Composites Using the Cure Reference Method
Physical Description: 1 online resource (83 p.)
Language: english
Creator: Strickland, Nancy M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: composites, cure, interferometry, method, moire, reference, residual, stresses, woven
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Aerospace Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Residual stresses develop in composite materials due to the high temperature manufacturing process that is required. As the structure cools from the curing temperature, those stresses develop because of a mismatch between the coefficients of thermal expansion values of the two constituent materials of the composite. If the residual stresses are not accounted for when designing a composite structure, premature failure can develop due to the lack of available strength remaining in the structure. A significant amount of work has been focused on determining these stresses for unidirectional composites, but as the woven geometries become more popular, the research must shift. Woven composites consist of a complex geometry, so because of that, a full-field measurement technique must be used to obtain the stress patterns throughout the repeating unit of the geometry. The cure reference method was designed to measure surface residual strains of unidirectional composites with very high sensitivity. That method is extended to the woven composite; however, to do so modifications are required to achieve optimal results. The final procedure chosen for this research is a less time-intensive and more repeatable process than the original method suggested for the original embodiment of the CRM. The experimental data show that significant strains develop throughout the repeating unit and the strains alternate between highly tensile to highly compressive over a very small distance. The approximate range of strains that are occurring is between -3000 microstrain and +2500 microstrain. A finite element analysis is also completed to verify the experimental method developed and to calculate the residual stresses resulting from the experimental residual strains. The verification analysis shows very good agreement between the analytical and experimental data. The tensile strains are occurring over the fiber bundles and the compressive strains are occurring in the resin rich zones. The range of strains that is occurring experimentally is larger than that predicted by the analysis but that is due to some assumptions made in the finite element model. The residual stresses that are developing are between -13 MPa and +4 MPa which are small with respect to the strength of the resin (82.7 MPa) but they cannot be ignored.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nancy M Strickland.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Ifju, Peter.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021829:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021829/00001

Material Information

Title: Residual Strain Measurement of Plain Weave Composites Using the Cure Reference Method
Physical Description: 1 online resource (83 p.)
Language: english
Creator: Strickland, Nancy M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: composites, cure, interferometry, method, moire, reference, residual, stresses, woven
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Aerospace Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Residual stresses develop in composite materials due to the high temperature manufacturing process that is required. As the structure cools from the curing temperature, those stresses develop because of a mismatch between the coefficients of thermal expansion values of the two constituent materials of the composite. If the residual stresses are not accounted for when designing a composite structure, premature failure can develop due to the lack of available strength remaining in the structure. A significant amount of work has been focused on determining these stresses for unidirectional composites, but as the woven geometries become more popular, the research must shift. Woven composites consist of a complex geometry, so because of that, a full-field measurement technique must be used to obtain the stress patterns throughout the repeating unit of the geometry. The cure reference method was designed to measure surface residual strains of unidirectional composites with very high sensitivity. That method is extended to the woven composite; however, to do so modifications are required to achieve optimal results. The final procedure chosen for this research is a less time-intensive and more repeatable process than the original method suggested for the original embodiment of the CRM. The experimental data show that significant strains develop throughout the repeating unit and the strains alternate between highly tensile to highly compressive over a very small distance. The approximate range of strains that are occurring is between -3000 microstrain and +2500 microstrain. A finite element analysis is also completed to verify the experimental method developed and to calculate the residual stresses resulting from the experimental residual strains. The verification analysis shows very good agreement between the analytical and experimental data. The tensile strains are occurring over the fiber bundles and the compressive strains are occurring in the resin rich zones. The range of strains that is occurring experimentally is larger than that predicted by the analysis but that is due to some assumptions made in the finite element model. The residual stresses that are developing are between -13 MPa and +4 MPa which are small with respect to the strength of the resin (82.7 MPa) but they cannot be ignored.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nancy M Strickland.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Ifju, Peter.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021829:00001


This item has the following downloads:


Full Text
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 E20101203_AAAAFM INGEST_TIME 2010-12-04T04:48:32Z PACKAGE UFE0021829_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 5366 DFID F20101203_AACLGX ORIGIN DEPOSITOR PATH strickland_n_Page_66thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
4b0407457efd4d65d34399f5c14d9c63
SHA-1
10db41164478d532e5dbda01b9c7525c907bdcea
246664 F20101203_AACKXB strickland_n_Page_01.jp2
de48bf38146a8541fb2754559ba19570
64727ecfcb8a0539992b16b8f3e5569f4417eee9
83895 F20101203_AACKZZ strickland_n_Page_53.jpg
999cbb6e8e83d64e00b8ea9300cc8f73
30990471c29870c091e416a5bd9df8ecacfea973
1051973 F20101203_AACKSE strickland_n_Page_74.jp2
e8615a91de558cc3c590d51734a2e985
c14aa1c094cc2a61b25bf6978fd04dd21c5c4712
5161 F20101203_AACKNH strickland_n_Page_06thm.jpg
9a9067b8030a965f93e0bc6510ba4d37
c791ff55b8b6226d424db2dc44a644169c939afe
20691 F20101203_AACLGY strickland_n_Page_69.QC.jpg
8f614df27d29b3346226ffaf6f7f3d69
643cbe8cf15e020398e66145e779595aa35d5584
2208 F20101203_AACLEA strickland_n_Page_13.txt
0512565b94efbadd3ecf7392e8f315fd
1be6728e85fa34ba95e002399a5fb9a19843e4db
83594 F20101203_AACKXC strickland_n_Page_25.jpg
5904a4bcef1187868653bfbc65689bfb
2146cfd07f819cbc1fb56f1e5ab275cff79c8d91
7111 F20101203_AACKSF strickland_n_Page_59thm.jpg
971869d729cf05f5ef1f4fcf28ea1c45
9da5741a1dd43804373b6630e90d4da880104740
2010 F20101203_AACKNI strickland_n_Page_59.txt
64fc328c822c49bf58bc997da0cb6dff
549eebf6f1e268efd8927b8a6e54555f6783fcef
3803 F20101203_AACLGZ strickland_n_Page_72thm.jpg
0754d9f58cbb5a2b7a055a393f8e8430
4ff852c69b421a861da80ff80d10aa18638f5f5f
2090 F20101203_AACLEB strickland_n_Page_15.txt
306452537c16589302b9aa2ddf333c15
4ed80cdfa3bfb00d3ab4d4eaa23ddf777a153ac0
719684 F20101203_AACKXD strickland_n_Page_10.jp2
7282850bd284f002f7dc07e8c915aad8
83b0124a3971ca3a9eca08e5d3b1a4f91517697e
847958 F20101203_AACKSG strickland_n_Page_44.jp2
fbdd60e3e163f454856fdd89cf36e682
4b275d5c96c44724c2d5c9e0cb139bef92efcbf3
25271604 F20101203_AACKNJ strickland_n_Page_15.tif
54a2eecd7dca9f51b59bbb310f92597b
83d64108d0cd5b027c51c618f7dcf43a6b52b488
118 F20101203_AACLEC strickland_n_Page_16.txt
77bb1dd40ff0891a4e52079f71176fa4
3cdf4acc136174050761056f0230223be32d656c
45235 F20101203_AACKSH strickland_n_Page_49.jpg
b4415d74c2c0190fc93a23810dfaedb0
ced1d97334fff6c9aea819cfeef9b1e51283c37d
7211 F20101203_AACKNK strickland_n_Page_25thm.jpg
400b73a71feda51abcf5c8a3cf04473c
702dfb8e2b232f2bf961cfd3a2987286e0b56322
3988 F20101203_AACKXE strickland_n_Page_12thm.jpg
d9d7a22b51e5e26d7dcc982b52f35835
ade4ca1b4ad033f1dff6652bd0bded21066167c9
26011 F20101203_AACKSI strickland_n_Page_22.QC.jpg
65cf99bee631d3ea784289115b08d15c
1c989f91c266924ccc1c6edf3dffa6fd501fa50d
28837 F20101203_AACKNL strickland_n_Page_26.QC.jpg
6ffd1fc71338993f99700eb560135597
454f352058959ad5f180264bd2417ebc0b926e5c
82183 F20101203_AACKXF strickland_n_Page_22.jpg
a7874c125c3c1d3d394352eea36f0b13
064dacb47fb4a1733a8ac743bb0b8d9f955004c0
2015 F20101203_AACLED strickland_n_Page_17.txt
657f9cc2796c41f3fdd9611f18cee729
e0c271ed2fe7427aa39ea49e27d50191f78b5e6a
6167 F20101203_AACKSJ strickland_n_Page_16.QC.jpg
80002e24ec4b57ec71f4f53f13efa54c
1dbf09fd6df9a06a171a3607dc593b278e9dabfc
989 F20101203_AACKNM strickland_n_Page_64.txt
a7c8423fca2736d93b4f0578b0e37ba4
53b03cc8063278352f12398657b5a559b8ee8745
71677 F20101203_AACKXG strickland_n_Page_06.pro
77be47d2c55a3aad0336e033b7083a7f
1cb3ba1f09a3b0235379eaf3003a77dfd2609a28
1985 F20101203_AACLEE strickland_n_Page_23.txt
c1225830a3ea8da82cd1588dff637b34
d8af2658f58b5edb31679141926642afec1d203a
85476 F20101203_AACKSK strickland_n_Page_63.jpg
876d04ee87a568ec102a2b8d58ac9805
784dac672e8badd540f37fa121b301dc6bc6845e
F20101203_AACKXH strickland_n_Page_33.tif
63189a198d7124744957d5a51b3b234b
6f5dd2ffaead0937539089b5f5b50d77d4c33c1c
2227 F20101203_AACLEF strickland_n_Page_24.txt
8b48539e7ec396f2622dd43fa68949d7
3ed3a3c45862458268c0dd1c3fbdd78ae1e7b180
56961 F20101203_AACKSL strickland_n_Page_14.pro
3ca6809e13217c757da844276cf25a70
e5a8ca6481a5a0872620dd0cdc80def4c720c9d1
73577 F20101203_AACKNN strickland_n_Page_26.pro
079b571e68ed36daed68b06302ee97a2
c5be30772f4293d525d91d089ee664a75aa1aeb9
6187 F20101203_AACKXI strickland_n_Page_52thm.jpg
8527d660652fed509cd5da49a4bd09aa
b5b77c33b5446277d098a4658207783958f85806
2144 F20101203_AACLEG strickland_n_Page_25.txt
f0eb0750669cb4ee858f4080ef940428
bede1cce02dc559082c5ec18affb1f51bf6f94dc
1288 F20101203_AACKSM strickland_n_Page_83.txt
788f8d0b01cfe10a8acbd27ba0bc61c3
1d4be0e37fa6e77d193f6b4808b06d61cbed90a4
563956 F20101203_AACKNO strickland_n_Page_64.jp2
2deb590d095f749d188eeb4e8ebaeda5
58024d1fab9686e7df263e0834bcddc717ecc8c2
1051944 F20101203_AACKXJ strickland_n_Page_55.jp2
201a5039b5e3643ae67f860950cdd7a0
e8b6885383e289ec773e62d2835469e7f769cc4c
1995 F20101203_AACLEH strickland_n_Page_27.txt
59cbac2c23cff01b738e824318f4783b
0c065d8f954e7bf2ea7937bd695cd630e541f1ce
5150 F20101203_AACKSN strickland_n_Page_82thm.jpg
b4f27d0e60f021b0de1927115a1fa866
959082705407b5d432047f190deeaddebda0b5d8
26314 F20101203_AACKNP strickland_n_Page_37.QC.jpg
bde34a8af967f1a80d63bff046e1158e
7de05a49c350d4d1d097847eddfa07672c495068
54357 F20101203_AACKXK strickland_n_Page_63.pro
3866dabc2a87dae2286f3be2dfe0bef5
c147c7ec1c06e2b97fd66bc961c90dbb2d2352bc
2368 F20101203_AACLEI strickland_n_Page_30.txt
1f990ff19d2dc673100eccc4f78261b5
89e595a5e3b0c0232c322dee274631716452f03f
1051949 F20101203_AACKSO strickland_n_Page_19.jp2
63b9c9d0d18ba425351f9f5612036510
595ba896cbd353f8e8e4c7c3b8060f0c1a28dc48
1051945 F20101203_AACKNQ strickland_n_Page_15.jp2
2099cb6748dc9ef06cc63aaf67609eee
336174c04e7f01b3704b8f90932e1bdd6ee12a66
6908 F20101203_AACKXL strickland_n_Page_27thm.jpg
be0bfc5b9f13ca9bce5012c93b3a7c79
6b104e9bc1d68f30216f23108ac55ee9badb5ea5
2473 F20101203_AACLEJ strickland_n_Page_32.txt
a9ae643214e9fa9e6b8ac2a3fb08a341
58cf19b292932ff4a0126875fc144cade92a1e60
1346 F20101203_AACKSP strickland_n_Page_02thm.jpg
f0a7e1ee9059b1ae41e034289283e7a5
bacd9a2b7273f53070b0d5f23987783475cde4e5
53969 F20101203_AACKNR strickland_n_Page_75.jpg
e54fb7e09a5f9c63943c14797ede77e6
a5a4bd959404384ebbb96e6a8968b5313e3cb509
F20101203_AACKXM strickland_n_Page_67.tif
57db1c044801055055584fb9a073958e
195de037343902ae63ef61ccc34f88308c1f66b9
2525 F20101203_AACLEK strickland_n_Page_34.txt
8e7e62544c7780eab97a7a78551a5857
eea986b47e9b3d3ffd74d6ece9935e578d74edfa
24768 F20101203_AACKSQ strickland_n_Page_59.QC.jpg
9292f9ebf16f22b94e40b9ea7a8b5eec
55f82a284f659233f1e09a8233be8a3620696aae
1051980 F20101203_AACKNS strickland_n_Page_08.jp2
9e76da7f1741ce03278e28683f046b7f
734e7334dc4e9e7687403e4ed7de03ee58c48af8
46849 F20101203_AACKXN strickland_n_Page_11.pro
965060e223c78a96602a349cc447698c
f76245a13920266c28e950ef40ebe5665f8f6ad0
2066 F20101203_AACLEL strickland_n_Page_37.txt
d65b5c2defa4f459ebcedf779ad2dafc
11785330d145888aab59a869ffbd830af8cb6b96
22425 F20101203_AACKSR strickland_n_Page_05.QC.jpg
e076c96881ce193602bbe435e562b2e2
61544ba7f0ba71e969a92ccc6996a9f85a39e094
851942 F20101203_AACKNT strickland_n_Page_82.jp2
3cd61bebad9fcc486ca0559c2deb6c59
3520e109ea160787ff0edb229f26a6feb083b27c
51271 F20101203_AACKXO strickland_n_Page_37.pro
32d8124eafb0b46957a9ffcf9fbcd359
f7bc363c8dcf4626e93149432da9c667ff1648dc
234 F20101203_AACLEM strickland_n_Page_39.txt
8f323f4ae809cc1cc4e88a87acfb59c5
0d9963c1088aaa2c7733d069e183cd07ed5e7f1a
24091 F20101203_AACKNU strickland_n_Page_17.QC.jpg
91fe19773db0182aa98fe67e4fde53d3
7ffe4e650889896c2d3d5340ab2ebd5f4103919b
469 F20101203_AACKXP strickland_n_Page_01.txt
1d979cac5b434d386f6f618c985df0ce
9d95dd415f713c31ec77f0a91295b000ebaab0e4
347 F20101203_AACLEN strickland_n_Page_41.txt
dc56bdff5fe8c25015aef3fe24baf68c
d707bef3444bdd274c1b491bd2c36d389cd109c6
21188 F20101203_AACKSS strickland_n_Page_67.QC.jpg
3a1e5272994aaaf802bfe2eee1ba39d5
830b49d9d15a381f8cba3e7fb803d89d0c6f0b88
3568 F20101203_AACKNV strickland_n_Page_07.pro
32b6da80910a55d318b117cbac5618eb
19d76edb290f497f565616053b781a35c8ac75ec
F20101203_AACKXQ strickland_n_Page_43.tif
afb74af5ba790ea615a77a2bb88d9c4f
1a7549aad03b7ef5298fc15b9382dd86d19a5f17
295 F20101203_AACLEO strickland_n_Page_42.txt
55ac47ea16dbf9dbcf9feb4611014e73
d69bcf7ecd7ba5fba40cb0e250b401bfd307949a
36976 F20101203_AACKST strickland_n_Page_82.pro
271bf688354579cda606ed632aaea0de
0b598b5baaa4c2d03b9755a2fda7fdbc841ff425
5558 F20101203_AACKNW strickland_n_Page_68.pro
aece725a823e16da55f48dc7c1098fd7
8f13b60994d8903fde9655040663836022927894
49136 F20101203_AACKXR strickland_n_Page_70.jpg
cfc6fdbb1c3ba4afcf509f5fbdd49eb7
c2617c7d0053cd5df9a7dd33ad17e7cd0d6d7701
530 F20101203_AACLEP strickland_n_Page_49.txt
5fbfe070d4bd9e1cba4959d2215caee4
b1872c0401ce205b9497329b1aabcd4be7a46057
100367 F20101203_AACKSU strickland_n_Page_16.jp2
3407f59edfbd9c3c4ffe3abaef0d7148
88a71f5fe75b02472dc7dc75de80bc941ce4d943
5553 F20101203_AACKNX strickland_n_Page_58.QC.jpg
ec36b067e475da16c9f30e099f56a8fa
ca74c58273d83c8791861177110937542b22f040
F20101203_AACKXS strickland_n_Page_68.tif
b6bfaa2b8806704081097cfdd9e650f3
307c9e8d8740ca5a6445839d410f48e63278f78c
2169 F20101203_AACLEQ strickland_n_Page_52.txt
b372f1d028ca585a7a7cf30a2b2cc8a1
2c86b62324f9a72cf25f7d6b273538812ea1273b
1051917 F20101203_AACKSV strickland_n_Page_70.jp2
58c355263d215046d7ff8d5da783ea4a
ae460a364b3a1393bda515cbdaed036dc7e013b7
69237 F20101203_AACKLA strickland_n_Page_76.jpg
62c0158024610574d85e51339e33b141
d12a7d68c47f9c24d2267396d73769b169fad19a
2148 F20101203_AACKNY strickland_n_Page_60.txt
3e3681ce98631824fe14f4205b0d756d
6299df008f6c398b92077a52e4510deadeef0fa1
27542 F20101203_AACKXT strickland_n_Page_61.QC.jpg
5d9a6a4cc16b1ff9011086e57251a84d
4d27619096a91e923113ccc8fb60b47fa7ea0da0
2207 F20101203_AACLER strickland_n_Page_53.txt
dc4ef99368f58dd4240049954a8a7d4a
4f6ddda3998ccce2ed7c7dac8493ab68295a4fee
15806 F20101203_AACKSW strickland_n_Page_56.QC.jpg
2782277d765d570f7d5010856c536b8f
322b749288e5338906d0fc9b54bbe48446aa08a9
3553 F20101203_AACKLB strickland_n_Page_40thm.jpg
00846fa71595e699942eca46fbec764c
1e404c18e168c8f2256ab89b6eefab14cc7d453e
F20101203_AACKNZ strickland_n_Page_60.tif
f721335e2a40297fe1e76ba43c9cc41f
87d328c14a8c62ee3288d8021da009b8184a0d2d
4297 F20101203_AACKXU strickland_n_Page_44thm.jpg
c3c012f362893ee8d5326b2f7fbbaacc
6b28d67c21afc6adf235bf3ddbe31a69769201cc
379 F20101203_AACLES strickland_n_Page_56.txt
261b1f58098a7b51c587ada681045ed2
345c0face9e8b777f6b3a96380d6426df1a8b6b2
2053 F20101203_AACKSX strickland_n_Page_20.txt
7aeb3f6525ac86aa5d3405807e957cce
378123a535e314f36c3874c22d9e6725e7a0533a
1050156 F20101203_AACKLC strickland_n_Page_17.jp2
14af43d386161416daa09e42981bccba
912982ecbe3301868d62c9b656c7478cf214122b
60594 F20101203_AACKXV strickland_n_Page_32.pro
c036344e7159bdba6398f4fe0200e085
625888d052f67bed711f13a42421009f4f9dfb64
2184 F20101203_AACLET strickland_n_Page_61.txt
9cb74cea9cb0df51a1a5e15ab18cbf0e
f770b7c9d26105eafec46c860ae4a41efeb8d9aa
14115 F20101203_AACKSY strickland_n_Page_64.QC.jpg
46a4dd91c7a3812bd33e186e968fb97e
17a075e01fe6061460dfc7d73f505ee0771648bc
1181 F20101203_AACKLD strickland_n_Page_79.txt
0afd6bb04f50169db27a917e377f9179
e12dd1f44fc5f9ffc4dd57680ce12e4487db072e
2143 F20101203_AACKXW strickland_n_Page_74.txt
b673407191c38ecc62321e6bc80b5ea3
85069eeec82df1b0adca756139546b0574adc860
28235 F20101203_AACKQA strickland_n_Page_02.jp2
7ec1f9712450d0734347b086dc901da2
a537f1402ac266d6cc174e27e4490352879f482a
1168 F20101203_AACLEU strickland_n_Page_66.txt
625f75a0887ebe7b4b8273a6dec79908
5f49b28881dabb5f33f3ef52c6b11b0d9e92c128
1051982 F20101203_AACKSZ strickland_n_Page_81.jp2
83d5981de6bf136bc300c6dc0f002643
d60ce14de1cbcc3eca5ca4a2ea38818cb8f7e067
17339 F20101203_AACKLE strickland_n_Page_72.pro
372cd59af7023df9d0884f76686faa6b
5ad3f94c0b76dc1855e6c57482a806fcba42946b
84806 F20101203_AACKQB strickland_n_Page_60.jpg
b80e24fab0c2cd4b3fd60ae3e7bee875
bd3301d6faa329dbe9547b9021edea86e5b7e6f1
1208 F20101203_AACLEV strickland_n_Page_69.txt
a29b513ad05c4160a99184ee0d02af8e
f6559300529919156c560c4a374ab383596a8d8d
4913 F20101203_AACKLF strickland_n_Page_83thm.jpg
03cc99154589ae70527a93118050dd1e
359144e05affa405cc2835d185152ab04603d0a2
63738 F20101203_AACKXX strickland_n_Page_82.jpg
c294a099334d429f030e92327a3825df
21c4f55fa2ededc2c9088e496d3acb8f71333b04
2000 F20101203_AACKQC strickland_n_Page_62.txt
13a55394bffdc0c701e82e18c10cbbd8
8706b46efaf75bdff1217959a865da7593f474d1
965 F20101203_AACLEW strickland_n_Page_70.txt
ba2323447be795854880345c1865e578
b350932fdb8d07e841faeeab261b0a5c5d267c21
54641 F20101203_AACKLG strickland_n_Page_24.pro
cb860068ba2a179d0d6153dde478c2dd
47016b50f91da98b9212b1fa8593e38587c51140
12976 F20101203_AACKVA strickland_n_Page_03.jpg
c41bfce2b054fd067758d1b69b14b574
7eee4ae13d9b533c646ae30ffb16a67670038d65
87508 F20101203_AACKXY strickland_n_Page_08.jpg
165762617ce19ae73865e49daa2b19be
a89a08d51e9f4b89b650047616e210f2c112de40
7398 F20101203_AACKQD strickland_n_Page_15thm.jpg
ec5da0168959d4e120c31412fa569d12
51c4eb2a8f562f21898864240cc87c8465284c37
2113 F20101203_AACLEX strickland_n_Page_73.txt
79f6e0583078e965de63c7a7d9302fd6
4ed98af076f80f85922cbd053c91304a46d769c6
2818 F20101203_AACKLH strickland_n_Page_58.pro
510e1a5f74957b6dba5874c7ffc17196
b7018bcfcd576ef5eb4ec4c7dc58d8e91d3e76b6
411 F20101203_AACKVB strickland_n_Page_50.txt
22bce08d20c6937b32ed54ab576d26d5
6401cf864ce0342e5ddec82caa17fe292ccf7b07
7416 F20101203_AACKXZ strickland_n_Page_80thm.jpg
e580f32a0c3d7d159d68e0579b3c0902
5b8ade85b7f3e6da3d31ffc22b5fe83dc13ede29
1051961 F20101203_AACKQE strickland_n_Page_32.jp2
0b6798af999a1604caa9f5ea70610d00
3984b39d9f5618b398febd7ee83b55877d29cac6
437 F20101203_AACLEY strickland_n_Page_76.txt
d2c7a6ada0ccbf9e65b2320f02dcc3f8
a929a338f6bd0b2618d49b0cfefe40dade54edf6
F20101203_AACLCA strickland_n_Page_27.tif
edc06569e327d17cb326c13d5f2544ac
a0c45896b41a8b646603370f2704d3c2c56e7a93
1051916 F20101203_AACKLI strickland_n_Page_37.jp2
58ac30396720df0d3cdb5841c8aa78e3
e64238b50a8090a8a1abee289b807d6b56e1db1f
7246 F20101203_AACKVC strickland_n_Page_60thm.jpg
2049ca54f0e1acaac22fb4f6d2f8abb7
aa320563499eeff6d8bad739a10d908e3dc60f76
1968 F20101203_AACKQF strickland_n_Page_19.txt
203b5cb99830d8fd80018c50edaf091c
19dd062887d2b05974571d585f09b1929c33f79b
6300568 F20101203_AACLEZ strickland_n.pdf
2debfdafa86b3b0f0092799934da5ccc
4fc6cd9ce806bbc07180c1c05f04db5e300f1da2
6766 F20101203_AACKLJ strickland_n_Page_17thm.jpg
3fd24d162341c2a9785a9d0c528fc40f
bcf5208f75dca8f209011a14b5b80c7ae326fedc
10330 F20101203_AACKVD strickland_n_Page_45.QC.jpg
7e39f850f730477fb4b5c0cb869195c8
17c61489e9b8407eb822677c3d82159e2d03ad6a
6747 F20101203_AACKQG strickland_n_Page_19thm.jpg
3a46bcfa2de5510beb4202da28dc0c4a
b72371c80b7fc2c0c34236bf792a269619a23d62
F20101203_AACLCB strickland_n_Page_36.tif
d436be2c25a589b5944d28480d98921c
80479e772870d6467931ef0b15c4288ad7ff63d5
F20101203_AACKLK strickland_n_Page_27.jp2
1a52c8b5a5cea36c09dc06ce56a03ec0
8e078a78904db1e46660c6ce9f50e4f17a81ff68
15647 F20101203_AACKVE strickland_n_Page_58.jpg
258f02ea145e0ae032b9af329d362884
430d260b7e8806a45add28b1bb65f6807658cd30
F20101203_AACKQH strickland_n_Page_10.tif
16a630437a251209ab13ea8a5c2dbe8d
f3d5aad66c58c7a5b574b358117ad939cd19e693
16880 F20101203_AACLHA strickland_n_Page_75.QC.jpg
347f63e61e1f7ee816d621797d07ea52
8199e9ba5eb1fa5f12da134dab264ac50b401f70
F20101203_AACLCC strickland_n_Page_45.tif
3db6ceb5eadf2c5236da6bffb611dd5f
995a42034805f7313c0ec19c51f875a29486bb10
216 F20101203_AACKVF strickland_n_Page_40.txt
bee5d5a6caf2741fda2e5d9c02a44ab2
a60513fae0f02343922793528c8da307b271ec1f
26730 F20101203_AACKQI strickland_n_Page_29.QC.jpg
cf0740f2a215ab0199494fc336ef0f63
e29cfc4d76fcfcd22eaaff9530de80f435ad38d8
6533 F20101203_AACLHB strickland_n_Page_76thm.jpg
54b601dc1c6135e366e7d1af2ba2e4f0
a00903072e89e6f872403a37f3d38899ad059cd4
F20101203_AACLCD strickland_n_Page_46.tif
2e0a19327b31eea984173febd1342041
917206f71a111af3f60d454558dcb5cbc47816b7
1010228 F20101203_AACKLL strickland_n_Page_42.jp2
0b74fc83d4e020b3a584d43895945057
94c69a5476faf56f0ab8445b05ddcc01337434d9
59440 F20101203_AACKVG strickland_n_Page_80.pro
6fcd452e20c2812f64b02f1e005bedc6
3237ce290cd9ca7924514aa37761de5fa8ca26a0
6970 F20101203_AACKQJ strickland_n_Page_36thm.jpg
49a3b531f09fe92b6edeb865e19335f6
50b7bbd7b8c72ba06fa0991f165e9019e63d1a6a
17909 F20101203_AACLHC strickland_n_Page_78.QC.jpg
9d5c681e3e25728f4e5ec5913db175f9
62c0ffb6b7af7a1f02b0634e9202b1dee9637eb8
F20101203_AACLCE strickland_n_Page_47.tif
d698b2f33f666d56113698f1876ad20e
507f6ea05b8d4b83d22ca6acb9a389580a2120ff
578 F20101203_AACKLM strickland_n_Page_38.txt
be5b3dd3d3c63515f63a5696e129063c
74baa284fb34c7fbfc4980013f2e6de935e642d2
355 F20101203_AACKVH strickland_n_Page_46.txt
88132f18c6a1512ecc50b62a6b3eb75c
69f7be6e76408db1d4cd81c2aa68761247b9ae9e
2445 F20101203_AACKQK strickland_n_Page_81.txt
da7dc4c22d5d286190cf9608df21d98e
4d9c1cbf828c7f7755b09dcfaebcf3618cd1d24b
5287 F20101203_AACLHD strickland_n_Page_78thm.jpg
0d0e69b99e93639978c548509ca5332b
12da4910ca9a3c040e20976e6ce116a5d3ec527d
F20101203_AACLCF strickland_n_Page_49.tif
9cdc045593cd3e1bfbe3d393b0f9d3ad
1a0eed7aa795cfaa4de8ce8e11b41deec4533466
589 F20101203_AACKLN strickland_n_Page_55.txt
f61e2dbc617faab3b79d000e3b7426e5
e3ba48bdd044d579e5ce6cd1f6554e749169bef9
2189 F20101203_AACKVI strickland_n_Page_33.txt
f2580a8bb1ea318fbc0d3b6b94074f15
75be98deef943bdae1d07d445033985e68877c79
412029 F20101203_AACKQL strickland_n_Page_04.jp2
c16dff7cf9a8b5e78a1179d7591d2b62
f00bd9d74ab527dc314034d692326a4e4ba9b4f6
16313 F20101203_AACLHE strickland_n_Page_79.QC.jpg
d611881695636e015649a6551372782a
4b975e09a41a653ab2755697ece60d958c930940
F20101203_AACLCG strickland_n_Page_54.tif
4e4123c2446d293df2043b7401d3d414
60278e58719d2f56f30e32e8810b92e96c066d72
30682 F20101203_AACKLO strickland_n_Page_10.jpg
5ecbdd1cea409548f39f788583cd85e4
930f41cefe126fa4dcbdc84f249908aa03e2e454
F20101203_AACKVJ strickland_n_Page_04.tif
8f6e82d5204938bf179e80147709dd29
32548503e86c13f8ea86df52cc7a7081396ef38f
42776 F20101203_AACKQM strickland_n_Page_12.jpg
a603aa7dd8ac0cc589b7663329be13b9
0d4de81dbe7479403cba2e722af669645235c320
4940 F20101203_AACLHF strickland_n_Page_79thm.jpg
422df791f5a38ba00cd32d03d5aa3169
02d113193b0a0c93b67cd34c71aed82a09e0132d
F20101203_AACLCH strickland_n_Page_56.tif
a3428b5ac5d565f0f9befa118f8758b3
5e9528e1b18bb6570dc84905af2d249821c0d285
37496 F20101203_AACKLP strickland_n_Page_51.pro
d254e5d70c599f77418f2ca76edeb259
0dc86b9a899da474145a572fa60ad67aa8532818
1051939 F20101203_AACKVK strickland_n_Page_53.jp2
dbd42493ea981de6c306ab286f617122
888850e81f8af87daeae92755c50bd80135fbc3f
27498 F20101203_AACKQN strickland_n_Page_74.QC.jpg
265bb5d995b9e94d0512db9eddc052fb
cf4a8b68fdda04a2c7d8604f27603abf643cddcd
F20101203_AACLCI strickland_n_Page_58.tif
618c127949de5129f0f59e62a4677028
d5a46297f5cea349325fc88868fb8235be606d56
25384 F20101203_AACKLQ strickland_n_Page_24.QC.jpg
ff0a1f629bc3fe80f1ae2a56de128b43
8a8bd516a5b02ee8f022bd21125beedf1fe83b76
F20101203_AACKVL strickland_n_Page_69.tif
4a028358d7d6ed2f7ebf20ae9fb59aa3
815fbac7aa1f39f61717121a599c80c1a2c9e78c
1051964 F20101203_AACKQO strickland_n_Page_69.jp2
f33eba01796d58854bd7c975879d2f0d
c35d8ec7f2db30065ffb89beac055d8c6fc36c9d
26924 F20101203_AACLHG strickland_n_Page_80.QC.jpg
bb3606e53415d451f3d8f00144a9066b
1e3800816a9c6f63d28d10f11834abdde1d068fd
F20101203_AACLCJ strickland_n_Page_64.tif
3fa9e39b646dd0da408d7a6f5d223341
d98e4a4eb1353d67ac58cf25ad3e668ac7355b94
10347 F20101203_AACKLR strickland_n_Page_41.QC.jpg
6a1be3e796bca9f7b1104a289cce8cff
25c4d20743ba457db8174ea60760de25340ea497
F20101203_AACKVM strickland_n_Page_48.tif
b0570117dcc3bfc01f4cdd700929cc1b
9d204eb19ace4a695b6af6a9129cd52c22006ffe
1051986 F20101203_AACKQP strickland_n_Page_46.jp2
fc24d56e69c9276b9cd74046e2fabcaf
accd33c4da323e5e50ad9f5c5e03609d47910c52
27333 F20101203_AACLHH strickland_n_Page_81.QC.jpg
3d8dd1a153f13e23c11020c71a0ca008
538e87bbd78785812c17bc041bb5d96de37b31fb
F20101203_AACLCK strickland_n_Page_65.tif
53c0a200334a028543b656adb779e85e
6077146e32b8a9829c051971fe236d5fb41b5440
2347 F20101203_AACKLS strickland_n_Page_80.txt
8e67ae987740b1f024fa1719449a5277
6dae4fa296e22901a3bafcf6ac34c770b056f475
968 F20101203_AACKVN strickland_n_Page_02.pro
f7600db12eb6495f594f2175433d96a1
7a1def38e7cda487809339665f6df59515fe4667
17580 F20101203_AACLHI strickland_n_Page_83.QC.jpg
3373ac122bc313ae69dea592b232ff60
3ecffd31134b7272e2ab063dff791a383c9f0966
F20101203_AACLCL strickland_n_Page_66.tif
2d685e9ee807bed615bbed1ff4767fbd
db228ca5ca6b0374623c48142bbc4e24df61f86c
2934 F20101203_AACKVO strickland_n_Page_06.txt
e9fd062271bcde518ad8f8b7b18b3e87
948f3eb3eecf73dd598a93fdeba2469f2140ca1f
5677 F20101203_AACKQQ strickland_n_Page_47thm.jpg
6d00fc80c3a2afebf10fefaf0af5cf04
a6a29740df510c2f8b8e0a8e243995c6ac7eef6d
81673 F20101203_AACKLT strickland_n_Page_15.jpg
fcaa82eef8c638ec137af55f740d2c67
ea25cf5e7986a1aa2ed5cb6e4a8c93b44e43235c
98686 F20101203_AACLHJ UFE0021829_00001.mets FULL
ff47d365a0528fec67de830664ae970d
f84a196fba06f31a99e75daac066ff93f8bbde6d
F20101203_AACLCM strickland_n_Page_71.tif
d9183ab02f9bacefbb82ddeffed3d415
ca452cf43a2165c2ee0285926587821436e6cce4
F20101203_AACKVP strickland_n_Page_44.txt
dcff9ae63a56b0a1badfefb3b509c7ed
9c50e0ed250354d9a48988728f7ae18b115e7eee
29175 F20101203_AACKQR strickland_n_Page_34.QC.jpg
71b72dc21b2564557dd69f87c40010b5
40bd3350883100e1dca61598279d6551d69c4777
511 F20101203_AACKLU strickland_n_Page_21.txt
57cee2d5b6b0ad528249e38f4a6fdde5
3e5538b086223886b624d1a6396b8d3b43090849
F20101203_AACLCN strickland_n_Page_72.tif
7dbd4201ddcbb3c2c43ba813c833863c
c7acbc499d6efbb14820c8740968cdbbea044072
53390 F20101203_AACKVQ strickland_n_Page_74.pro
523c81400f6083d9b7833b8bbc842445
b0bdf6ed193acc0b07e90221b92f89a1eb8a38b4
78122 F20101203_AACKQS strickland_n_Page_19.jpg
adb4269f95636f5901695205c55e3bc5
9a1f2a5d22daf187346fda36b0dd3f6497d129e7
12848 F20101203_AACKLV strickland_n_Page_70.QC.jpg
6fa3b7e990d2b2d471251ca810cc556d
75ad167363b2fafe7e1ad901b7b39cb0f7661bfc
F20101203_AACLCO strickland_n_Page_73.tif
1536036ddc6a4c8ae3250d810729a731
2bfeeefda4ed2163df4f784724906a0f12798f51
6771 F20101203_AACKVR strickland_n_Page_69thm.jpg
86b934d6ed6b353e9033fe483086af13
950a0ec5f77124acd9d0bc495adcfd423d31919b
49764 F20101203_AACKQT strickland_n_Page_19.pro
f04967dc65fe101c0f37027051bad030
d6c9fc8628378459251bc8f8e261b87aa5d61f9f
636083 F20101203_AACKLW strickland_n_Page_41.jp2
93ac1baa30583472b0131359ac7b7e81
c09248b7ca8ef037565fbf3f85063674f8226323
F20101203_AACLCP strickland_n_Page_78.tif
0f5c902e1f15b44ebb3210b11c399d63
862197fef6b4904c01c4bd1cc4f9f67bf5659ae8
727885 F20101203_AACKVS strickland_n_Page_83.jp2
9787f03b62faf792292542ee06aa4bb7
60983f8b58e3b0e65ee0e3792dc7d8fb8de07e6f
1051914 F20101203_AACKQU strickland_n_Page_67.jp2
39d263470f5f2a5d62fff37a6212d922
d976169274765eff8a0481477b3ea916340b0e0b
12031 F20101203_AACKLX strickland_n_Page_54.QC.jpg
3fefe4a31500bfe7900b02d493d4cbe5
71bd9227630abe43646030de76cd8ce99221b836
F20101203_AACLCQ strickland_n_Page_79.tif
077577bb923ec7abf97626ac82a81698
00cdf53e5676a425e5643f5688c6c524ee0f3f8d
152 F20101203_AACKVT strickland_n_Page_07.txt
8e9872440d4b7a2c8fa8ee97e3168b47
67949ea0f8d61b124000df05012e39087a1492d0
F20101203_AACKQV strickland_n_Page_44.tif
70118175721a9b2e84d5b3fea6a334a8
490772b1a543bcd4b79db0c33e55ebf0ebbd300a
F20101203_AACKLY strickland_n_Page_23.tif
752942c28bc1230852447a22739ec7c2
81e62492225a894996effcd4aacec8ddc7cc47a5
F20101203_AACLCR strickland_n_Page_82.tif
b2d029ddcab74fa995d303a71bf716bd
77f429f2f7b153de33f97a8881c7428fab9f7b2e
F20101203_AACKVU strickland_n_Page_42.tif
0dc402111fa983f313c9c4a02a6a73ed
54d80e15bc7fccceb132958c8d51f9bc6a9f0033
19628 F20101203_AACKQW strickland_n_Page_47.QC.jpg
3965185ab1318cbcea88d050f6d82678
c16d599e34d72641e6bfa2a8c36f035f148f7304
F20101203_AACKLZ strickland_n_Page_32.tif
0f426a813bc885b8869d1392ac4c541a
fcbcbb8aca96b989ea6672a4b5d9df5e80ccbf3b
8045 F20101203_AACLCS strickland_n_Page_01.pro
ff1e55088e918a2c31c54060f89a2969
63ee480f59bde4302e88e929912e4d165d68a8bb
17245 F20101203_AACKQX strickland_n_Page_57.QC.jpg
69cdfb44f44a8704025e1dee817e1c1b
be748eb86d7cfc0cdb51f3b37d5b8ef296299abd
3622 F20101203_AACLCT strickland_n_Page_03.pro
fbe205b63ea4d9b9ad18fc863dbfd7d7
edb46eddb827e8fa8ab1cc2455e4091126edf935
7458 F20101203_AACKVV strickland_n_Page_61thm.jpg
3a85ed89b4b011bf615bcf556d1654e7
d9e2111caebf0fa1cc5e48ed52bb580205254240
9147 F20101203_AACKOA strickland_n_Page_10.QC.jpg
93e68a86582ebefd6068c1e12127fbb7
d15f1ae378b26bc183a438082adcf8a1611bcb1d
4011 F20101203_AACKQY strickland_n_Page_70thm.jpg
812edee14fe4bfc8f44f727474efdd49
6778503b73b29eb5518c81f9095e8f0176e9a749
17804 F20101203_AACLCU strickland_n_Page_04.pro
d257949336016c2435f35dc8ba009ffc
552b611e37f4cf45723a972fef91aedb3ddd25d5
30363 F20101203_AACKVW strickland_n_Page_41.jpg
477024897d418edab5d7f0cecc1b4822
08aa2a9f25386bff79e4d86fffe82416d7d55069
7539 F20101203_AACKOB strickland_n_Page_63thm.jpg
73be283df3c8470a76c0ed5d1ce29640
6455f67e23528605c90b38eba24b6b2967dca682
842 F20101203_AACKQZ strickland_n_Page_71.txt
37f1edde1dbc73c49c33d84bd1907fd1
46ce84baafd799c1d655924b1be95e5e8321b9ba
94512 F20101203_AACLCV strickland_n_Page_05.pro
c0c64df4df569a485b4e864f91a8aced
1de303e155f1226c0e934e1d61d12c4ddefb5704
86632 F20101203_AACKVX strickland_n_Page_61.jpg
ffe0685a0bcbd6f55cc184dc825a7c7d
f9a87a210f46fba570276f345b5fa91b5272a588
5419 F20101203_AACKOC strickland_n_Page_68.QC.jpg
fda309722dc339b636a6d430db51f8a4
f94d85388fb7a653603c6e658ff707c07f4c3c8a
71659 F20101203_AACLCW strickland_n_Page_08.pro
78a5786c9dd554400f18ebdbc76249b8
30696add5929035f156a9776ad1f9a83b0fddc6c
1051935 F20101203_AACKVY strickland_n_Page_80.jp2
65b2301d9bdd4375e0be5954fb6b4fc5
1a2d9928f2f1e25278201cc5b457aad23a9d9c64
2301 F20101203_AACKOD strickland_n_Page_51.txt
a55bb30dc10770518634665b2733ce16
aedc396a1f7f0a1d7e64ae36dca835f35fdfef11
17496 F20101203_AACKTA strickland_n_Page_16.jpg
363cbe6b4054668d304675dae1f83a8e
17b0ee572dd43c8a60d614ced5a2e4eef7d96137
69979 F20101203_AACLCX strickland_n_Page_09.pro
998af62d030f72b0755b9a52809a8097
e00bf1fbd9719f9b816ca9000b7e307bc0e681cf
26219 F20101203_AACKVZ strickland_n_Page_15.QC.jpg
9608a927b77973a2c4cb33843aecb424
31b826a8c403787da892e07b797641d6cfc8b244
F20101203_AACKOE strickland_n_Page_81.tif
6968c0fa41603afca21d5ab5fa839a1d
941c37f72841ec83b109b1ec1cb969fbb1ff2e7c
25652 F20101203_AACKTB strickland_n_Page_35.QC.jpg
62b6749308b06e1403ed75f295136e2c
fd8a80c1df37b1140918fccfd062e1a1fa956430
24164 F20101203_AACLCY strickland_n_Page_12.pro
701ea161b2bcd4a5424996bef4244054
a4a0a545457d72caffc073a9b10debb1044405a5
1051976 F20101203_AACKOF strickland_n_Page_34.jp2
614171dc095478f31226933f35f3431d
e2e35c4f6e656a9e6ae8ed1e522fb97f703dd9c8
568 F20101203_AACKTC strickland_n_Page_47.txt
0d97b1db6e180e8911d4542610ceeab8
9a24f1936acf0fb7c534fc3914cce3560600386c
53221 F20101203_AACLCZ strickland_n_Page_13.pro
a1896b100f841371bc4e8c0d0946a0d4
a90596bce26aec145d77eaa2f5159273d965e264
7470 F20101203_AACKOG strickland_n_Page_34thm.jpg
934cedb861cb6f278147b811398a4273
e7799427dbaa0d27610326b11547a3bff7b0fe87
36840 F20101203_AACKYA strickland_n_Page_71.jpg
c182106322963d8582cb5d8071550e15
ded03c7ad905c8d9002e03e356122c932455ceb7
38865 F20101203_AACLAA strickland_n_Page_54.jpg
964e98aaaa94283d29b0f18c22d2a99a
23818c3f972ebe2c4392b3b9a90f54adc764c8f2
208865 F20101203_AACKTD strickland_n_Page_68.jp2
ac98bb261c61d2bc1a9c3bb5c30f8121
9a4fdf3a6a4947202d5e2bc1e32cd0988709b8f0
F20101203_AACKOH strickland_n_Page_62.tif
63fc306042709adc142db5424a70ae7f
508910620f5a8dc32f41b29c287f4bd158079498
1051953 F20101203_AACKYB strickland_n_Page_26.jp2
061444d83260b0e3d94cd1ae2db1fbc1
6898d7aa95db99c3979d976c7fb771bfe03bbd97
57661 F20101203_AACLAB strickland_n_Page_55.jpg
1b0571f3e7e8280ca3bb4b41bb35ce5b
2e75155ac4690dc2f4b61844458c392350d39233
F20101203_AACKTE strickland_n_Page_07.tif
11a81a886b1a262072864b28a9e4443a
a558d4ce85397871536445ebad99ffca4a10d018
7694 F20101203_AACLFA strickland_n_Page_01.QC.jpg
e2cd29d4a4d9cb4c5d12310dd03b7331
041e0f41e7f0ca7e4344adf78c2770f8ebabfc4a
F20101203_AACKOI strickland_n_Page_13.tif
ca736cacd2b6285eddd2de1a419a14ae
6a46964131df0e932fd4f3a9f1583e85cb891478
1051970 F20101203_AACKYC strickland_n_Page_79.jp2
3703fbea5b23c45dce7dd5ec422771ed
09dc746fcad395c17d204f73751ad792f00be4b6
52669 F20101203_AACLAC strickland_n_Page_56.jpg
ba43d7c2b8c8156e7a4eb0fb6f19c758
9bc8ca0edb05f671e9aaaea237b868a9c07d99af
3499 F20101203_AACKTF strickland_n_Page_45thm.jpg
19ffaf2c7738c889b710062311c7f791
e6617f6890d2bf547f44f72805d73ffca279c339
1774 F20101203_AACLFB strickland_n_Page_03thm.jpg
65bc39a457819ffffd633461a5e1ad4a
7de160d6f433ef2b226b62ddea1cf582cb3ecccb
F20101203_AACKOJ strickland_n_Page_37.tif
a1891291dd5e427006d2935f2f60cb87
4327f9cc48955d32d4d5a10726b782c1975255c3
7447 F20101203_AACKYD strickland_n_Page_30thm.jpg
5188ea604001b04ef28c250f1e88b4a8
68a3694a52212a399ad50a11ce52810651783414
44358 F20101203_AACLAD strickland_n_Page_64.jpg
877b4f20e0da27d85c4db81e7f674752
8e5463ef19a4ee13b3f8aa583849553b310569c0
14608 F20101203_AACKTG strickland_n_Page_46.QC.jpg
408b328b94e2a92e85273bf1504e0ff8
07992b296e719ef9027f60048edf1a8ac02fc17e
10652 F20101203_AACLFC strickland_n_Page_04.QC.jpg
002f54fadfcf7e23a4233db66eea983a
c648759b9932546e2d13c55c991e03598e0c5012
26148 F20101203_AACKOK strickland_n_Page_30.QC.jpg
858ffa1bb3b4a69b891f8ff6b03b47a9
07cfbac39addd350d19c8ec833d6210847a7ed12
F20101203_AACKYE strickland_n_Page_20.tif
e1a6d6c2b53ad41e2c44ab309f7e39f9
b85d9b6f1fd2bc341d784e6ab3d096221589e556
60460 F20101203_AACLAE strickland_n_Page_65.jpg
f1c6fb1331adbed97e8f84e2481224a2
3a136ed3ce3d5b4539a16f4c586a170103b7a856
59334 F20101203_AACKTH strickland_n_Page_30.pro
ca3a32d004f935bdae43abf5e7be65b5
37bf76cc28f8d89f0a44e39673716bb77a56bf89
5602 F20101203_AACLFD strickland_n_Page_05thm.jpg
e87b2cd430fb2d2df45bb644610b33a7
f5f503b619d881673e68e65a4276e4f9a883a1b6
3726 F20101203_AACKOL strickland_n_Page_54thm.jpg
d0706c2a2b28528692301810c72e76f6
db3c5996363809e4939cc0465f67cca2edb3be3d
F20101203_AACKYF strickland_n_Page_51.tif
44f524650addfa426b4df13e918de277
2a77ce0894099d74e5abe4d1bfbc5507b7ab59c8
17930 F20101203_AACLAF strickland_n_Page_68.jpg
f06ff6548861c9bd215a8b9705290685
eb76407a54f87ccb0d3fd2d1b1d6c94ef7282bd9
2051 F20101203_AACKTI strickland_n_Page_28.txt
a0309f0da825046affa13f17e1033acf
56a872e246046e87bac0df80b95394b4b07d88ce
47424 F20101203_AACKOM strickland_n_Page_57.jpg
1fca48777b660a23075d517a36ba2f23
4e549984c8c85f7d49f5dda90b0be472bf0158e1
158220 F20101203_AACKYG strickland_n_Page_07.jp2
7ae2cd34acd899f90b9f333a3b9f3c2d
3a1d34c69d9c7557988feb986de714b90894d536
61704 F20101203_AACLAG strickland_n_Page_69.jpg
d6e50a88749fa7349baa9c5d3977133f
59f634d98c51b13aacd5082c520d9a031e37b894
2043 F20101203_AACKTJ strickland_n_Page_11.txt
437574975442aa92dfc5312564b9a14f
4e75c28e3510f2225c5dbb1ad4cd8671c37c1abf
19429 F20101203_AACLFE strickland_n_Page_06.QC.jpg
87a476fac4a6040af89654429d9720cb
7334d15d42ab30c0f1ba1c97203fc85ef3df7b74
10822 F20101203_AACKON strickland_n_Page_40.QC.jpg
4047bd966e0fec18f6a517a3bdac232d
a830720c54bd70cbdba57363b95a90e236d3b43a
F20101203_AACKYH strickland_n_Page_55.tif
455174fd56081649645e59114f4a98e1
e4be41f0ef9fa34b44ff57ef5b8a10901fc3a104
39394 F20101203_AACLAH strickland_n_Page_72.jpg
31d5060eafca5a8c3f3850124864c68b
4179b5f5835aa58f9dbdb04c96d493f75346d4d6
78712 F20101203_AACKTK strickland_n_Page_11.jpg
3685a9ea4a542ae249f7a62bed3a6e2a
a4ace6bb88f101b5a0ea78dc8dc57315d6e87b48
1781 F20101203_AACLFF strickland_n_Page_07thm.jpg
dd334fdc72c296f1d3e1f9ffa36d0728
b383ffd0ab8e06c03c49c5d688bb901b2476fb10
84422 F20101203_AACLAI strickland_n_Page_74.jpg
6fb89ecfc6510754c9b72ccfd815a4c1
b9aff365d9d698d75dbb36fa2913a907fb9d5cb8
2047 F20101203_AACKYI strickland_n_Page_35.txt
ea45bb711ed62de4abcb07fbc9aabc85
41dadf89169bdea3e7c863bd345dcd0c1839676a
6212 F20101203_AACKTL strickland_n_Page_41.pro
54c6e16e1df616939444b8932a1e4122
1accd4cf506fbf82ae1f4f655c363717f0de43db
2885 F20101203_AACLFG strickland_n_Page_10thm.jpg
64a30b78cef65d449907c8b8b2305293
d362eb54df4b84533646555dc7608f1058dfbd3b
57149 F20101203_AACLAJ strickland_n_Page_78.jpg
e4803d1991477dc01ca8a19ba02f81d8
865334984c3db3cc9eca2827f73566de9e802932
F20101203_AACKOO strickland_n_Page_01.tif
984300ab7ec1f1c943811ad971e330aa
74ff43456d23133386b4feba9368357644a31d60
80746 F20101203_AACKYJ strickland_n_Page_37.jpg
3a24e1a4bf4baf5e11735dd4d3c996ff
43868380e42816662322bc39283cd188e1779696
11694 F20101203_AACKTM strickland_n_Page_55.pro
893a5e73cc75e53e62872eae97acb984
b5d601f7197b9bded7d6f2f2fd98371e73ce543f
6894 F20101203_AACLFH strickland_n_Page_11thm.jpg
f3ac91f69ad4a4b641bc987d73343026
53503c415fab3ed1e3fb746bc3be55110bef9a46
55300 F20101203_AACLAK strickland_n_Page_83.jpg
10c4aa1b5c8c6ff9c4628898c17c07bd
4c546448d22f33115d03c57450290efe8970d37c
F20101203_AACKYK strickland_n_Page_39.tif
38eaab1ba2810857548320daa8e418ee
df4e643529963794bd14f1aef058c927ac387d76
F20101203_AACKTN strickland_n_Page_24.tif
71daaedfe33b98812baa7a066b4f5fe5
60f670c0076417eb47240e1c9d2b1a436024b429
71407 F20101203_AACKOP strickland_n_Page_47.jpg
b2cb918838e608978afd330ce46a05cd
1706dab76a543e48d79fca4c6568dbc8d33cbc0d
26403 F20101203_AACLFI strickland_n_Page_13.QC.jpg
ebdadc30f8fe6863442b21ac8ee98aeb
f401ea2eaa519594455a7d952adf0ab5d6357ccd
1051983 F20101203_AACLAL strickland_n_Page_05.jp2
a8e8c1f7a49427894ca5bc68fa0855ad
803d9633d2d598a3f79d341a6f2ca7789b95976e
828966 F20101203_AACKYL strickland_n_Page_51.jp2
106bb78a412696fe73c4904b320e5d8a
3f4c7c5c724614101b6331b6efb57a208fb99014
7187 F20101203_AACKTO strickland_n_Page_74thm.jpg
6a2605aacd1d78c4d3b7545d2655f0ca
323dc9610688b2d096859097be4a531496121e91
6291 F20101203_AACKOQ strickland_n_Page_08thm.jpg
727c3e81ac38bc5577cb5338a40038d5
a5d19d582cd38e3209fd50899a45b29228f14bac
7189 F20101203_AACLFJ strickland_n_Page_13thm.jpg
84810683e1076e3dcfba3ac0be251cbb
5b8fdec40b0490628278d81a4671e76d209d5f98
1051975 F20101203_AACLAM strickland_n_Page_06.jp2
efa9c95449571db881a303a7380b076a
17b57515b6727508b2c3cfb1ac7d5b2354161d41
F20101203_AACKYM strickland_n_Page_17.tif
de2f6968d1e042622fc01f29a9a26d76
eaee6b44f43bf0d1799c7aa4794c6d0b921c77db
80846 F20101203_AACKTP strickland_n_Page_73.jpg
14fe8a97de20ccaba82ffb8c1cc6fab3
bd4a391d1e61889b54baf665bd2cfafd8b937200
2123 F20101203_AACKOR strickland_n_Page_29.txt
a4c7a6eb3a0653b6e133380d90431251
bd29b81909cd4de98ff2d904d55b4beac327c554
27690 F20101203_AACLFK strickland_n_Page_14.QC.jpg
13b402cb72ee38f364c9a49df5bbbd95
f5ab98a098498b2035295d16fb847e5c884958d0
F20101203_AACLAN strickland_n_Page_11.jp2
aa5a1b8086073574e2b9d0227cc95294
e3437d20c1bf985f63922748ad9d671e8395a8fd
2284 F20101203_AACKYN strickland_n_Page_31.txt
abaf9d53050a109ffdfd7a470bc9bdac
d86b1dbdea1580535945da6bb02e32f290ad0b4c
7403 F20101203_AACKTQ strickland_n_Page_26thm.jpg
75b66d590750f0954400b642d39309d3
436b9bd418aa71aa3f53cf80024c0e406234664d
F20101203_AACKOS strickland_n_Page_63.tif
6dde452e732eed7b2fe6361449439eee
dda8b6fde954eff422e9e6205ad6e0d6e4ae400f
24579 F20101203_AACLFL strickland_n_Page_18.QC.jpg
18cfb1739bdda5d2c2cd677ddd051d1b
87dc7d3f0c37ae8282cf1b146df6bdcf16277b81
F20101203_AACLAO strickland_n_Page_18.jp2
312b44322304e714f061da36712cbed4
13df58c80ccb0b7f1b2db2f7c7a1f80e9a835f16
3462 F20101203_AACKYO strickland_n_Page_04thm.jpg
ccb9a3ed9bd4a997c1eb09719a307f31
61717372f79d98da7a3307c080aa044106d618ae
6276 F20101203_AACKTR strickland_n_Page_56.pro
5a6f8e1489608ad4946e01bcc4276adb
1ca9057dd5ee1ff802d11c85e208a30861acef7f
50409 F20101203_AACKOT strickland_n_Page_27.pro
223afec076688f9d782fcc5dacd66e54
b940579c3b2b71aa3ba6f324c93c0f9b67c4bc5a
6659 F20101203_AACLFM strickland_n_Page_18thm.jpg
ba31691b81451b00d515b0f2a8cc206a
72ace069e199d8e0998b4ec451c66d43c5235646
299090 F20101203_AACLAP strickland_n_Page_21.jp2
cd674988acfdf56eacb9d90866d4688b
9848d2ff7daafdc38e974aa34c2a3da4c721ad5c
23705 F20101203_AACKYP strickland_n_Page_11.QC.jpg
2a320f3a25292955c6cac2f7f9922b3c
3b31f6c768406090fdda4a625ac0b72e48d1eca1
84195 F20101203_AACKTS strickland_n_Page_29.jpg
4075d86fced32e3d33bcf38415996a1f
2284d90fb21f81e756b783382746e765982fe4ab
F20101203_AACKOU strickland_n_Page_80.tif
c4765c7adde73174046510d5fc542b1c
9a86c90838dc038f4336407560e5c6e039764ce6
24246 F20101203_AACLFN strickland_n_Page_19.QC.jpg
34c8b0b72094d4f8320fe4bdaf78b89d
6967d548c9bc1713d97bfd323ea03db44910857b
1051889 F20101203_AACLAQ strickland_n_Page_22.jp2
b502d80221058ded247d8eac488cafff
76ab340a1d507d8c2e27ea64c73e7dd84f7384ff
F20101203_AACKYQ strickland_n_Page_52.tif
4bcd92fb1637c2420cc81ade307093d7
f6d6ab6c0824ca2accfcb97e04abe2093dd9c1f7
7271 F20101203_AACKOV strickland_n_Page_37thm.jpg
25ad7d3ac9c681498b474de445ab49f9
7531f8058bb5284eae8e426042543c60004c3c41
2655 F20101203_AACLFO strickland_n_Page_21thm.jpg
844f405bcc1b60f176c7aa2af79fd0a6
4d6808f328f5da210cf086f6a6492cc6ce57e886
F20101203_AACLAR strickland_n_Page_23.jp2
6a41c1215cb3ff0318937968ec547a9c
2471541aad37a73733360cc51551ca7d43cda389
61360 F20101203_AACKYR strickland_n_Page_66.jpg
1b65827ddc7a19ef86559d57368f290e
62b95e042eb7ec9f7beec822b53e067a708c2d3d
1051984 F20101203_AACKTT strickland_n_Page_25.jp2
7b28851f3240860c39b25feaef074325
edaaf1c6d3aeb14ce9d791e43d83729a72fd79fb
F20101203_AACKOW strickland_n_Page_34.tif
2c4ef330f5efbf157219fcb2d54ededc
388e78eed9be348f912d782704962cf190b922e7
7148 F20101203_AACLFP strickland_n_Page_22thm.jpg
189eac4a11d6094c6ee3e65cb155817b
aa0308c41980664596323432b116e2fe57206fcd
1051974 F20101203_AACLAS strickland_n_Page_28.jp2
0843f3d6bff9d478a61fa7535568267c
23f3ccbbe47ccbf963b8de89d97138b36d316c92
235 F20101203_AACKYS strickland_n_Page_58.txt
02f7d3aeb481b4e2cd74ba0505848541
f663560c66fd4e37e3dc588a5c7b9bda619e3598
12225 F20101203_AACKTU strickland_n_Page_72.QC.jpg
d1b081568d15f783b3406699a5c3d414
de229fd0e8d1d13bff11c033616a7e1dbec0842e
52844 F20101203_AACKOX strickland_n_Page_25.pro
f99ec43b1c6c4ce76268de10190629ca
16cbaceff76d388a44e445a4608c1dcbf28796b0
24324 F20101203_AACLFQ strickland_n_Page_23.QC.jpg
98a7ab29b002fac28203adbb7b8d9e11
388b3799bacf1ceb1820f726229efa1feae8f529
1051981 F20101203_AACLAT strickland_n_Page_29.jp2
df1d6aa408206ed4fd6d7a6bda5a90f2
dd872790802675d36dff0bc77f0ca5ffa9c3408c
13674 F20101203_AACKYT strickland_n_Page_77.pro
6f529669ad317612d6e2fc7981dd54e2
8858218fe984111c4c9c696be5c820999d19525b
4252 F20101203_AACKTV strickland_n_Page_03.QC.jpg
9993fb505e916bed40452dcf7e5b69e0
c9eb895f77224129efc6b8388adb6edb1db27317
17994 F20101203_AACKMA strickland_n_Page_82.QC.jpg
90d00d3642243ac68fdb94571a2ba19d
b0cb5abdd0df2786653689cb2d7972cdf8fa7f44
1051968 F20101203_AACKOY strickland_n_Page_20.jp2
8e4fa8b42764aa5ce615d5f5b2807037
772b544160756c22a9de51673e2c1ad5c41f00c8
6742 F20101203_AACLFR strickland_n_Page_23thm.jpg
054417f0306be8a95ba4e44dc0c0bf5d
1b57e530010f3b4e51b871de37c76287a50eae5d
1051936 F20101203_AACLAU strickland_n_Page_31.jp2
c417df2dbfe0f8e67a85ea012e952c97
ce4a678c73fffa80da4c878543853dbcb3a4a2d4
22319 F20101203_AACKYU strickland_n_Page_67.pro
cf9adb2c26f19a7bdd0eca0a9c0ca720
fa0e78a31e1329da73a95af3eb0708743005888a
19479 F20101203_AACKTW strickland_n_Page_77.QC.jpg
e81bd77e7361e5b6ffa0cdae08dd4301
a588dd39e0344e3972987815472969a683eecc6c
F20101203_AACKMB strickland_n_Page_30.tif
0522bc7c293202c58edeb9cf4af76c76
42202a7929a89569f1dcafc12ceca2d8d8de916b
F20101203_AACKOZ strickland_n_Page_57.tif
12e8a0f297db5d2b97d64ffedbc7134a
71b5b80f40f7c36ee112e034f7f121e2d42fd7c1
7045 F20101203_AACLFS strickland_n_Page_24thm.jpg
2f0dbcbeadb82ea880e4600cd1d019b0
acff7b55dd7bbd12eea9046c41de4616087ebc7e
1051909 F20101203_AACLAV strickland_n_Page_35.jp2
e6bfd0161d0eab1d71d9342aac6d965a
4b175245940599125d56b52b394560c50b9cde8b
2376 F20101203_AACKYV strickland_n_Page_01thm.jpg
e2ab7357d03d99742354d46047b89873
dec940c178bc428a069bb26664d97789927f6214
27182 F20101203_AACKTX strickland_n_Page_21.jpg
06b7547d6be7dc53757a214a7d2e4465
210f0f8b2920051000aafa7bda0dd32510feb5f7
4815 F20101203_AACKMC strickland_n_Page_75thm.jpg
f74e37ba2fe5deb858295c534742c6e6
13ba7d220178be04d2cb43d96e5a4e9555588e14
25736 F20101203_AACLFT strickland_n_Page_25.QC.jpg
76547fddccbf8d7f66c2622558847e1a
9942c197b111c3915efcd900f6300ed3b31f1983
1051921 F20101203_AACLAW strickland_n_Page_36.jp2
f1bd9dad311a30e715fa26f0c71276e0
71fef05b7c2f346bc8f74aa50f81642b73bfa7d7
786 F20101203_AACKYW strickland_n_Page_10.txt
061f3e122943468d52e865ea0307ef1d
f2f47196ea462fe66f631b95dd9a3a024c27c66d
5870 F20101203_AACKRA strickland_n_Page_57thm.jpg
7d94644e85abd63bd4ceea22b1fbb44e
a7eea31f6fad2e6036a2e8f345c2b2bf80d54e91
51240 F20101203_AACKTY strickland_n_Page_35.pro
216cbd342092021c296fa6706a7a9848
ab4e03b1e0ff159ea6ace1bec96b2fd45a8d0fc4
25234 F20101203_AACKMD strickland_n_Page_73.QC.jpg
c7350da7dfdb4f2d06e03e67d24c336c
1242a44ba7402382f3e798fea90eec63389c9e81
24974 F20101203_AACLFU strickland_n_Page_27.QC.jpg
97c258be70ffe7c4e5758a89ceec3eba
cfdf2ef61c8bd3227dca5f14a08f66b6f6385ad6
326692 F20101203_AACLAX strickland_n_Page_38.jp2
f1ff5949f7add4c698d403ebd64e916a
7f0a1510a4ea2c04d827b1f0fff7e642f54c1746
127845 F20101203_AACKYX UFE0021829_00001.xml
e4aee17a23d67b8cb2f041e2e3b703f9
f37ec4c8a35a5d5d4d2d8db6e527b12ef6428e77
19710 F20101203_AACKRB strickland_n_Page_10.pro
715ec7420475f5dc6ba6ecdea70fff8c
ad80f5c0cc78057162252b0315f867c96d8f3032
F20101203_AACKTZ strickland_n_Page_74.tif
59f321c3bd49d7b940ea615779e248b9
4beefb7ddc4e5f5821060c93b739c4d17ffe2559
2191 F20101203_AACKME strickland_n_Page_68thm.jpg
b6e72f734f440e744e61bb1bb8c84e75
31d6effd882e33e5d89f420039dc0e96111f7f9e
6609 F20101203_AACLFV strickland_n_Page_28thm.jpg
7e5ae7a59b8d365e9f3124902b4c4f26
4bc5c4954aaf78335d4fb96b87382df86a30abc1
836194 F20101203_AACLAY strickland_n_Page_39.jp2
4dc126f36b1bcef897e769351e49798b
59a57c64127a43651ffe29553a7435d0c4534ba7
F20101203_AACKRC strickland_n_Page_63.jp2
547ce168c34a31d80908c2e590056351
022e077f38a0ff877bf3cf20877c4e5781e46099
333 F20101203_AACKMF strickland_n_Page_45.txt
f446998606715cce86ad51ac648b063e
27085dcf0f8abec77ec588144fc05bdaf5f1bcce
7336 F20101203_AACLFW strickland_n_Page_29thm.jpg
c09e60bb5b2b9072f892abc8cd2b0df8
a98734fa5b402ee7714006083ece08edc6ad2782
965561 F20101203_AACLAZ strickland_n_Page_43.jp2
e3a30f8387243418c3497908d1ac59b2
d55591206c9294c57b6ca3fd7ba5cebb1f46886f
91 F20101203_AACKRD strickland_n_Page_02.txt
e0be0fdef75437ba91ee2763c9dbdc91
ebde4de14316c99e3b98279fbe8ec892317d1900
1051969 F20101203_AACKMG strickland_n_Page_73.jp2
f7f422051fb9b5b2e9247260ba1c6f82
c3b5a5f84d96a9c6de7aa27a20fd6e714c833bb8
3866 F20101203_AACKWA strickland_n_Page_71thm.jpg
626e3bf47534723ae46162c92025d28e
ccc693d900da20d86bd3fd4630d799091dff8cb8
28437 F20101203_AACLFX strickland_n_Page_31.QC.jpg
6967cf44d8d771885fc3cec234c76a8a
53e63c25868c754b65f203e0fd43370d30482aef
64591 F20101203_AACKRE strickland_n_Page_67.jpg
3bd30386a5df13cd91c7f1d9dec8e6ab
67e90b8d0f4355e0863e99e881f303639de2e4ed
1051952 F20101203_AACKMH strickland_n_Page_77.jp2
719b8c8f92f55ad6cd2063f18a933de2
ec83681d8bab1b013290e83926e094f261ca9880
33079 F20101203_AACKWB strickland_n_Page_40.jpg
085ff3a55e8a4b15657d119b856e2932
021976816efe317579822ad1de5415be426afbab
7586 F20101203_AACLFY strickland_n_Page_31thm.jpg
122ac2e6eef9e679261df2e886e75673
737654989d40f5a024b489ce2d140f1963cacfb7
49379 F20101203_AACLDA strickland_n_Page_18.pro
135df5d1cdb8f72a446206eed823c67b
311a313aad1b79b0ab7ff58a5d2388b228231789
431 F20101203_AACKRF strickland_n_Page_57.txt
992143dd9477919b03a095f648d4352c
663c91e39049333e7893cc9cbf35c5a78a9c4022
25840 F20101203_AACKMI strickland_n_Page_20.QC.jpg
16d3c72ef4b9a1cbd9b72b5e6f5c9287
a80b797b6e2443fd76e2b0a0527bf9397e806d70
90401 F20101203_AACKWC strickland_n_Page_30.jpg
c6677dd61b2ecc39afa385e9e69c018c
69ebe386c7c0fa51229ea16b69433ac500e657e6
28372 F20101203_AACLFZ strickland_n_Page_32.QC.jpg
c260c30b97b366534f9542cc68e190bf
36500362806087e8ab2e222f7572076472615714
51785 F20101203_AACLDB strickland_n_Page_20.pro
442091e2f1e8e930100e6722af971f52
69462dc7f5326cf461a018d7e84186c6eea46355
50007 F20101203_AACKRG strickland_n_Page_62.pro
dc01c39e0efbbf820eeafc7c06180975
2d73d8f07d62a0f61ecbdcbe02d2045a17697c52
55002 F20101203_AACKMJ strickland_n_Page_53.pro
118789fc9106f5c706d5cc40304c6497
b9cc448c5e1a9d5b5d770c4a2cb8926afb3b426f
2864 F20101203_AACKWD strickland_n_Page_08.txt
4524b2a411662080c28ce5f696a45682
693bc4d82085b3e9661488011d179e51bb7407c0
7145 F20101203_AACKRH strickland_n_Page_35thm.jpg
284268ceb4739483e8e0f6c6653e64d0
ab5a78ce34790b6ada6537eac5b7f640110afc91
55338 F20101203_AACKMK strickland_n_Page_61.pro
814815c0d7a4339ae10204a0b4021c71
6c664b38993b21aa84769ba19fdd2a0f5c56f06a
4066 F20101203_AACKWE strickland_n_Page_40.pro
db45dac4fa16f6a620588619c973ca5b
b16f683f864c04007a894803a5f408f244c51850
49203 F20101203_AACLDC strickland_n_Page_23.pro
30091a2ed9e9f90bdf5e901672a12287
be181996772f9b15446998ff376667ac338033d9
61335 F20101203_AACKRI strickland_n_Page_81.pro
267141170a77f29679829f8f1006d223
3b6e17e4c76cf77f2a52815041c066d559f2d3f9
1051971 F20101203_AACKML strickland_n_Page_14.jp2
0b47e80829b5e662e31c7561c059ea50
8f84380bb8d92d5bea3606083f2ed3461fef032a
80321 F20101203_AACKWF strickland_n_Page_62.jpg
7c16bbd82bdb7dcd31dcddf2cd96c088
b09603ac9438ce6294e71b6cbcb4ff7445f2d5e5
51009 F20101203_AACLDD strickland_n_Page_28.pro
a7ee0b42f56cdada1a80ac8d5e7984b5
a219128f294df50fdf1690cc64207ce9f7c298e8
F20101203_AACKRJ strickland_n_Page_38.tif
6160154f770efe9852fcf656a87d399f
d377ff2ef33878c53bd8223dc0e13b723dd3b570
18377 F20101203_AACKWG strickland_n_Page_70.pro
84701eb15af375a11febcdf9a8ba2997
42eca91da44e6ca61988454090a518e3123bfb4e
54022 F20101203_AACLDE strickland_n_Page_29.pro
f9e97e920b733d9155a35b4eac3d3f42
3f635c32f8504838f1b8f2ba9e8be53ee893034c
F20101203_AACKRK strickland_n_Page_50.tif
0b06ea1d1ae547f8f8077942d3c1564a
6189cd1efb65f33d278b13e36c2cdf1c1fb78843
7714 F20101203_AACKMM strickland_n_Page_33thm.jpg
f159585fc0ef4dbb28f371ed06f1a9a7
ccccb862de5d74ece4290a6e116488787e0759eb
77909 F20101203_AACKWH strickland_n_Page_59.jpg
e6f4f6a1e11fa7568878c3939a482b01
d2ee24119ab08dba3a8ac73d2ecca308b8341341
57794 F20101203_AACLDF strickland_n_Page_31.pro
091efba139ff18caaab3f9fb71a62e93
a33f2b3508e4200f132cf9830f99d3172c905311
F20101203_AACKRL strickland_n_Page_31.tif
154cea4cc75060fb648c30a5d17e5ee7
4f0ec93d9f06ebec30e2f4ed9941f6a263d49404
260 F20101203_AACKMN strickland_n_Page_68.txt
82d660b8a8fc1a3002937c5270bfebf7
182e6872ccbaa7cecdcd21a5235a1f3851e63cf0
11652 F20101203_AACKWI strickland_n_Page_49.pro
c0098f9d09726726ec359f81b61fae97
cb9a399581ad87d20ac15e4ae8140b99a4a9d610
51266 F20101203_AACLDG strickland_n_Page_36.pro
3fe78a067ad8c514bbcb0543c8fbb77f
dd6c85c95783962b9e994f6720ff880d3ca326f1
9422 F20101203_AACKMO strickland_n_Page_38.QC.jpg
1467373a6e4d73843d314aba66ad2afc
918be7ee660895e8f391f6083a96e9c3571faaca
1488 F20101203_AACKWJ strickland_n_Page_82.txt
3254380c3e2992d30f998f2e717566b5
18fe8f5fd4ece2df8d80331ac8bf629a67ce1414
3132 F20101203_AACKRM strickland_n_Page_02.QC.jpg
c342837f988b1f4cb7dc0bc2841eb8b5
c9c55c65c5c8bfdbbf8100ca9e8268f4104aed02
14405 F20101203_AACLDH strickland_n_Page_38.pro
a618afdee6d6126ea445835a76bce830
761b671d8059748fb33cc88b348844052d1f1806
1742 F20101203_AACKMP strickland_n_Page_67.txt
91307c67c1ce700278c04630042aee7c
d7cd15aad723b4fb76ba5f178b55b8323919bf74
F20101203_AACKWK strickland_n_Page_41.tif
37f8e26f3decd49c57372c09192a9fef
9428cef9831c0dfb08d57cd6617f2c6e9aea55a9
52234 F20101203_AACKRN strickland_n_Page_15.pro
7f12fbe527e92643a85e0320cc3e1050
ceb9d5973de1235298658f4135a913877cb75e30
5121 F20101203_AACLDI strickland_n_Page_39.pro
ef2dc566cbb5061c3dfacc076505f63b
309307b1c2ddaa89dd65d26c53286c17c784ba65
2095 F20101203_AACKMQ strickland_n_Page_16thm.jpg
08aa3e1d48c67afa5ab6e17e9e11388f
c97527daf2211e113496b80f843b2809dfed90ec
51049 F20101203_AACKWL strickland_n_Page_73.pro
9a4596042304972f90f3838a1d116b4f
cad81b37fafd74b6a239b8244226fc507e36b7bb
3643 F20101203_AACKRO strickland_n_Page_41thm.jpg
f316c82320f1b67d03ccfe40f8e25a31
90e4b3593054d2aee94d837f9c3a6a5efa2a964d
12624 F20101203_AACLDJ strickland_n_Page_43.pro
f4235fcc45d2ef84ee0202484fc2bba6
36f5bc58ca98953999b87fb6638f3ce9230e10d4
21094 F20101203_AACKMR strickland_n_Page_76.QC.jpg
990a790744379e3e901ce815272951f8
50c20d3908e84973503fc1318ff841dea1731d93
922 F20101203_AACKWM strickland_n_Page_77.txt
e7997286545d763ec95dbf34ca4f3aa2
2c0c0cc53c53e837ab995a1587c87ce4e282e1e6
11352 F20101203_AACKRP strickland_n_Page_71.QC.jpg
6d44649d3e8fc277a1e1ceae42f43c20
8b7db385de3a86c1aa2afcc022b89b3c0bd05431
9258 F20101203_AACLDK strickland_n_Page_50.pro
c9227e375c8836144f726bd84ac358d7
bd76a126d6afbda2c52e9ba9385f43462bc4dc13
2187 F20101203_AACKMS strickland_n_Page_63.txt
724082ea2aa392b21d044eb6e1f01d34
66a12d59d53842d9df9809b62d5f60c7287d4d2d
13371 F20101203_AACKWN strickland_n_Page_12.QC.jpg
55962f5461c226222b8edc258569ad5e
86ce86011317ee4a4320ca11f1ec83d1a0a82046
498 F20101203_AACKRQ strickland_n_Page_78.txt
ad5eb90b226f867af17384ecd921beb6
335de744aac4aa0a64d2cab5b1418fe41328b854
38673 F20101203_AACLDL strickland_n_Page_52.pro
c4afc85016c304675a050272cab25139
085c4fa2d4f96c0fb6ee292639cc7e7fa85bf8b7
19668 F20101203_AACKMT strickland_n_Page_52.QC.jpg
8cb322ce545b5be6aa65806043eeebba
181506013fd64e18fc34ed7e68c04b7ad99af582
7313 F20101203_AACKWO strickland_n_Page_20thm.jpg
8b2b30e69fea172f0d69a46f9fe16176
d1b996991eb533d3f67465dd354aff89f0b16577
21486 F20101203_AACLDM strickland_n_Page_54.pro
d7c05901771722768f608e503f38c697
a3c55e2bfc4fd7915607bee084dd8e1ab3ed0ae0
664651 F20101203_AACKMU strickland_n_Page_45.jp2
1bb590ddd3e8393cdd8a35c3c6217299
e754fe17e3ac81101f8dc21e69536137e11a33c6
1051985 F20101203_AACKWP strickland_n_Page_33.jp2
2944f471e9366ca5e501580135925158
31603f1c16a67a0da5a0c3aba90139633e1dd875
47375 F20101203_AACKRR strickland_n_Page_17.pro
33ca5b90744296a1dc8625683ce9efb6
58f2e1906f3519826e0bf7065403d9599e8c3614
8789 F20101203_AACLDN strickland_n_Page_57.pro
068bf8bfc29d97283658ec757d3ea9df
73647fc5d02fc20b3c7c752bc2f9b3e5773d5c69
48101 F20101203_AACKMV strickland_n_Page_59.pro
8153c2c0ffd5d7fc4d37e7530fe477a4
729e1162cb400f8e31910a5b6e78fc9db44edfb0
F20101203_AACKWQ strickland_n_Page_40.tif
69da1b23221443ec712aebe717a74627
27b248fd928cf2d997b98ce40d500d0f8d60c235
F20101203_AACKRS strickland_n_Page_08.tif
0639dd21524c6ef0de2835aa740c50e2
0dab95e2c5c241b32a6312e4d8bb96331e858a3c
53905 F20101203_AACLDO strickland_n_Page_60.pro
7327699a1cfadd39d75cc0fdadffe34d
cf64e1a79ad67256dbfff9de548e326fae4b102b
93780 F20101203_AACKMW strickland_n_Page_80.jpg
5d653a78cf32a670f627b960fed4fe9d
08226023bbe7272ecd9e749fb06013a24f5ec3c3
1267 F20101203_AACKWR strickland_n_Page_75.txt
180944836bbcd271d4d16f74d0d7b068
52a8d47797517875373b25bc82b03780a8d01325
8683 F20101203_AACKRT strickland_n_Page_21.QC.jpg
2654192df7435c4a7c3b440d6af77c5f
faa8a879ba61fde7d93f6f63b300a95c14482c26
24777 F20101203_AACLDP strickland_n_Page_64.pro
1e2ec55b8d6cdf41fdccdac78267d1d6
e99d81c1b1f32ffaf57cedda25bb4b727bc89215
611 F20101203_AACKMX strickland_n_Page_43.txt
974ba8adb2e84070daf26885af5eda3d
1dd5f90893a58d9f53d28548bdd3c4e1343a25b8
97390 F20101203_AACKWS strickland_n_Page_81.jpg
748fb13ef81f253cae132972c66bc54f
8f3355f3632f72d1ca1568a251b236067725fd65
F20101203_AACKRU strickland_n_Page_53.tif
dd822444ea5aaeef3b9ea33e80635284
fb0f807f8f164128a39f5697a85f3fde153ddcea
25472 F20101203_AACLDQ strickland_n_Page_65.pro
38a993a319ad11a3f3ed367347e7f0f9
a924f6ddc1d48cae762a5dde2d72b28f65558909
6309 F20101203_AACKMY strickland_n_Page_46.pro
f9e0366a25b550df963008b16a378fe3
559d7b303495875f2c7f36a8d6caab192bade0db
F20101203_AACKWT strickland_n_Page_29.tif
61b4830ea90a74260a7c350d388f974b
74fbd1de657926a7b6f51f0ef82950325e8799ea
86186 F20101203_AACKRV strickland_n_Page_03.jp2
e55d465176bd6513b0319a6c92960b6e
e2370b86c794c1f8027695e71ae531e0aa76eafb
25193 F20101203_AACLDR strickland_n_Page_66.pro
9fba8fd73aa9e50e927c9b5f60caae1c
833914df754b3389386cf3cb656617aa66583c58
1051972 F20101203_AACKMZ strickland_n_Page_59.jp2
0f31677de0f877e1062933886066b3af
fdca2ac4aad6907ba8bf259348b5fe6ba54b9603
3903 F20101203_AACKWU strickland_n_Page_05.txt
70dabffacc87c4294e06e5ba150a9c55
6346f185f409a6d49d43d46ec09e4ba087f4cca5
1051978 F20101203_AACKRW strickland_n_Page_30.jp2
240fd2488418044d11e6e21342aa2d1a
57a4c0ae0221817049dd484f9711e60335de36c3
16099 F20101203_AACLDS strickland_n_Page_71.pro
0d7e1ec73a7c6c55a80c739989188aa5
16ee1c3024b5b1b2e8698f04757d80651752e4c5
2265 F20101203_AACKWV strickland_n_Page_14.txt
b24453b0e15dfbe39b5e8db8e0a478bc
027d30bc731d545f24ecb6622ab667933a5d3cf2
4747 F20101203_AACKRX strickland_n_Page_50thm.jpg
74a9844265b52b784b20a328f078862f
f42a97fabfacf271699f83b8b88705c2707cd3e9
31816 F20101203_AACLDT strickland_n_Page_75.pro
df1227e6583a748c955046cccbc60cb3
1bfc91c8ce662f7ef52d6cd876201d0f0a7f213d
12819 F20101203_AACKPA strickland_n_Page_21.pro
2e2db776b7b297151a6adbbeb0d05c73
1a927025343cdd03df125387a48888ea16359ee5
F20101203_AACKRY strickland_n_Page_25.tif
942279e4d52ded4415def1105da2d064
2bdb35bc20354d9f6c6dd143b45f33dcd4822c45
F20101203_AACLDU strickland_n_Page_76.pro
1c39b02e67cdbb49d229252331c03e35
fbb6aaa8fc8c535818bb7f330eb075ab5c599566
7149 F20101203_AACKWW strickland_n_Page_67thm.jpg
bbffc4a5077a3748be4dc526e5b70e92
bf241833f8a66279d9a3262c5ec2e17da7a7f06b
84841 F20101203_AACKPB strickland_n_Page_24.jpg
5940d5ee659b4a3bc1fa88c0bfde49b6
1a6c009c30807ca4d4290a259923b440a1d045b8
12897 F20101203_AACKRZ strickland_n_Page_47.pro
68aed348f5cc3478aefff715f5df4ec5
d9cbfce482a514125e1263757073fee76b4a6ccc
16761 F20101203_AACLDV strickland_n_Page_79.pro
b7fa59f344bd9089b5e9d2c02b562929
2d11db43b94da786aa5d867fdb1de077e06b25df
52546 F20101203_AACKWX strickland_n_Page_48.jpg
e24345264df8a24c9ea9ec779489ecc2
bcd06e7ee52b5f4fe4e4cfcfe8cff90cf8878aa7
F20101203_AACKPC strickland_n_Page_21.tif
0de5847cb587ca242958f02f0e6a9384
06b8bb0705fe870ccba3e90169e11c4609827687
31452 F20101203_AACLDW strickland_n_Page_83.pro
a19b985f13774fe13b25924b821f9ef9
244a18d4dcf2377ebd8b17ad9881c5cd8bb7c14e
6461 F20101203_AACKUA strickland_n_Page_42.pro
c88f6eb11147d2e6c14a1b8ab4e5cad0
576f4f65eb60a0bcd6a683386520913d4dd66a97
7418 F20101203_AACKWY strickland_n_Page_53thm.jpg
6ffa280121329a1792f3090321130ada
d85be288803dd331ce1beb58b1f5425b3826a332
24014 F20101203_AACKPD strickland_n_Page_28.QC.jpg
34649404750d686a7240f9e224037216
ac5518281ec240c48071ad6ed120300b6d365ea8
256 F20101203_AACLDX strickland_n_Page_03.txt
0ce6a7785378de5db0ae7e3a8434a68c
e524ab53e66fcab5c1b2b089cf32debcb309d632
F20101203_AACKUB strickland_n_Page_83.tif
c77417bffbf4d2330d939a959055c164
45b71ff6a4ec6d1fb5ca5d625e5f67b8525dad77
7432 F20101203_AACKWZ strickland_n_Page_14thm.jpg
0abd54188084a1ca30bebd8409f49f93
3544e17531b0e36e7ea1e2ab9145c59f4c0f11ce
F20101203_AACKPE strickland_n_Page_09.jp2
62d8d56a414a7c13ad5f1683f1d728d2
06438279e7b4d62e992295bd52d052e112abb5be
751 F20101203_AACLDY strickland_n_Page_04.txt
886c3a87ba1cf9b3f421f1d90664d71d
b5f5fe4da5c982e5c76d3d5577e3fb89b4bc3254
14338 F20101203_AACKUC strickland_n_Page_07.jpg
3e9b7ee714120b45935f7901962afac6
35a3cd520c463edcaa225f0370ec430d0bb9e266
860 F20101203_AACKPF strickland_n_Page_54.txt
90d33fcd8a6e3a92137c87a547bb0f28
1a4b7d96b48ea6cc6e48d64513421c341c27e976
963 F20101203_AACLDZ strickland_n_Page_12.txt
1a24ad801a705c2b16975bec23f1caa6
7a167627636165e63def30a9c473000f7852b750
1051934 F20101203_AACLBA strickland_n_Page_48.jp2
8bebde5e4178d218b759c07166fc8c19
4fc38ca30cc292551fd903ab941a0d7d87b2496c
14997 F20101203_AACKUD strickland_n_Page_48.QC.jpg
a0ae30f75b449ffac9c1b4a1d4d5b0c5
bcca539bf799b765aa3d25032eb8984d1e9232b1
25468 F20101203_AACKPG strickland_n_Page_08.QC.jpg
51b510737409f8fd2ad0c4bf88383bcc
202662d1d6d29fb1f83d73f9c42a2677665a7bd4
10222 F20101203_AACKZA strickland_n_Page_02.jpg
d5b2a22d1d33c12ca4cf93bd426491cd
d9cc0f95100fee6581ae069c01456ec1d90df75c
876613 F20101203_AACLBB strickland_n_Page_49.jp2
04d0c9df6887bbbecc8eb7171aafbae9
b51ddefb168338d27b74bc12f370fd83b0d8d650
6294 F20101203_AACKUE strickland_n_Page_77thm.jpg
a162fbcd905a5fcf4efb0144101f9a27
b159358f1abe60082d4205bc7de39d3ac7e3ae43
4655 F20101203_AACKPH strickland_n_Page_43thm.jpg
36aea493d702ebe2f79a7f9e0d26cdbe
9e537ce958a8a0732a81e202e5ecdf5f61c433c3
83818 F20101203_AACKZB strickland_n_Page_05.jpg
4a1c2be2529b2196f8f912019a27eb36
6809522e8a6b0eb1d249dad5f7d45aac981fa58c
7455 F20101203_AACLGA strickland_n_Page_32thm.jpg
5f4e036826265e33cec7b1435e3919f6
48fb922dbff9352ea59bc4d00a7b0a08a22492a2
919964 F20101203_AACLBC strickland_n_Page_50.jp2
3036b252dbbe13513a8793a018a1ca64
96829e911dcca5963061d1e3278f68fa1a3a5007
4868 F20101203_AACKUF strickland_n_Page_56thm.jpg
3e97528b545029a39e877fea0214c7b2
763f40e447c057c924cefefb90d4eb417a59a0e2
1051923 F20101203_AACKPI strickland_n_Page_56.jp2
1f0924b4cd40a6f7b1ce9738c7ada1c3
1686aced8ed2ea77a1241ea36cdebe21a31c6b27
96186 F20101203_AACKZC strickland_n_Page_09.jpg
75284df1d9216bb2fa83c4be368c9110
3b58466453f608d9e65df03a3f69aabaf4aa8da8
24509 F20101203_AACLGB strickland_n_Page_36.QC.jpg
e430314eb84fe78f5fcc3ed926164c6c
fe7623a9a1d775c7052f912b921fdd5ed203d3ea
876155 F20101203_AACLBD strickland_n_Page_52.jp2
067efe521dde8ed2ec8a198c177fc8fa
c1aad32ba44f004c2cbc58a98ffb7fd570e850ad
F20101203_AACKUG strickland_n_Page_35.tif
11d14b584f753cdb7ae946d88e456260
943adfdbae4b77cffe5ee01945ac867ae74528c3
F20101203_AACKPJ strickland_n_Page_61.jp2
ccfd47c249d4e0ee352a28fe199a3664
27c701024207384e2dcaa04157d9d07e0a38e531
83477 F20101203_AACKZD strickland_n_Page_13.jpg
a563199dac7b60a037ff1118166bae67
9b01ad8ff03d1b8da02dfd6875f772b1dba3ae19
2795 F20101203_AACLGC strickland_n_Page_38thm.jpg
951f9d348bd5b1bfe09629dee9207447
bc1f4c936d9d34d350882ae514cb2a3db645f53a
488132 F20101203_AACLBE strickland_n_Page_54.jp2
821516f4b5359055130d7c19fe37240a
181962c8bf98fa56c1180d01532cb61dbdf6e5ae
8390 F20101203_AACKUH strickland_n_Page_48.pro
9b4b675734f7cbd5eeabe9802a19a9d2
22dd15d96db8331ee31985ef3bbb160befbfdeee
542965 F20101203_AACKPK strickland_n_Page_12.jp2
caaf2257b755994053383d185b04c376
caf71d0610f5a7660da17dde1d10f55b7e3b8d36
87321 F20101203_AACKZE strickland_n_Page_14.jpg
2ad4304f3f707397b0786a407e03a801
dffe55c68837c72769c44432c0b3f2db90a64abc
13047 F20101203_AACLGD strickland_n_Page_39.QC.jpg
e582b9e1f947e5f879b8156949ce4ac6
eff1aa4e1d40da09e7183bbc60330e703c84d179
1051963 F20101203_AACLBF strickland_n_Page_57.jp2
a2652d99da032fef1b6d66f09820919c
442992f7f4277f4fd05969988685d1d8b8ebc834
2152 F20101203_AACKPL strickland_n_Page_22.txt
525db5d139a527b32ab9a109c8693423
4385cdca14dd46dbc3ba43c2d79f82a1304c6ed9
76586 F20101203_AACKZF strickland_n_Page_17.jpg
3fda065939a39182b5334aecb77bb27d
c87daa150f8e7febd88768e5ca19e2fd3dd5af03
F20101203_AACKUI strickland_n_Page_78.jp2
041f964a0ff4d8507eaff2984413170e
e2f334df089d0b0eed5db4012ef469cac40e71c8
4164 F20101203_AACLGE strickland_n_Page_39thm.jpg
5ab13d5001670a39ee628dcd57db83b3
e7e616bf36788ddd7c0b77b5c2e271843a3b6987
231573 F20101203_AACLBG strickland_n_Page_58.jp2
2b2a36c443e7f27c303ec473e5d0a993
87e5fdb2d3c2fa7a823d9d26868fa9b65fb34d62
55821 F20101203_AACKPM strickland_n_Page_33.pro
f8d3dc8f190963aabf700d7a875b133e
554fbb54c83bbb5af09a1042bbb052dbad6a16bd
76779 F20101203_AACKZG strickland_n_Page_18.jpg
06f9db0b2df8654e7b43cf1f40f343cd
46dcd37ae968e2fd9c14c4427fa3947b86f7baad
4309 F20101203_AACKUJ strickland_n_Page_07.QC.jpg
9817afe31e477fb2764db5382a2e1d1c
7bced5b16c37c5c4b57dbae24eef08de1f070113
F20101203_AACLBH strickland_n_Page_60.jp2
373d255b07699fce2630aa5f77a04e0f
024b5f02788b0d95f4ec77ccb6b63e454ee2710c
565697 F20101203_AACKPN strickland_n_Page_40.jp2
afbf30d3153875f4c2d02688cc357cd9
ff060b79932b50b9c62a89c5e46aa1114a2dbeaf
81929 F20101203_AACKZH strickland_n_Page_20.jpg
0cbdd91200b39c1cf4c6d538fa1d2ffe
29bf2238b9108c88e05e08bdec1a07f2ca4d07cd
66455 F20101203_AACKUK strickland_n_Page_77.jpg
a895621cf067e09a2702644e00651288
2cd2e059a078ba78dce7c86c48ef811723338155
12361 F20101203_AACLGF strickland_n_Page_42.QC.jpg
f9e8338967ebe4580ca6be68eb70eb86
8c7e304c91907552406069f2824d7d0ef2bfb4ec
1051967 F20101203_AACLBI strickland_n_Page_65.jp2
8cf8f51664b68111ac96b9bbdaab3a0f
0f505ee646bd1bcb9bc423f994e87e8a5637660b
8296 F20101203_AACKPO strickland_n_Page_78.pro
610e8b21d86bf75e6d7f008814b85bf5
c643de03c8562547833c22eeac6472ddc2003cf3
77055 F20101203_AACKZI strickland_n_Page_23.jpg
b7da028a82d0636c933327bf8a1926c9
6603b0da0cd7f7d51a4f2445301a52cd1d08fb88
1978 F20101203_AACKUL strickland_n_Page_18.txt
a614758bb70ff118d37cd66d7131a627
eb83f953a68d6645e540133c3599586aa3c3b09d
4305 F20101203_AACLGG strickland_n_Page_42thm.jpg
ba0ebefcc071a693b4f3b4992a419f85
f0e621fdd8f530607b4176521665d12048c38457
1051959 F20101203_AACLBJ strickland_n_Page_66.jp2
612fb41930b7aa2a54856b55d748f871
96af0e8039249cd86c60f85c1fb94b117a9dadc8
109011 F20101203_AACKZJ strickland_n_Page_26.jpg
906ba6da277686f93648ab45763b5269
b9885850588ef13ba1efe202d5ac07aa0351fb31
4463 F20101203_AACKUM strickland_n_Page_49thm.jpg
f5143533ace0b5fd2263737cd4cc2e9c
92768a5059aed281f20bb4c585203cee1e60aa4d
14430 F20101203_AACLGH strickland_n_Page_43.QC.jpg
72e2727fe6d6ad2c67135106a11908a7
652a307b60dcd23b913f07a9da502a1e69bda81d
564530 F20101203_AACLBK strickland_n_Page_71.jp2
4bfbe6c8d8a7150ecea3c913fc91f04b
bd4ecdd568f6e456647e72547fde51784ecdd1bf
F20101203_AACKPP strickland_n_Page_24.jp2
b6d745525d871c711687d7ffc9363cfe
57993f99969d97c413b7e998846ea3d2a5da8764
80926 F20101203_AACKZK strickland_n_Page_27.jpg
7e3c532b8de69d488239e73c4cf81a7f
a27e6ab463fe2247edb54257811e35f63345f755
7542 F20101203_AACKUN strickland_n_Page_45.pro
3b641a99bddf7575d437607b262ca922
f9116489c949c0712ba3dbe44d4bfa7805bef93a
4007 F20101203_AACLGI strickland_n_Page_46thm.jpg
d965d135e3952244208dc5f901b747f0
ccead1360e6ee52c0d3d160b5669dcca88476b75
615994 F20101203_AACLBL strickland_n_Page_72.jp2
d259acadce3dc5779dd6b1b27f94a1b8
919f61ee10f4b39205b3f9b476d20351de0c6800
863 F20101203_AACKPQ strickland_n_Page_72.txt
7503c6034f373947aa15957c3f85da3c
e467ebbec80c8c9eb26a0193751da39b87f43f8b
80452 F20101203_AACKZL strickland_n_Page_28.jpg
a40cd2949c9aa239a873934d744556c2
8b71ef7775e880fe98b662f4cab660a8efc056a6
13352 F20101203_AACKUO strickland_n_Page_44.QC.jpg
e14c37eca0ee08b42b5edb4d98de6eb3
97a5739ed5985effb6a76bda7e100a7b23843a01
4528 F20101203_AACLGJ strickland_n_Page_48thm.jpg
ac0b1310796172819c0401d0feb9d1d1
e0c93e0ac0a95514a815d3f04a43e83ad23d3483
716537 F20101203_AACLBM strickland_n_Page_75.jp2
2f6d6ba56f5e9a6aaa0d187abfdf9dd4
56749a0bb624025fa53a0ec64014bf1c3fbbfd35
1255 F20101203_AACKPR strickland_n_Page_65.txt
d271d1057fe47895ce46e98a2a62e450
de07294bfc4cb129a5ae1bead43a07c8737f222a
91570 F20101203_AACKZM strickland_n_Page_31.jpg
9a98efea54f780102cf792c962f943f0
790cf2588682a50a9d32f97f6a24a4ee59697c30
7472 F20101203_AACKUP strickland_n_Page_81thm.jpg
eed18157805ff352b488398134e0fed2
8f500fd9815c8b6274c8470e5a85855111d7afab
13598 F20101203_AACLGK strickland_n_Page_49.QC.jpg
56f431406a469fbb1e5589b39b5da711
7feebec34136f7229f44af52c8affbb043879dde
F20101203_AACLBN strickland_n_Page_76.jp2
8c5b06ce12e61c2fd9832785d62f2e71
de3ad1fdbea5b49954af572940bc4a9ec3530ba1
27979 F20101203_AACKPS strickland_n_Page_33.QC.jpg
de1c171ddae01dc60019371b6bf34c0c
79153d6dbb7ccae49cc87a87cf6d0cce8cb9c944
95443 F20101203_AACKZN strickland_n_Page_32.jpg
a17bee822ab76abc0d259b195d699918
8f0dee87fc6965db1c1929cea0abb03930fe3b46
28778 F20101203_AACKUQ strickland_n_Page_09.QC.jpg
4f9c575630854abf2bff7b795fe426b7
56a8484a40b0733b281037d1053c3ed8f1df33d2
13972 F20101203_AACLGL strickland_n_Page_50.QC.jpg
8b6c7cf07930755dface5d78bd9f3bf8
1c4846ec45f87e94165e984aed15ec40e94a222f
F20101203_AACLBO strickland_n_Page_02.tif
5921d0f9603c3802050db1b6d0ba5399
6add439f95e678e6c1b5b13f24217b7559151e8f
6619 F20101203_AACKPT strickland_n_Page_62thm.jpg
091eccce0c434294b919be04964bd215
0757b3f95aef08b7fc883909ee0563536121aeba
89156 F20101203_AACKZO strickland_n_Page_33.jpg
c1d95c247f81b0aa52cfeedb08eacad7
5fba7996b1eca4f7ac4e92bafa3ad4ab77982f85
7306 F20101203_AACKKW strickland_n_Page_73thm.jpg
943f8c22e7ed2c046d813e15193b1b4a
01a609e65e7989ef1eaa6497293f0aeeddf4767e
34813 F20101203_AACKUR strickland_n_Page_04.jpg
c95be9498557790ba1aab337229aadf4
64a7c9aca713974c75c5ac0e8b4d451b16d9c27e
19993 F20101203_AACLGM strickland_n_Page_51.QC.jpg
50577d222b12fdcf0ac7c298b7089a89
bb4de3ccce7d347913e4a3e548afa0a934c4098a
F20101203_AACLBP strickland_n_Page_03.tif
a2783e1d91e565bcf32aaecf1c89c8c9
67a7b6803ab1a0676da7d3c4a25f2c0f0cbdab05
56616 F20101203_AACKPU strickland_n_Page_79.jpg
fc961ab194837ffa9d5f87f195231f4f
17e6655bcf0f2673c509476cea6f619ba72854ee
81476 F20101203_AACKZP strickland_n_Page_35.jpg
8abbc5925023f40c0ffb2a11a1c82f01
ff7496de7baaa5c366af70ab1b2b4aa1a8974dfe
F20101203_AACKKX strickland_n_Page_77.tif
36e0bc99cfcc28e1ccff1124cbc75575
6f4ac43bee3466b9745fe1a161f0a710c2585a03
500 F20101203_AACKUS strickland_n_Page_48.txt
9457b51fe3a4523b87572429509b2687
6069898e2f5912873f26c6767fc6d59fb1fc4f9e
6023 F20101203_AACLGN strickland_n_Page_51thm.jpg
2c85737c0f065fd694d09fec9b917763
d047771fbcaf52fe67462a7ab9f3f5ce73e13205
F20101203_AACLBQ strickland_n_Page_05.tif
5fc4945cde90458dbf40cdf07ebc11df
29e7a16993143ce79e40190006c77968b6bc6c51
93024 F20101203_AACKPV strickland_n_Page_34.jpg
5c59a8827ed7ccc20ca018048c7b6b66
afed51c34856b5a4ef8e48a7f5d0f4e82f23cad3
79402 F20101203_AACKZQ strickland_n_Page_36.jpg
d5e351601bd308ac53bfaa03517ce76d
12d66c92ffe3d8f5551320afd45414bc93aed6fc
F20101203_AACKKY strickland_n_Page_70.tif
1aa0007c09980f1e0df7d9825461743e
09e30c22a38e6a5a4324227b40fd6baa13e990d1
F20101203_AACKUT strickland_n_Page_61.tif
472c627c37c1a307516f2f3167dff477
a58b483ac128a40ec8e34d9e478e5813089b4dd4
27263 F20101203_AACLGO strickland_n_Page_53.QC.jpg
1ba027a14620e930ddc8b8c4078625e1
0142a6c36221e3673f65902d7117fa0a9e7c1f4a
F20101203_AACLBR strickland_n_Page_06.tif
9cc54b657a8473616b44f0dd88306f22
aac511350c59c97b5e9d2a9e0162f91ddcf8be74
3046 F20101203_AACKPW strickland_n_Page_26.txt
94252c7d150b410600e89e7d31badd5d
1b4df9c2c1815f14b10ce5230848dbc9913a02d3
39090 F20101203_AACKZR strickland_n_Page_39.jpg
0cf5f83c92345e6c439c49d3d4223d3e
369d966d3074ae11b8635319ffefbd871ca62b4f
2202 F20101203_AACKKZ strickland_n_Page_58thm.jpg
27e2ba5ab8f19097e0a9872a00733059
ecbfc5e18bea710331161f1145727c6db804d1c8
16954 F20101203_AACLGP strickland_n_Page_55.QC.jpg
362f339ba38b2e42b989b94ffc478b6c
bb21a1b1daee5fd3ec2345f67b2196ab0ec47656
F20101203_AACLBS strickland_n_Page_09.tif
132b77b735a59e2d921a575e50152121
b9d0f1a6b7b7d4ee2b7d6ca80ffa1bd18f0c161d
2069 F20101203_AACKPX strickland_n_Page_36.txt
de43377b75e206784f620d561747f405
d055ec0ea5f024ff8483ba95dbeeb3c44f89aea7
42072 F20101203_AACKZS strickland_n_Page_42.jpg
2392dbd01afb8316cfe9ff6a3de686e4
b35397b2c0600ae700f170c45b98753c80b22bf8
7415 F20101203_AACKUU strickland_n_Page_09thm.jpg
1e321a9ff4947f3e221fcd2755b3aa55
ce141ac856bf3ea0faec69cf2288d8c3f81f80b8
5267 F20101203_AACLGQ strickland_n_Page_55thm.jpg
63b9288d7bd1fb5195177eedf75205c4
f0744f704a06d5b0009db30fffc0c86609465859
F20101203_AACLBT strickland_n_Page_11.tif
7074e5fde99790f11982812c11b77c0a
d779ed03543fac34ad7defdbdefe304381127005
1051960 F20101203_AACKPY strickland_n_Page_62.jp2
4bb9ef04f59db0aab6e8ecbedba77fc0
d8f3e2e203b3ae2d7c02e36228fc77d1a77ffe62
50717 F20101203_AACKZT strickland_n_Page_43.jpg
602c24ac48e821ee675e66429751be10
226b3db69a504f91555d4536e5a619626946a71e
F20101203_AACKUV strickland_n_Page_47.jp2
226c82060407a9388fe15460e8411dee
3d12fcb6290f7b19459ce7924e89f3d2c50e9c24
62683 F20101203_AACKNA strickland_n_Page_34.pro
553421aaed02a46656b2c7a4db02c21c
44380626d6784f1ace26df016e7a61fed2d36456
27028 F20101203_AACLGR strickland_n_Page_60.QC.jpg
da624dd91882137b2478dcb0f0fcf90c
10dad40704449eff39475ea9de66a71a0e1f5d37
F20101203_AACLBU strickland_n_Page_12.tif
bc15c2caac6f80675c9bfe8b7f7b520a
ad89ed433ac9b7c3930300a6db220e4c50a1533e
52099 F20101203_AACKPZ strickland_n_Page_22.pro
a6ac1635a92adea66e8d34a78a7d0ab6
ab2fa9c7fcf100a23c5ae2a87921892ae8a94d2e
41271 F20101203_AACKZU strickland_n_Page_44.jpg
ad196e408b5dfc658e68c4aaf076b768
b23fcf5aadeb477f38c2cda22c99a294c2c9c7d2
21030 F20101203_AACKUW strickland_n_Page_69.pro
4f7c4b11356eb0a63f89a44aa1c0b39a
dab429185ed27ea6fda7ac8681449dac75601fd8
F20101203_AACKNB strickland_n_Page_75.tif
1f48030d60ec860d739b3f9715e7e0a6
950d8c1010f80b566b472353fe703a588071410d
25354 F20101203_AACLGS strickland_n_Page_62.QC.jpg
5e5550de77236036b40560f44eb9bd93
af4d12c8b1f242fae1848e4e057ed52f0f6074e4
F20101203_AACLBV strickland_n_Page_16.tif
4b48a037187b8be9c61f529b03f78afe
33d7e0c9d252539b0a8fbbf57cc7ff7634afd439
34064 F20101203_AACKZV strickland_n_Page_45.jpg
5855bcee6207fd04fae1c8f6082b30ee
77775c3eeeb9535f4cb91832e8868a571ae0f8c6
F20101203_AACKUX strickland_n_Page_13.jp2
51a96d74323ad75495ca09eaab27d41a
7d23da024ec1bea35c661f416ebc23d37e501560
4652 F20101203_AACKNC strickland_n_Page_65thm.jpg
59b94fd721fa661a0cc00771651e2139
7a441393bebac1990a00a64a7b26e2b2af48ceec
27760 F20101203_AACLGT strickland_n_Page_63.QC.jpg
d74211aa4f106e4a4fec98d28d64a4a1
9906bf9472841c09821aa6479a0f593496237844
F20101203_AACLBW strickland_n_Page_18.tif
60b15f1549610fa39888260eba888d58
925054eefc883f166647567e5f632859236ac47c
51134 F20101203_AACKZW strickland_n_Page_46.jpg
ff2bda4b01f98e0612177a8f8d6451cb
2a438c1bbaac6dea1b7e19411e24cb120c9fdefd
65881 F20101203_AACKSA strickland_n_Page_52.jpg
04cfbad87bedaf55a9375a561314492a
f444fa32dbf132af6c76bfd3c96b8c23293a5265
10565 F20101203_AACKUY strickland_n_Page_44.pro
130f5f70dd3531a73c432d034637cbf1
e337f0c6dfb51e0a3b114a782257006d07717505
F20101203_AACKND strickland_n_Page_59.tif
c86beaecd7ecd8215ef7a2b8404d6d12
c9dac145e25a90c1aeb636660bf5f05558a8e64d
4165 F20101203_AACLGU strickland_n_Page_64thm.jpg
c8afcb44990dcf7859d9dea44c4281d3
5066f152755808d14f4756a264895a42c8364d29
F20101203_AACLBX strickland_n_Page_19.tif
3abf82a6e27548668aea4d8b6b810f85
3ce7fdaae6a084f10462d3f72bd075109411bb2f
42814 F20101203_AACKZX strickland_n_Page_50.jpg
845f936067c8ed90cf3b2f67d873be40
c36ed68398bba2e3b1c5af4e2a0d834d8fd3709e
F20101203_AACKSB strickland_n_Page_14.tif
56797649e5875409ba162e92d0328920
d32fc3901e89f4e556af3cee7e9b6483f3b9aae3
F20101203_AACKUZ strickland_n_Page_28.tif
b7347b9bf79f933d660c07edb77a18dc
3886acb861215d2a9413c576665b6c5fdf887212
26670 F20101203_AACKNE strickland_n_Page_01.jpg
3d14695a9838bb041b991dee76b7fdd8
3ebf5660d1cdc67e86177024c31a10a68e3fe4d6
16963 F20101203_AACLGV strickland_n_Page_65.QC.jpg
2af6918161c7cb5dc27b222c56c8849c
da9000b76c77fd1fa261e93a0b6810c9332f0b9a
F20101203_AACLBY strickland_n_Page_22.tif
21635babebedb347d3bb284e6524665f
7d783f0cafec4f63a00ac85ea3c48d7928b88ca4
63057 F20101203_AACKZY strickland_n_Page_51.jpg
3aa4f8d30cc79cb6901fb1984b4cfe8f
bc40955c6b48e0b4276a6bcd2d6259df613b3c56
2344 F20101203_AACKSC strickland_n_Page_16.pro
60f1bb48557a6948092880f1b5097f91
a73a313ba6452577a1988df4cb5761b8d835e5de
2771 F20101203_AACKNF strickland_n_Page_09.txt
4b145a7dad7058ebad02435b9ec3afc4
680e68b4c078edc34af3b62c4b99f300b4ff2ff2
17462 F20101203_AACLGW strickland_n_Page_66.QC.jpg
8f06e70d2233b058a2e20add73cc32e2
573dd1b280f40528d17e17d2f5a21465546d8397
F20101203_AACLBZ strickland_n_Page_26.tif
4dd41447f57d41473169fecb236173c9
9dcd3235fc686001536182ed110fdbf5e6e54a2e
28820 F20101203_AACKXA strickland_n_Page_38.jpg
bbb954c5f175953b7308681a9ca130d8
0163e53a0c44a34657d9317a5cbb70db80bb7709
72762 F20101203_AACKSD strickland_n_Page_06.jpg
47e025c8a41f9d9fcb8ee5d6821e16e5
8f65835478c5d0f1a5cf423801062d2506de63b3
F20101203_AACKNG strickland_n_Page_76.tif
25da272cac84d97e3524ee8db09f1b65
6a81f32fedd0528b19fee9203036306910d0e617







RESIDUAL STRAIN MEASUREMENT OF PLAIN WEAVE COMPOSITES USING THE
CURE REFERENCE METHOD




















By

NANCY M. STRICKLAND


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


2007
































2007 Nancy M. Strickland





























In memory of my mother.
Also to my family, since none of this would have been possible without their encouragement and
support.









ACKNOWLEDGMENTS

I would like to thank everyone who provided support and assistance throughout this

research. First, I want to thank Dr. Peter Ifju for all of the advice and support he has offered. I

would also like to thank my other committee members, Dr. Bhavani Sankar and Dr. Raphael

Haftka, for their advice and expertise.

Also, this work would have not been possible without the help and shared knowledge from

Weiqi Yin, Tzu-Chau Chen and Diane Villanueva, my colleagues in the Experimental Stress

Analysis Lab. Finally, I would like to thank my mom, dad and the rest of my family for all of

their support and encouragement they have provided throughout the years.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

LIST OF TA BLE S ................. .......................................................... 7

LIST O F FIG U RE S ................................................................. 8

A B S T R A C T ........................................... ................................................................. 1 1

CHAPTER

1 IN TR O D U C T IO N ................................................................................ 13

W oven Composites ....................................................................... ......... 13
R esidual Stresses and Strains ...........................................................................................14
Research Objectives ............................................. 15

2 L IT E R A T U R E R E V IE W ................................................................ ...............................17

In tro du ctio n ................... ...................1...................7..........
D destructive T techniques ................................................................17
H ole D killing M ethod ...............................................................17
F first P ly F ailu re T e st ................................................................ ...............................18
N on-D destructive Techniques .............................................................18
X -Ray D iffraction ............................................. 18
E m b ed d ed S en so rs ..................................................................................................... 19
W arpage M easurem ents .................................................................... 20
Cure R reference M ethod........................................................................ .. ..............20
Analysis Methods Used in Predicting Residual Stresses in Woven Composites ...................20

3 CRM DEVELOPMENT .................. ........................ ............................ 22

Principles of the Cure R reference M ethod..................................................................... ...... 22
Moire Interferometry ................................. ...... ... .................. 23
Specim en P reparation .............................. ...................................................................24
D evelopm ent Process................................................................................. 25
Original Cure Reference Method .............................................................25
Original cure reference method procedure ...............................................................25
Results using original CRM ......................................................... ............... 26
Silicone Rubber Based M methods .................................. .....................................28
Silicone rubber grating using resin from prepreg ............. .......... .................. 28
Results using silicone rubber grating using resin from prepreg ...............................29
Modifications attempted: Thin silicone rubber grating ........................................29
Modification attempted: Additional epoxy cured onto silicone rubber ..................30
Results from silicone rubber and 3501-6 .................................. ..................31









Kapton Polyimide Film Based Methods...... ................. ...............31
Results using the Kapton film method ......... ...... ............. ......... ............. 32
Modifications attempted to the Kapton method............................... ...............33
H igh Tem perature Epoxy B asked M methods ........................................... .....................33
Pooling the epoxy onto the grating ........................................ ....... ............... 35
Spreading the epoxy onto the grating................................. ...............35
Spreading the epoxy onto the prepreg before immediately applying to grating ......36
Spreading the epoxy and letting it set before placing on autoclave tool ................37
C onclu sions.......... ............................... ................................................37

4 RESIDUAL STRAIN MEASUREMENTS ........................................ ....................... 51

P h a se S h iftin g ................................................................................................................... 5 1
R e su lts ................... ...................5...................2..........
C o n clu sio n ................... ...................5...................3..........

5 FEM SIMULATION OF PROCESS INDUCED STRAINS ............ ............... 59

M motivation for Perform ing FEM A analysis .................................................. .....................59
M odel D description ............................................................................. 59
Sim ulating the C during C ycle ............................................................... .............................60
Using FEA to Validate the Experimental Method...... ................. ............61
Experimental Process Induced Residual Stresses...........................................62
C onclu sions.......... ............................... ................................................63

6 CONCLUSIONS AND FUTURE WORK............................. .....................73

C o n c lu sio n s.............................................................................. .7 3
Future W ork .............. .................... .................................... .................. 74

APPENDIX

CRM IMAGES AND PHASE SHIFTED RESULTS.....................................................76

L IST O F R E F E R E N C E S .................................................................................... .....................80

B IO G R A PH IC A L SK E T C H .......................................................................... .. .......................83










LIST OF TABLES


Table


5-1 Material properties used in ABAQUS model for the plain weave composite RVE..........65


page









LIST OF FIGURES


Figure page

1-1 Representative volume element of the plain weave composite geometry .........................16

3-1 Schem atic of a typical four-beam interferom eter ........................................ ...................39

3-2 Fixture used to align the diffraction grating with the autoclave tool ..............................39

3-3 Oven and vacuum lines used for the curing cycle .................................. ............... 40

3-4 V acuum bagging assem bly ........................................................................ ..................40

3-5 Cure cycle used for HM F plain weave prepreg ...................................... ............... 41

3-6 Tool used to separate master grating from the autoclave tool .......................................41

3-7 Tools required to m ake Teflon device ........................................ .......................... 42

3-8 Specimens made using Original CRM procedure. ................................ .................42

3-9 Specimen made using original CRM with successful transfer. ............................43

3-10 The shapes of the diffraction gratings when the master grating is hard from a high
temperature epoxy (a) and when a silicone rubber grating is used as the master
g rating g (b )............................................................................ 4 3

3-11 Fiber reactions with different silicone rubber thicknesses. ............................................44

3-12 Silicone rubber grating showing exposed fibers indicating that the grating surface
w would not be perfectly flat ....... .. .......... ........... ................................ ............... 44

3-13 Specimen made using thin silicone rubber grating ............... ............... ............... 45

3-14 Grating production procedure for silicone rubber with cured 3501-6 resin ....................45

3-15 Specimen made using the silicone rubber with cured 3501-6 resin................................46

3-16 Moire images taken from the specimen shown in Figure 3-14. .......................................46

3-17 Stacking method used for the Kapton polyimide process......................................47

3-18 Specimen made using the Kapton polyimide method............................. ...............47

3-19 Moire images taken for the specimen shown in Figure 3-17. ........................................47

3-20 Specimen showing the locations of strain discontinuities occurring on peaks of fiber
bundles (a) and the correlated location within the RVE (b). ...........................................48









3-21 Various specimens manufactured with different polyimide materials ...........................48

3-22 Procedure used to produce base grating for high temperature epoxy methods .................49

3-23 Specimens produced by spreading a thin epoxy layer directly onto the aluminized
g rating g su rface ................................. ......................................................... ............... 4 9

3-24 Pooling the epoxy onto the prepreg as done in the high temperature epoxy method ........50

3-25 Sample specimens made after allowing the epoxy to begin to set on the prepreg.............50

4-1 Fringe images for Specimen 1 for the U-field(a) and V-field (b)................................55

4-2 F1 strain contour map via phase shifting from Specimen 1.............................................55

4-3 S2 strain contour map via phase shifting from Specimen 1............................. 56

4-4 Fringe images for Specimen 2 for the U-field (a) and V-field (b)............................... 56

4-5 S1 strain contour map via phase shifting from Specimen 2.............................................57

4-6 S2 strain contour map via phase shifting from Specimen 2............... ............... 57

4-7 The representative volume element of the plain weave geometry.................................58

5-1 Simulated process induced residual strains in the 1-direction (s1) resulting from
cooling the RVE from Tref to room temperature and introducing it to atmospheric
pressure. Region A is the resin rich zone and Region B has a lower resin content..........65

5-2 The contour plot of Si after the simulated cooling cycle to room temperature ..............66

5-3 Plot of F1 along the length of the midline of the warp of the RVE in the 1-direction. .....66

5-4 FEA Si results for the RVE from the simulated curing cycle obtained ...........................67

5-5 Experimental phase shifted e1 resulting from the cooling cycle................. ....... ........ 67

5-6 Plot of the residual strain, ei, along the resin zone in the 3-direction .............. ...............68

5-7 FEA 83 results for the RVE from the simulated curing cycle ........................................69

5-8 Experimental phase shifted 83 resulting from the cooling cycle ....................................69

5-9 Stress contours for the 1- and 3-directions using experimental displacement data..........70

5-10 The stresses are plotted along the fiber directions in along two different bundles............71

5-11 The stresses along the resin zones are shown along and transverse to the resin zone. ....72









A-1 Silicone Rubber and 3501-6 Resin, U Field ........................................... ............... 76

A-2 P1 phase shifted results for silicone rubber and 3501-6 resin......................................76

A-3 Silicone rubber and 3501-6 resin, V Field.................. ....... .............. ............... 77

A-4 82 phase shifted results for silicone rubber and 3501-6 resin......................................77

A-5 High temperature epoxy spread and set on prepreg, U Field......................................78

A-6 S1 phase shifted results for High temperature epoxy spread and set on prepreg ...............78

A-7 High temperature epoxy spread and set on prepreg, V Field......................................79









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

RESIDUAL STRAIN MEASUREMENT OF PLAIN WEAVE COMPOSITES USING THE
CURE REFERENCE METHOD

By

Nancy M. Strickland

December 2007

Chair: Peter Ifju
Major: Aerospace Engineering

Residual stresses develop in composite materials due to the high temperature

manufacturing process that is required. As the structure cools from the curing temperature, those

stresses develop because of a mismatch between the coefficients of thermal expansion values of

the two constituent materials of the composite. If the residual stresses are not accounted for

when designing a composite structure, premature failure can develop due to the lack of available

strength remaining in the structure. A significant amount of work has been focused on

determining these stresses for unidirectional composites, but as the woven geometries become

more popular, the research must shift.

Woven composites consist of a complex geometry, so because of that, a full-field

measurement technique must be used to obtain the stress patterns throughout the repeating unit

of the geometry. The cure reference method was designed to measure surface residual strains of

unidirectional composites with very high sensitivity. That method is extended to the woven

composite; however, to do so modifications are required to achieve optimal results. The final

procedure that is chosen to employ for this research is a less time intensive and more repeatable

process than the original method suggested for the original embodiment of the CRM. The

experimental data show that significant strains develop throughout the repeating unit and the









strains alternate between highly tensile to highly compressive over a very small distance. The

approximate range of strains that are occurring is between -3000te and +2500tE.

A finite element analysis is also completed to verify the experimental method developed

and to calculate the residual stresses resulting from the experimental residual strains. The

verification analysis shows very good agreement between the analytical and experimental data.

The tensile strains are occurring over the fiber bundles and the compressive strains are occurring

in the resin rich zones. The range of strains that is occurring experimentally is larger than that

predicted by the analysis but that is due to some assumptions made in the finite element model.

The residual stresses that are developing are between -13 MPa and +4 MPa which are small with

respect to the strength of the resin (82.7 MPa) but they cannot be ignored.









CHAPTER 1
INTRODUCTION

Composites are composed of two or more different materials combined to form a single

macroscopic structural unit. A common reason for manufacturing a composite material is so that

a combination of desired properties can be achieved yielding optimal results. It has been shown

in studies performed by Griffith that materials in the fiber form yield higher strengths than they

do in the bulk form [1]. However, it is also very difficult to use fibers structurally since they

cannot maintain compressive loads and the properties transverse to the loading direction are very

poor. This can be solved by embedding the fibers in a matrix, or filler material, so that many

fibers bound together by the surrounding matrix can be used to carry the load. Ultimately, the

structural advantages of fibrous materials, such as stiffness and strength, can be used in a low

weight structure yielding a part with a high specific stiffness or strength.

Composite materials are becoming more widely used by various industries. As the desire

increases for advanced material properties that composites provide, more research needs to be

continually done so that they can be fully characterized and understood.

Woven Composites

There are two categories of composites, plies composed of unidirectional or randomly

scattered chopped fibers (laminates), or fiber bundles woven into a regular pattern (textiles).

Unidirectional composites provide strength and stiffness properties in the direction that the fibers

are aligned. By stacking multiple plies and aligning the fibers in different orientations, material

properties can be enhanced in desired directions for the part being designed. The second class of

fiber reinforced composites is the woven textile. Similar to the unidirectional composites, the

weave pattern can be adjusted so that certain material properties are achieved, particularly if

more strength/stiffness is desired in one direction as opposed to another. Additionally, the









overall stiffness or strength of the part can be enhanced by stacking multiple layers to create a

thicker structure. Several weave patterns exist, including the plain weave and various satin

weaves. The focus of this work will only be on the plain weave geometry.

As opposed to the unidirectional composite laminates, woven composites exhibit a very

complex geometric structure that results in property trends very different from many other

materials. Most material classifications yield a uniform strain field during tension or

compression tests as long as strain concentrations do not exist in the region of interest. In these

cases, the average strain is a reasonable estimate of the local strains. However, this is not the

case in woven composites because of the complex geometry exhibited by the fibers. Because of

this, even though the large scale averaged strains may appear uniform, the strains on the scale of

the representative volume element (RVE), shown in Figure 1-1, are not uniform. Typically,

under most loading conditions a repeating pattern would occur in the strain field indicating very

different strain values at various locations in the RVE. A characteristic of woven composites

that is distinctly different from most materials is that in a given loading condition, it could be

possible for the structure to exhibit both tensile and compressive strains throughout the RVE.

For example, moire interferometry was used to measure the strains transverse to the loading

directions on a plain weave composite during a uniaxial tension test. Repeating regions of

tensile strain were measured across the specimen even though typically only compressive strains

would be expected to develop because of the Poisson effect [2].

Residual Stresses and Strains

The manufacturing process for many composite structures requires a high temperature

heating cycle. Those high temperatures are required so that the resin can fully heat and complete

its polymerization process as well as wet the fibers and finally cure into a hard structure.

However, this necessary process introduces a significant amount of residual stresses into the









composite specimen after the heating cycle completes and the specimen cools back down to

room temperature. The residual stresses in a composite arise from both chemical and thermal

shrinkage [3].

The chemical shrinkage is due to polymerization of the resin in which the two monomers

in the epoxy come together to form the final compound. The majority of the chemical shrinkage

occurs during the initial heating period of the curing cycle and therefore does not contribute to

the overall residual stresses. This is because stresses cannot exist before the composite fully

cures and the resin transforms to the solid state. However, there is some additional chemical

shrinkage that does contribute to the final residual stress levels in the specimen that occurs after

the solidification of the resin.

On the ply level, a significant amount of the residual stresses develop because of the

mismatch that naturally exists between the coefficients of thermal expansion (CTE) of the two

materials. The resin typically has a much higher CTE value than the fiber; therefore a larger

amount of thermally induced shrinkage would develop in the matrix. However, the higher

strength fibers restrict the contraction of the matrix. That restriction is what introduces the

residual stresses into the composite.

Research Objectives

The goal of this research was to employ an experimental method to measure the residual

strains that develop in the plain weave composite due to the manufacturing process. Because of

the complex nature of the material, a full-field experimental method was desired so that a strain

contour map could be obtained for the representative volume element of the weave pattern. The

process induced residual stresses were to be obtained via finite element modeling by using the

experimental displacement data as inputs into the analysis. Additionally, the finite element

model will be used to perform a comparison study to the experimental data.




















Figure 1-1. Representative volume element of the plain weave composite geometry









CHAPTER 2
LITERATURE REVIEW

Introduction

Many methods have been used to experimentally measure or analytically predict the

residual stress levels of composite materials. The experimental methods can be divided into

destructive and non-destructive techniques.

Destructive Techniques

Hole Drilling Method

The hole drilling technique is one of the oldest and most widely accepted methods for

measuring residual stress levels of most materials. It was first discovered by Mather in the

1930's that by removing a section of material in a stressed structure, a change in stress state in

the remaining material will occur so that static equilibrium can be maintained [4]. Doing so

results in the material surrounding that hole to relax therefore altering the dimensions of the hole

which correlates with the surrounding strain field.

Much advancement to the hole drilling method has been made since the origin of the

technique. The development of the hole drilling strain gage rosette has allowed more accurate

measurements of the area directly surrounding of the blind drilled hole [5]. Also, as technology

has improved, the gages have become smaller giving the ability to better capture the large strains

that are developing close to the hole edge that cannot be measured due to the significant

averaging larger gages must perform. The hole-drilling method has been applied to orthotropic

materials by developing intensive analytical relations between the stress and strain fields [6,7].

Because of the extensive calculations required for the application to orthotropic materials, either

many assumptions must be made to simplify the calculations or a finite element analysis must be

performed to gather the coefficients required by the analytical relations. By introducing









interferometric methods to the hole drilling procedure, the strain fields surrounding the hole

could be measured as opposed to a averaged values close to the hole [8,9,10]. However, even

those methods have difficulty accurately measuring the large stress concentrations around the

hole. Although much advancement has been made in this technique, there is a major drawback

associated because of the lack of full-field information it provides.

First Ply Failure Test

The residual stresses that exist in a composite can also be found by the first ply failure test

[11,12]. A [0/90] specimen was loaded until failure, and throughout the loading the strains were

recorded. Using elastic stress-strain relations, the approximate load was calculated at which the

first ply failed in the test. That value was then compared to the ultimate strength of the ply, and

the difference was called the residual stress.

Non-Destructive Techniques

X-Ray Diffraction

X-Ray diffraction was introduced into the composites community as a possibility for

measuring residual stresses by Predecki and Barrett in 1979 [13]. By embedding metallic

particles between prepreg plies, the residual stresses of the particulates could be obtained using

X-Ray diffraction. The stresses of the particulates could then be related to the surrounding

matrix material to determine the stress values of the matrix. This methodology was then

extended to measuring the residual strains of the composite and using numerical modeling via

the Eshelby method to determine the residual stresses [14]. After having success with the

unidirectional laminate, X-Ray diffration was used for measuring the interlaminar residual

strains of woven composites [15]. A drawback to this method is that it requires embedding a

foreign object into the matrix which could ultimately modify the material properties.









Embedded Sensors

Two different types of sensors are commonly embedded in composite materials to measure

residual strains: electrical resistance strain gages and fiber optics. By embedding the sensor

between layers of the composite laminate, the deformation experienced by the specimen is

recorded by the sensor and after using the necessary correction factors the strains can be

obtained.

Electrical resistance strain gages were first embedding within laminated composites to

measure the residual strains in the 1970's [16,17,18]. The gages were placed between plies of

the laminate and on the surface of the specimen during the manufacturing process prior to cure.

By doing this, any strains that developed during the curing process would be measured by the

strain gages. Daniel made sure that the gages and required wiring were chosen so that the

measurement equipment did not introduce any local thickness differences to limit the

introduction of errors in the strain results. The method of embedding strain gages within the

composite specimen has also been used to help determine the warpage developed in asymmetric

woven laminates combined with laminate theory and shadow moire [19].

The ability to measure residual strains by embedding fiber optics within composite

laminates was developed under the same principle as embedding electrical strain gages, and was

first used by Lawrence [20,21]. Both an extrinsic Fabry-Perot interferometer and a fiber Bragg

grating were used to measure the process induced strains in a composite laminate. The fiber

optic is laid between plies before cure and is nulled at room temperature. Assumptions are made

that the optic bonds well to the epoxy and that it does not structurally reinforce the specimen.

Work has also been done to extend the use of fiber optics to woven composites by using a

distributed sensor system [22].









Two main drawbacks of using embedded strain sensors are that a foreign object must be

placed within the specimen and that only localized strain values can be obtained. The

introduction of the foreign object is always assumed not to reinforce the material or alter the

properties in the area of interest, however, some errors although minimal are bound to develop.

Warpage Measurements

When assymetric composites are manufactured, warpage will ultimately result. Much

investigation has been done on determining the warpage of composites and relating that to the

residual stresses [18,23]. However, this method is not desirable because it is seldom that an

assymetric composite is desired and due to the inherent symmetric nature of plain weave

composites, measuring warpage is not feasible.

Cure Reference Method

The Cure Reference Method (CRM) was developed at the University of Florida as a

nondestructive technique to measure full-field residual strains of composite materials [24]. By

applying a diffraction grating to the specimen at the cure temperature, which is assumed to be a

free stress state, moire interferometry can be used at room temperature to measure the residual

strains that develop from the curing process. This technique is the basis of this work, so an in

depth description will follow.

Analysis Methods Used in Predicting Residual Stresses in Woven Composites

For traditional laminate composites, the classical laminate theory (CLT) can be used to

calculate the stresses in a structure once the material properties and strain data are known.

However, in advanced composites, such as woven textiles, the complications of the material

geometry no longer allow these straightforward calculations. Both analytical and finite element

models have been used to predict the process induced strains in woven composites. Several

groups have worked to develop analytical models based off of CLT that account for the complex









geometries contained in woven composites [25, 26, 27]. Both TEXCAD and MESOTEX have

been extended to predict the process induced thermal stresses for the plain weave composite.

Other work has been done by using finite element models to predict the residual stresses in

woven composites, however; some of these models are still either simplified two- or three-

dimensional models that are not able to completely predict the stress field occurring throughout

the geometry [28, 29].









CHAPTER 3
CRM DEVELOPMENT

The focus of this research was to modify and improve the original process proposed for the

Cure Reference Method [15]. This technique was then to be used to measure and obtain the

residual strain field of a plain weave composite. This chapter will go into the details of all

methods that were attempted in modifying this experimental technique and explain what was

desired when creating a grating to be used for the cure reference method.

Principles of the Cure Reference Method

The cure reference method (CRM) works in conjunction with moire interferometry to

measure the residual strains that develop in a composite material because of the high temperature

curing cycle required for fabrication. During the curing cycle, the composite undergoes a

transformation process from its original liquid state as a prepreg to the solid state once the resin

has cured. Until the point it transforms to a solid, it can be assumed to be in a stress-free state.

By attaching a diffraction grating to the composite surface at the point that transformation

occurs, any changes in the frequency of the grating will only be attributed to the deformations

that develop as the specimen cools to room temperature.

The diffraction grating was replicated directly onto the specimen from the autoclave tool,

which was a piece of ultra low expansion glass so to ensure that the grating frequency would be

the same at the cure temperature as it was at room temperature. The two main assumptions that

this method employs were that the grating bonds perfectly to the specimen surface and that the

bonding undergoes the same process as the solidification of the resin in the composite. Also,

because it is assumed that the grating adheres to the specimen at the transformation point to the

solid, the exact time the adhesion occurs does not need to be known, since the room temperature

measurements will always be in reference to the free-stress state.









After the specimen was given the appropriate time to cure, dictated by the curing cycle, the

composite was separated from the autoclave tool before the cooling cycle began. This allowed

the specimen to undergo free contraction as it cooled and ensured that the frequency change

measured in the grating was due to the residual strain occurring from the cure temperature.

Moire interferometry is then used after the specimen completely cools to room temperature to

measure the residual strains.

Moire Interferometry

Moire interferometry is a laser based optical method that is used to measure full-field

strains with extreme sensitivity [2]. The sensitivity of this method is 0.417[tm, and it provides a

contour displacement map throughout the area of interest which is constructed via interference

from light diffracting off of the specimen. A diffraction grating is adhered to the specimen

before tests are conducted and it has an initial frequency matching the reference, or master

grating. As the specimen deforms, the frequency of the grating changes and that change is

directly proportional to the deformation that occurred. A four-beam interferometer was used for

this work and is shown schematically in Figure 3-1.

To measure the changes of frequency that occur, the interferometer has to be tuned to the

initial, or reference, grating. This is done by directing the light sources towards the specimen so

that no, or very few, interference fringes can be observed. Then when the specimen replaces the

reference grating in the interferometer, interference of the light sources will occur because of the

changes in the grating frequency. As it is applied to CRM, after the composite cools, the residual

strains that develop due to the cooling process can be measured by comparing the specimen

grating to the reference grating in the interferometer.









The interferometer used in this work was a self-contained, portable moire interferometer

(PMI) in the Materials Characterization Lab of the Civil Engineering Department. It was a four-

beam, 2400 lines/mm interferometer that consisted of a 628nm He-Neon laser.

Specimen Preparation

The material tested for this research was a plain weave composite manufactured by Cytec

Industries. The typical test specimen was a lay-up composed of four 3.0" x 3.0" layers. While

stacking the prepreg layers, careful attention was directed towards aligning the fiber bundles with

each other so to obtain minimal misalignment. It was critical to obtain the best alignment

possible so that the strains measured via moire interferometry corresponded with the strains

along the fiber directions. Also, prior to bagging the specimen and autoclave tool, the alignment

of the grating needed to be determined with respect to the autoclave tool. Using the fixture

shown in Figure 3-2, a laser source was reflected off of the grating surface and the diffracted

beams were aligned with the markings seen in the figure by adjusting the position and rotation of

the autoclave tool with respect to the optical bench.

The specimens were required to be cured in a vacuum oven that is shown in Figure 3-3.

Irregardless of the CRM production method used, to prepare the specimen for the vacuum curing

cycle it was bagged according to the process described below and shown schematically in Figure

3-4.

1. A circular hole was cut into a piece of non-porous release film large enough to expose the
grating on the autoclave tool. That piece of film was then laid directly onto the autoclave
tool leaving the grating exposed.

2. The prepreg was then placed over the release film so that the fiber bundles aligned with
the diffraction grating and so that the grating was exposed to the composite surface.

3. The composite was then covered with a porous release film, another non-release film
layer and then a breather cloth that was used to pull an even vacuum over the entire
specimen.









4. The autoclave tool and respective layers were then placed inside a sealed air tight vacuum
bag that was connected to the vacuum line of the oven and checked for air leaks before
beginning the cure cycle.

After bagging, the specimens were cured according to the cure cycle stipulated by the

manufacturer shown in Figure 3-5. The critical step when implementing the CRM is separating

the specimen from the autoclave tool after it has fully cured, and that point is also shown in the

figure. However, that separation must be done before the cure cycle reaches the cooling process

so that the entire cooling regime can be measured in the deformations. If the specimen is not

separated from the tool, it is not allowed to freely contract and any strains that are measured

would be from an unknown temperature.

Development Process

When work initially began on this research, a few CRM gratings were available that were

produced using the process described in the original CRM method. However, after those were

used, new gratings needed to be produced for more experiments to be conducted. During that

replication it was determined that the existing procedure needed to be modified so that it could

be applied to woven composites and so that it would require a less time demanding procedure to

produce a single grating.

Original Cure Reference Method

Original cure reference method procedure

The original fabrication method proposed for CRM was a time intensive process that

required approximately sixty hours, over a total of seven steps, to produce a single autoclave tool

[24]. After the four available gratings were used to manufacture test specimens, new gratings

had to be produced. The following procedure was used to replicate that process.

1. A silicone rubber grating was replicated at room temperature onto a piece of ultra low
expansion glass (3.0" x 4.0" x 0.5") from a Photoresist master diffraction grating. The
silicone rubber was a two-part mixture, GE 615RV.









2. The master silicone grating was used to replicate the diffraction grating onto another
piece of Astrosital (3.0" x 4.0" x 0.5) using a high temperature epoxy (Epon 862 with
curing agent W). This would become the intermediate grating. The parts were mixed
together using a mass ratio of 100:26.4 and then heated for five minutes at 1300C.
Afterwards it was centrifuged for five minutes to remove any bubbles before being
applied to the grating for replication. The epoxy was then allowed to cure for ten hours
in a 1300C oven before the gratings could be separated. The gratings were separated
using the device shown in Figure 3-6.

3. Two aluminum layers were deposited onto the grating surface using vacuum deposition
with a film of Photo-flo used as a parting agent between the two layers. The Photo-flo
was allowed to sit for twenty-four hours before the second layer of aluminum was
applied.

4. The grating was then replicated onto the autoclave too, a piece of Astrosital glass (5.0" x
5.0" x 0.5") using an epoxy resin, 3501-6. The resin was heated for ten minutes at 1750C
before a vacuum was also applied to the oven for an additional fifteen minutes. While
the resin was heating, the intermediate grating and the autoclave tool were also placed
into the oven to preheat. After liquefying the resin, it was then pooled onto the autoclave
tool and the intermediate grating was lowered onto it for replication. The resin was
allowed to cure for ten hours at 1750C before being separated.

5. A single layer of aluminum was applied to the autoclave tool over the resin layer.

6. A piece of Teflon was tightened over a piece of glass using the device shown in Figure 3-
7. The device was then placed into the oven for an hour at 1750C and the Teflon was
tightened at several points in that time to ensure a smooth surface. After the Teflon was
tightened, the 3501-6 resin was heated and vacuumed using the same procedure as in
Step 4. Afterwards a small amount was pooled onto the aluminum surface of the
autoclave tool and the Teflon device was carefully lowered down onto it so that no air
bubbles were trapped. The autoclave tool and Teflon device were then heated in a 175C
oven for ten hours. After that point the Teflon device was unscrewed and the Teflon was
peeled off of the epoxy at a steep angle so that the epoxy does not separate from the
aluminum.

Results using original CRM

After following the above procedure, grating production was approximately a thirty

percent success rate. The method was extremely sensitive to the thickness of the final layer of

resin applied to the autoclave tool. If that layer was allowed to be too thick, as it cooled it would

crack at several locations in the grating. In several instances, as the epoxy layer cracked it would









chip the glass. When this occurred, that grating was no longer usable and the six step process

previously described needed to be completed again.

Four specimens were made using the pre-fabricated gratings, and fringe images were

captured. The number of fringes was drastically lower than expected and the fringes were

indicating almost zero strain. It was thought that the epoxy on those gratings was too stiff,

therefore not allowing any shrinkage to occur. Therefore, new gratings were fabricated using the

described procedure and those would be used for testing.

When specimens were produced using the gratings made from the procedure explained

above, several issues arose when transferring the grating onto the composite and separating the

composite from the autoclave tool. Some examples of specimens produced from these gratings

are shown in Figure 3-8. In most gratings that were produced, the grating was either severely

cracked, as in Figure 3-8 (a), making it impossible to obtain usable fringe images or the grating

transfer was very incomplete as in Figure 3-8(b). It was believed that the reason for the poor

grating transfer was due to insufficient resin available in the prepreg. The grating did

successfully transfer in a couple tests. However, due to the low resin volume present,

discontinuous fringes developed at the locations which should be resin rich zones. As can be

seen in Figure 3-9(a), those areas are completely void of resin in many locations, and because of

that, discontinuous fringes developed as shown in Figure 3-9 (b).

However, even though a few gratings were successfully produced, because of the difficulty

and time required for the procedure it was decided that efforts should be targeted towards

modifying the process to one that would be more repeatable and less time demanding. The

following sections describe the methods attempted in achieving that goal.









Silicone Rubber Based Methods

As opposed to other methods that were tested, this technique initially required only one

grating replication. A silicone rubber grating was replicated onto the autoclave tool and was

used to transfer the CRM grating onto the specimen. Philosophically, the silicone rubber grating

would provide two distinct advantages towards improving the CRM procedure. Firstly, the use

of silicone rubber would provide easy separation between the autoclave tool and the composite

specimen. By improving the separation it helps to ensure that the grating surface will separate

cleanly so that smooth and continuous fringe information could be obtained. Also, if the gratings

are difficult to separate, some bending of the specimen would inevitably occur which could

introduce erroneous information into the diffraction grating. Secondly, the production process

for the grating would be drastically reduced, by using silicone rubber, therefore shortening the

time required for one test.

Silicone Rubber Grating Using Resin from Prepreg

The first attempt at this method assumed that there was sufficient resin in the prepreg. By

making this assumption, as the composite cured, the resin within the prepreg would replicate the

diffraction grating without any additional epoxy being applied. The detailed procedure follows

the steps below:

1. A diffraction grating was replicated using a two part silicone rubber onto a piece of
Astrosital ultra low expansion glass at room temperature using a Photoresist master
grating. The silicone rubber (GE 615RV) was carefully mixed and centrifuged before
use. Three days were allowed for the silicone rubber to fully cure at room temperature.

2. The prepreg was then placed directly on the silicone rubber grating before being placed
into the autoclave. When the resin of the prepreg cured, the diffraction grated would be
formed in the resin of the composite surface.









Results using silicone rubber grating using resin from prepreg

When silicone rubber was used to replicate the grating onto the autoclave tool, the

resulting fringe image was not able to be obtained using the interferometer because of the lack of

a smooth reflective surface on the specimen. This occurred because the silicone rubber was not

hard enough to prevent the fibers in the prepreg from indenting the grating. Ideally, the grating

surface should appear similar to that in Figure 3-10 (a) so that all light directed towards the

grating is reflected directly back towards the camera. However, when a soft medium, such as

silicone rubber, is used to transfer the grating, the fibers of the composite cannot remain

perfectly rigid and eventually sink into the rubber grating as shown in Figure 3-11. Therefore,

the resulting grating surface appears similar to that shown in Figure 3-10 (b). As can be seen in

the figure, because the grating surface is not completely smooth, the light is redirected when it

reflects away from the surface providing an incomplete and inaccurate fringe image at the

camera. This can also be seen by visually inspecting the grating on the specimen, Figure 3-12, as

one can see individual fibers in the grating area which indicated that the grating surface was not

perfectly smooth.

Modifications attempted: Thin silicone rubber grating

An attempt was made to create a thin silicone rubber grating that would behave stiffer so

that a flatter grating surface could be created on the specimen. It was believed that if the

thickness of the silicone rubber was of a similar magnitude to that of the fiber diameter,

significant indentation would be inevitable yielding a grating surface similar to that achieved in

Figure 3-11 (b). However, if the thickness of the silicone rubber was held to a magnitude much

less than that of the fiber diameter as in Figure 3-11 (d), it was hypothesized that the grating

surface would flatten out and yield a more complete fringe image. As can be seen in Figure 3-









13, making this modification did not improve the grating surface and a moire image was still

unattainable.

Modification attempted: Additional epoxy cured onto silicone rubber

The next modification was to add an additional layer of epoxy between the grating and the

prepreg during the lay-up process. This additional epoxy was pooled onto the silicone rubber

grating immediately before the prepreg was laid down. By doing this, it was thought that better

replication could occur between the reference grating and the specimen due to the addition epoxy

available to form the grating. However, this proved not to be the case after attempting this

procedure after using both 3501-6 epoxy-resin and a two-part epoxy, Epon 862 with Curing

Agent W.

Pooling the epoxy directly onto the grating before immediately prior to the cure cycle did

not provide desirable results. Therefore, the next modification was to cure a layer of epoxy onto

the silicone rubber grating, and then at the transformation point in the cure cycle that grating

would adhere to the specimen under the same principle that was used in the original CRM.

Because of the very low viscosity of the Epon 862, continuous gratings could not be produced

because of the repellant nature of the silicone rubber (GE 615RV). Therefore, the epoxy was

changed to the 3501-6 resin. The procedure used for this process is described below and is

shown graphically in Figure 3-15:

1. A silicone rubber grating was replicated at room temperature onto a piece of Astrosital
glass (5.0" x 5.0" x 0.5") using a two part mixture of GE 6428. That silicone rubber had
a short working life which allowed for quicker replication. The silicone rubber was
allowed to cure for one hour before the gratings were separated

2. A Teflon device was prepared using the same process as in Step 6 in the previous section.
After the Teflon was tightened on the device, a small amount of 3501-6 was heated for
ten minutes before a vacuum was also applied for additional fifteen minutes. After the
resin had been sufficiently vacuumed, it was the pooled onto the silicone rubber grating
and the Teflon device was carefully lowered down so as to not trap any air bubbles within
the epoxy.









3. After allowing the epoxy to cure for ten hours in a 175C oven, the Teflon was carefully
removed by peeling it away from the grating surface at a sharp angle. The exposed
epoxy grating would later adhere to the prepreg during the cure cycle.

Results from Silicone rubber and 3501-6

After producing several gratings by this method, the diffraction grating was found to crack

after it cooled to room temperature, similar to the problem that was encountered when the 3501-6

resin was used in trying to reproduce the original CRM method. However, the cracking was

much less severe, so the gratings were still capable of being used from this method. A test

specimen produced using this CRM method is shown in Figure 3-15 and the resulting moire

images are shown for the U and V fields in Figure 3-16 (a) and (b) respectively. An example of

the cracks occurring in the diffraction grating can also be seen in the images of Figure 3-16 as

there is a crack visible along the bottom and left side of both images. The existence of that crack

leads to discontinuous fringes along that area. The point defects that can also be seen are due to

air bubbles that could not be completely removed from the epoxy before it was cured. The

presence of those defects introduces significant strain concentrations that do not correctly

describe the actual strain state of the specimen. A slight repeating pattern was evident in the

fringes shown in the moire images of Figure 3-16; however, it was not as prominent as was

expected. Therefore, because of the inconsistent diffraction gratings that were being replicated

and the low quality fringes, new methods needed to be attempted.

Kapton Polyimide Film Based Methods

Because many of the problems encountered when attempting the silicone rubber based

methods were due to the inability of replicating a perfectly flat grating onto the specimen, it was

thought that if a stiffer medium was used to replicate the grating, those issues could be resolved.

If a material with a glass transition temperature (Tg) lower than the curing temperature of the

composite was used, it would adhere to the specimen at cure. Then by assuming perfect









adhesion to the specimen, the deformations occurring in the composite would transfer to the

medium as it cooled down. This is similar in principle to the original CRM; however, instead of

transferring the displacement information through an epoxy layer, it was transferred through a

polyimide medium. When selecting the medium to use for this technique, the strength needed to

be lower than that of the resin in the composite so the deformations could be transferred and the

Tg needed to be lower than the cure temperature. The production time was very minimal using

this procedure and required approximately thirty minutes prior to manufacturing a specimen.

The process is described below and the stacking sequence is shown in Figure 3-17.

1. A master silicone rubber grating was replicated from a master Photoresist diffraction
grating onto a piece of Astrosital ultra low expansion glass (5.0" x 5.0" x 0.5") using a
two part silicone rubber (GE 615RV).

2. A thin layer of Epon 862 with curing agent W was pooled directly onto the silicone
rubber grating at room temperature. The high temperature epoxy was mixed using a
100:26.4 mass ratio of epoxy to curing agent. It was then heated for five minutes to
reduce viscosity and then centrifuged for five minutes. The purpose of this layer was to
replicate the grating from the autoclave tool onto the Kapton film.

3. A small piece of the Kapton film (1.0" x 1.0") was carefully lowered onto the epoxy
pool so as to not entrap air bubbles within the epoxy.

4. An additional layer of the high temperature epoxy (Epon 862 with curing agent W) was
pooled directly onto the Kapton film. This layer was carefully spread into a thin layer
so as to not introduce any air bubbles.

5. The prepreg was then lowered onto the layer of uncured epoxy before being cured in the
autoclave.

This method was originally completed without performing Step 4. However, due to

insufficient resin in the prepreg, the additional layer of epoxy was added in hopes that better

adhesion would occur between the specimen and the Kapton film.

Results using the Kapton film method

An example specimen using this method is shown in Figure 3-18 and the fringe images are

shown in Figure 3-19 for both the U and V fields. As can be seen from the moire images,









discontinuous fringes are developing at very regular locations throughout the field of view.

These locations correlate with the areas in the weave that are closest to the surface as illustrated

in Figure 3-20. The discontinuities were occurring because the epoxy that was pooled in Step 4

wicked away from fiber bundle peaks and concentrated in the resin pockets. That produced areas

in the film that could not fully adhere to the specimen which ultimately produced small air

pockets between the composite and the film. Therefore, because the displacement measurements

in those areas were only recording the deformations caused from the air pockets, extremely high

density fringes developed resulting in discontinuities within the overall fringe patterns.

Modifications attempted to the Kapton method

Different combinations of epoxy and polyimide films were used to modify this method in

attempts of achieving a replicable grating. The epoxies used were Epon 862 with curing agent

W and Tra-Con F230. After attempting various materials, the resulting specimens produced

gratings very similar to those shown in Figure 3-21. The discontinuities could not be avoided

after attempting several combinations of epoxies and polyimide films. Although highly

repeating fringes were developing, the images could not be correctly analyzed due to the extreme

strain concentrations at so many locations.

High Temperature Epoxy Based Methods

The high temperature epoxy based methods tried to correct the main problems that arose

from the previous attempts. It was decided that the features that needed to be included in the

grating production were: easy separation of the specimen from the autoclave tool, the master

grating had to be hard and epoxy needed to be able to cover and fill the weave pattern. In some

of the other methods attempted, separation of the specimen from the autoclave tool was difficult.

This resulted in incomplete grating transfer and it introduced bending into the specimen during

separation which could give inaccurate results in the fringe maps. It was decided that the best









separations had occurred when two layers of aluminum were applied onto the base grating with a

layer of Photo-flo between them. Through the combination of aluminum layers and a release

film between the prepreg and the autoclave tool, separation should become significantly easier.

After using silicone rubber as the base grating material for the autoclave tool, it was apparent

that it was too soft to yield a successful grating transfer. Therefore, it was preferable that the

base grating was to be made from a hard high temperature epoxy. The initial four steps were the

same for all of the attempts made using this method. Those steps are described below and shown

schematically in Figure 3-22:

1. A silicone rubber grating was replicated at room temperature onto a piece of ultra low
expansion glass (3.0" x 4.0" x 0.5") from a Photoresist master diffraction grating. The
silicone rubber was a two-part mixture, GE 615RV.

2. The master silicone grating was used to replicate the diffraction grating onto another
piece of Astrosital (3.0" x 4.0" x 0.5) using a high temperature epoxy (Epon 862 with
curing agent W). The parts were mixed together using a mass ratio of 100:26.4 and then
heated for five minutes at 1300C. Afterwards it was centrifuged for five minutes to
remove any bubbles before being applied to the grating for replication. The epoxy was
then allowed to cure for ten hours in a 1300C oven before the gratings could be separated.

3. Two aluminum layers were deposited onto the grating surface using vacuum deposition
with a film of Photo-flo used as a parting agent between the two layers. The Photo-flo
was allowed to sit for twenty-four hours before the second layer of aluminum was
applied.

Several modifications were made for the remainder of this technique which included

applying additional epoxy between the prepreg and grating so that the voids and surfaces were

sufficiently filled. In each method that was attempted, both the Tra-Con F230 and the Epon 862

with curing agent W were used to test the procedure because the two epoxies have very different

viscosities. The four methods that were tried involved pooling the epoxy onto the grating,

spreading the epoxy onto the grating before applying the prepreg, spreading the epoxy onto the

prepreg before immediately applying to grating and spreading the epoxy onto the prepreg and









allowing the epoxy to begin to set before applying the prepreg to the grating. Those processes

and reasons for doing so are described in the following sections.

Pooling the epoxy onto the grating

The philosophy behind this attempt was that additional epoxy needed to be applied

between the grating and the prepreg, and the simplest method to achieve this was to pool the

epoxy directly onto the grating. The procedure is described below:

1. A thin layer of epoxy was pooled directly onto the aluminized grating. Epon 862 with
Curing Agent W, Tra-Con F230 and 3501-6 resin were all used during different attempts
at this procedure.

2. The prepreg was then carefully lowered onto the pool of epoxy so as to not introduce any
air bubbles.

After producing several specimens through this technique, several flaws were discovered.

By placing the prepreg onto a single pool of epoxy, a uniform epoxy distribution was difficult to

achieve. With the majority of the epoxy localizing in one location, warpage was introduced into

the specimen at the pool location. All three of the epoxies tried in this approach resulted in some

localized warpage and pooling. Because of that, any strain measurements that could have been

taken would not have been true cooling induced deformations. Incorrect information would have

been recorded because the resulting strains would be strongly influenced by the epoxy instead of

the fiber weave.

Spreading the epoxy onto the grating

It was thought that the coverage of the epoxy onto the prepreg could be improved if the

epoxy was initially spread into a thin layer on the aluminized surface prior to laying down the

prepreg. This would also prevent areas with significant concentrations of epoxy which would

introduce warpage into the specimen as was obtained by the previous method. The procedure

used was the same presented in the previous section; however, before lowering the prepreg onto









the grating, the epoxy was carefully spread into a thin layer so that air bubbles were not

introduced.

This method did produce slightly better results; however, the gratings on the specimens

were not fully continuous. There were many instances that the resin pockets did not have

sufficient resin, therefore producing a discontinuity in the grating surface. Another problem was

that there were several locations where there was insufficient epoxy where the fiber bundles were

closest to the surface which resulted in discontinuous fringes to develop. Some examples of

specimens made using this method are shown in Figure 3-23.

Spreading the epoxy onto the prepreg before immediately applying to grating

Many issues seemed to be caused from having insufficient epoxy in the resin pockets and a

lack of complete coverage over the fiber bundles in the previous two techniques. Therefore, to

attempt to improve the coverage by the epoxy onto the specimen, the epoxy was spread into a

thin layer on the prepreg before it was placed down onto the grating. The prepreg was prepared

for curing as described in the following steps:

1. A small amount of epoxy was pooled onto the center of the prepreg, as shown in Figure
3-24 (a).

2. Using a flat edged spatula, that pool of epoxy was spread into a thin layer so that it
covered the area of interest on the specimen (Figure 3-24 (b)). This was done carefully
so that air bubbles were not inadvertently introduced which would cause discontinuities
in the fringes.

3. The specimen was then lowered onto the aluminized autoclave tool so that the epoxy
region matched up with the grating area on the glass.

The grating surfaces did show improvement using this method. However, there were still

regions yielding discontinuous fringes caused either by flawed areas at the resin pockets or at the

surface contact zones. It was thought that this was still occurring because although the epoxy









was spread onto the composite filling those voids, when the prepreg was pressed onto the grating

the epoxy wicked away from those areas, still leaving voids.

Spreading the epoxy and letting it set before placing on autoclave tool

By examining the gratings on the specimens, it appeared that the epoxy layer intended to

fill the resin pockets and cover the fiber bundles was not remaining there after the prepreg was

placed onto the autoclave tool. The next attempt allowed that layer of epoxy to begin to cure so

it would not easily wick away from the critical locations. The process followed was identical to

that in the previous section except that after the layer of epoxy was spread onto the prepreg it

was allowed to sit and partially cure before it was placed onto the autoclave tool. Examples of

specimens made using this procedure are shown in Figure 3-25. As can be seen from the figure,

the transferred gratings were repeatedly smooth and continuous. Also, all of the specimens

produced from this technique were easy to separate from the autoclave tool. The fringe images,

which will be included in Chapter 4, also yielded areas within the region of interest of

continuous fringes with minimal defects which was ultimately the main goal so that continuous

strain contour maps could be obtained.

Conclusions

The original CRM process that was developed was a time intensive method that was found

to be very difficult to repeat. In addition, because of the complex geometry of the plain weave

composite, the original method was not able to repeatedly separate from the autoclave tool to

produce a continuous grating. Therefore, the focus of the research had to be shifted to modifying

that technique to be applied to woven composites.

Several methods were attempted using different philosophies to transfer a diffraction

grating from the autoclave tool to the specimen at the point of the liquid to solid transformation









point of the resin. However, most methods yielded undesirable results in either the separation

abilities or the quality of the final grating and resulting fringe images.

The last method attempted, allowing the spread epoxy to begin to cure on the prepreg

before being placed onto the autoclave tool, provided the most repeatable and highest quality

results. Therefore, for the experimental testing, that method will be used to measure the process

induced residual strains developing in the specimen. Those results are examined in Chapter 4.












the Reference Frequenc.
Figure 3-1. Schematic of a typical four-beam interferometer
..L


Figure 3-2. Fixture used to align the diffraction grating with the autoclave tool.


























Figure 3-3. Oven and vacuum lines used for the curing cycle




p3orus rfeeaa? Min Bratherf ply


Pre-preg


Figure 3-4. Vacuum bagging assembly


















S150

100

50


0 1 2 3 4 5 6 7
Time (hrs)
Figure 3-5. Cure cycle used for HMF plain weave prepreg


Separate Specimen
from tool


Figure 3-6. Tool used to separate master grating from the autoclave tool





















Figure 3-7. Tools required to make Teflon device


(a) (b)


Figure 3-8. Specimens made using Original CRM procedure. Cracking along resin lines is
shown in (a) and incomplete separation is shown in (b).















(a) (b)
Figure 3-9. Specimen made using original CRM with successful transfer. Regions void of resin
seen in (a) and the resulting discontinuous fringes for the U-field can be seen in (b)


Light Source


-f II


rI ,\X


Light Source
(bl
Figure 3-2. The shapes of the diffraction gratings when the master grating is hard from a high
temperature epoxy (a) and when a silicone rubber grating is used as the master grating
(b).











(a)


Figure 3-3. Fiber reactions with different silicone rubber thicknesses. Figures (a) and (c) show
the initial conditions for different silicone rubber thicknesses. During curing the
fibers compress into the silicone rubber as seen in (b) and (d)


Figure 3-4. Silicone rubber grating showing exposed fibers indicating that the grating surface
would not be perfectly flat


C~XXi:I:
(br









~-r---'jy-'~prr nrr'PrF i-IT.


Figure 3-5. Specimen made using thin silicone rubber grating


Silicone rubber grating



Teflon device



Prepreg

Cured 3501-6 resin


Figure 3-6. Grating production procedure for silicone rubber with cured 3501-6 resin


Step 1




Step 2




Step 3
























Figure 3-7. Specimen made using the silicone rubber with cured 3501-6 resin














(a) (b)

Figure 3-8. Moire images taken from the specimen shown in Figure 3-16. The U and V fields
are shown in (a) and (b) respectively.









Pool of Epoxy
Kapton Film I PrePreg Release Film


Autoclave Tool Silicone Rubber Grating
Figure 3-9. Stacking method used for the Kapton polyimide process


Figure 3-10. Specimen made using the Kapton polyimide method


(a) (b)
Figure 3-19. Moire images taken for the specimen shown in Figure 3-19. The U and V fields
are shown in (a) and (b) respectively. Fringe discontinuities can be seen at several
locations in both images.




















(a) (b)
Figure 3-11. Specimen showing the locations of strain discontinuities occurring on peaks of
fiber bundles (a) and the correlated location within the RVE (b).














(a) (b)











(c) (d)

Figure 3-12. Various specimens manufactured with different polyimide materials









-- Silicone Rubber Grating


Silicone Rubber Grating


Autoclave Tool



Aluminized Autoclave Tool


Figure 3-13. Procedure used to produce base grating for high temperature epoxy methods












Figure 3-14. Specimens produced by spreading a thin epoxy layer directly onto the aluminized
grating surface. Several locations on both gratings can be noticed that include many
discontinuities in the grating surface.


Step 1


Step 2





Step 3




















(a) (b)


Figure 3-15. Pooling the epoxy onto the prepreg as done in the high temperature epoxy method
(a) and spreading the epoxy into a thin layer that fills the resin pockets and covers the
fiber bundles (b).


Figure 3-16. Sample specimens made after allowing the epoxy to begin to set on the prepreg.









CHAPTER 4
RESIDUAL STRAIN MEASUREMENTS

Phase Shifting

Moire fringe patterns develop by interference caused when two light sources reflecting off

a common location. Within that interference there is a corresponding intensity and phase

associated at each location within the fringe image as can be seen from Equation 1 where s is the

2 = ae +2 2 2
S =+aaleil +a2eiP22
= a, + aj + 2ala2cos(P
Equation 1

intensity and (p is the phase. Therefore, the fringe images can be analyzed using either the

intensity or the phase information.

Traditionally, the intensity based techniques have been used. To calculate the strain at a

desired location, the number of fringes in that area, Ni, and the exact locations between those

fringes, Axi must be measured. Then using Equation 2 4 the strains can be calculated.

1 ANx
Exx(X,y) = Ax
f AN
Equation 2
1 ANy
yy (X, )= Ay
f Ay
Equation 3

= (x, + y)AN
Equation 4


However, this method is not always desired as it is rather time intensive and errors in the results

are inevitable because of the uncertainty involved in choosing the exact location for a fringe.

The second method uses the phase information from Equation 1 to calculate the strain

information. However, because that expression uses a cosine relationship, it is impossible to

determine whether the phase is positive or negative from a single fringe image because both have









the same intensity value. This problem is solved by introducing a known carrier fringe pattern,

or phase ramp a, into the equation as is seen in Equation 5.

Ix(, y) = Iback(X, ) + Imod(X, y)COS [0(x,y) + na]
Equation 5

By capturing several fringe images that are shifted by that known phase ramp; the complete

phase information can be obtained for the entire region of interest. The strains can then be

calculated using Equations 6-8.

1 A ,(x, y)
f 2nAx
Equation 6
1 A ,(x, y)
f 2nAx
Equation 7

Y (x, y) 1 A + APXJY)
Yxy f [ 2tnAy 2nAx ]
Equation 8


For this research an automated analysis tool was designed for Matlab by Weiqi Yin (a

student in the experimental stress analysis lab at the University of Florida) [30]. By inputting

four sequentially shifted fringe images into the program, full-field displacement and strain

information could be obtained. This was critical for the research since full-field information was

desired so the strain trends throughout the RVE could be determined. Had an intensity based

method, or another non-automated technique been used, the ability to gather full-field

information would have been drastically reduced.

Results

The fringe images are shown in Figure 4-1 and Figure 4-4, and the phase shifted residual

strain results for two specimens are shown in Figures 4-2, 4-3, 4-5 and 4-6. The maximum

repeating tensile strains that were developing were approximately 2500pt, and the maximum









repeating compressive strains were approximately -3000te in both the U and V fields.

Additional specimens and their resulting strain contour maps are included in Appendix A.

The strain pattern that developed can be explained by relating the trends to the geometry of

the weave. As can be seen in Figure 4-7, there are two main regions within the RVE, the resin

and the fiber regions. The resin zone expands the entire length of the RVE between fiber

bundles. However, along that length, the resin depth from the surface changes as it crosses over

transverse bundles. The fiber regions have much lower surface resin content levels since they

are much closer to the surface.

The regions of high tensile strains seen in the figures were occurring along the tops of the

fiber bundles in the fiber regions in both fields of view, whereas the compressive strains were

developing throughout the resin zones. The CTE value for the resin is significantly larger than

that of the fiber, so as the composite cools to room temperature, a considerable amount of

residual strains develop within the resin because the fiber bundles restrict free contraction in

those areas. The tensile regions that developed along the fiber bundles were a direct result from

the compressive resin regions. To maintain static equilibrium, the thin resin zones that cover the

fibers must undergo tension to accommodate for the large compressive areas.

It can also be noted that the strain patterns and values are very similar for the U and V

fields. That was expected to occur because of the symmetry of the weave geometry.

Conclusion

The improved CRM was used to measure the process induced residual strains for the plain

weave composite specimen. By using an automated phase shifting analysis method, the strains

throughout the entire region of interest were calculated. Having that information allowed the

strain trends throughout the RVE to be presented. Significant strains were found to develop

within the RVE, both in compression and in tension. The displacement information that was









obtained from that data will be used in Chapter 5 to numerically determine the residual stresses

that were occurring via finite element method.

After examining the strain maps for several specimens, including those produced by other

CRM processes in Appendix A, it can be seen that fairly repeatable strain values were obtained

in the extreme strain regions. Although all of the methods tested throughout this research were

not able to produce strain contour maps with the definition that was obtained using the final

method, the overall strain values obtained were fairly constant between all methods. Therefore,

the results that were presented in this chapter do appear repeatable. Analytical validation will be

shown in Chapter 5 which discusses the simulated cooling cycle that was completed within

ABAQUS.















(a) (b)
Figure 4-1. Fringe images for Specimen 1 for the U-field(a) and V-field (b).

Stra )4000
Ou3
SM)I -
515
SIX1 ( | |



: '-4 I
a' 1a m _I ,


'(-- ,
1 i no00ph h n
100 200 300 400 500 600 70 Boo 900 1000
Figure 4-17. P1 strain contour map via phase shifting from Specimen 1
















IOUI






2MM















100 200 300 400 500 600 700 00 9000 IOw


Figure 4-3. S2 strain contour map via phase shifting from Specimen 1


















(a) (b)


Figure 4-4. Fringe images for Specimen 2 for the U-field (a) and V-field (b).











































w 100 150 200 250 300 350 400 460 500 550


- -I 4*00n


-6000


Figure 4-5. P1 strain contour map via phase shifting from Specimen 2


A.

a


400



3000



2000


1 J00
2f1f


Figure 4-6. S2 strain contour map via phase shifting from Specimen 2


50 100 150 200 250 300 350 400 450 500 550















Resin Zone

Fiber Zone






Figure 4-7. The representative volume element of the plain weave geometry.









CHAPTER 5
FEM SIMULATION OF PROCESS INDUCED STRAINS

Motivation for Performing FEM Analysis

Because of the complexity of the woven geometry, finite element modeling was the

preferred analysis method due to the full-field capabilities it provides and because most

analytical models are highly computation intensive and require many simplifications. There

were two ultimate goals for performing the finite element analysis. Those were to compare the

results obtained experimentally by simulating the curing cycle and to determine the experimental

residual stresses using the displacement data. Although experimental validation could be

performed using a unidirectional laminate and comparing against previous tests, the same could

not be done for the woven composite. Therefore, the validation was important for this research

to ensure that the data collected was reasonable, taking into account both the mechanical and

thermal properties of the material.

Model Description

The model used for the analysis was one developed by Karkkainen of the representative

volume element for the plain weave composite [31]. The geometry of the model was taken from

a literature source that had documented the dimensions required when constructing the weave

pattern and had taken measurements of a typical plain weave fabric using SEM and microscopes

[32].

Two materials were created within the model, that of the matrix and that of the weave.

The weave was assigned the properties of a unidirectional laminate, AS4/3501-6, taken from

experimental data [33]. However, because of the undulation that occurs in the weave, the

properties could not be assigned directly to the entire fiber bundle. Instead, when building the

model Karkkainen assigned a local material coordinate system to each element so that when the









laminate properties were applied they would correctly align with the local direction of the fibers.

By doing this, only one material needed to be defined for the weave, and it could be applied for

the warp and fill yarns. The temperature dependent properties for the matrix, 3501-6, were taken

from the literature [34]. All of the material properties used in the model are shown in Table 5-1.

Simulating the Curing Cycle

The application of temperature and pressure fields was required to model the cooling

process to simulate both the change in temperature and the releasing of the vacuum surrounding

the specimen. When the temperature dependent CTE data was input into the model, the

reference temperature was chosen to be 1300C so that any temperature field applied would

simulate cooling from the cure temperature. The room temperature conditions were modeled by

applying a full-field temperature of 220C and a hydrostatic pressure load of 101.3kPa. To

prevent rigid body motion, while allowing full deformations of the model, one node was fixed in

all three translational directions.

The results are shown in Figure 5-1 and Figure 5-2. Plots were also created along the fiber

bundles to illustrate the change in sign that occurred between the tension and compressive

regions. As can be seen in Region A of Figure 5-1, large compressive strains are developing

within the resin rich zones of the RVE. Conversely, the strains are highly tensile on top of the

fiber bundles where less resin is present. Those tensile strains are being induced from the

surrounding compressive resin rich zones. As the resin rich zones contract, the thinner resin

zones are forced into tension to satisfy equilibrium. Because of this phenomenon, undulating

strain patterns develop as a function of the amount of resin in that region. Figure 5-3 shows this

occurring, as F1 varies between tensile and compressive values along the length of the fiber

bundle in the 1-direction. Those trends were expected to occur because of the significantly









larger CTE value of the resin as opposed to the fiber bundles and the strain values show good

agreement to the experimental data.

Using FEA to Compare to the Experimental Method

Because it was not feasible to use experimental verification for the developed CRM

procedure, it was essential to simulate the cooling process associated with the cure cycle to

model the residual strains that developed. By performing a comparison between the two data

sources, the validity of the experimental results could be obtained. It would not be expected for

the results to match perfectly as the finite model did not account for any chemical shrinkage that

would be occurring in the actual specimen. Also, the geometry of the model and the actual

weave geometry used for this research were not exactly the same. The geometry for the model

was taken from a literature source when the model was developed. However, the actual

geometries of the materials tested were close to those used in the model, so the results provide

good approximations for the true values.

Figure 5-4 shows the e1 results for the RVE calculated in ABAQUS with the schematic of

the unit cell overlaid to show how the strain values relate to the weave geometry. For

comparison, the contour plot of the experimental data obtained for S1 using the CRM is shown in

Figure 5-5. By comparing the two plots for the U field data, it can be seen that good agreement

exists between the two methods. The overall trends throughout the RVE are the same. The

values differ slightly in some regions. One case in particular, is that along the fiber bundles of

the experimental data, extremely high tensile strains develop, approximately 2500[,e verse the

maximum strain developing in the finite element model was approximately 500es. One reason

that could cause such a significant difference would be the amount of resin in the specimen as

opposed to that in the model. Through the different procedures that had been tested it became

apparent that the resin level was low in the prepreg. Having low resin content over the fibers









would yield higher strains due to the lack of material to support the surrounding compressive

strains. Therefore, it does seem reasonable that such a large discrepancy would occur in the

values for the maximum tensile strains.

The undulating compressive strain values along the resin zones in the 3-direction shown

good agreement between the two plots. Both plots show that the strains in those areas are

remaining between approximately -3500[,e and -1000le[. The actual repeating pattern can be

better seen in Figure 5-6.

Similar plots were created for 83 and those are shown in Figure 5-7 and in Figure 5-8 for

the FEA results and experimental data respectively. Like the plots for e1, good agreement occurs

between both sets of data. Again, there is a significant difference between the maximum tensile

strains occurring in the two plots. However, the reason for this occurring would be the same as

the case for the s1 data.

Experimental Process Induced Residual Stresses

The displacement data obtained from the experimental results was used in the model to

determine the residual stress contour fields. When temperature fields are applied to a material,

strains develop due to the coefficient of thermal expansion as the change in temperature

increases. However, those deformations are not due to an applied force or stress, so they have to

be modeled as virtual loads.

The phase shifting analysis provides strain and displacement information at every pixel

location within the area of interest. Theoretically then, one could use that full-field data and

input the full-field displacement data into a finite element model to obtain the resulting full-field

stress data. However, due to the large number of data points from the experiment and the

number of nodes in the model, it was decided to input the displacements through the periodic

boundary conditions (PBC).









When Karkkainen developed the ABAQUS model, a dummy node was created to control

the PBC's. Therefore, in order to take advantage of the PBC's, the overall displacements for the

U and V fields needed to be determined from the experimental data. Because the displacements

constantly vary throughout the RVE, the average overall displacements were used for each

direction. Using this method, the displacements that were used for the model in the U and V

directions were -6.13x10-3 mm and -6.06x10-3 mm respectively. Those values were achieved by

taking the average displacement occurring in each direction along the length of the RVE.

The stress contours are shown in Figure 5-9 (a) and (b) for Io and 03 respectively. Also,

the overall trends for the stresses are plotted in Figure 5-10 and Figure 5-11. The stresses follow

the same trends that have been shown for the strains. Peak tension values are occurring over the

fiber bundles and the peak compression values are within the resin rich pockets. The peak

tensile stress was approximately 5 MPa and that was located in the very center of the RVE over

the fiber bundle where the resin content is lowest. The peak compressive stress was

approximately -24 MPa and was located at alternating resin zones throughout the RVE.

Conclusions

Finite element analysis was used for two aspects within this research. The cooling cycle

was simulated to obtain a prediction for the strain pattern and values that should be expected

from the experimental work. The strain pattern was described by relating it to the geometry of

the RVE and the different CTE values of the two different materials. The strains fell within the

approximate range of a maximum tensile strain of 1820tE and a maximum compressive strain of

-3700[t.

By comparing the contour plots obtained through the experimental and analytical methods,

reasonable verification of the method developed was completed. Because other experimental

techniques do not exist that are capable of measuring the full field strains of woven composites,









data was not available to perform experimental validation. Therefore, performing the simulation

within ABAQUS was the best method available to verify the experimental results. Even though

the results cannot be exact between the two models due to geometry and chemical shrinkage

issues, very good agreement occurred. Therefore, the results that were obtained through the

improved CRM procedure were those expected.

After using the deformations obtained from the experimental data, the residual stresses

along the surface for the RVE were calculated. Although the values are relatively low, those are

the stresses within the resin on the surface. According to the data sheet provided by the prepreg

manufacturer, the tensile strength of the resin is 82.7 MPa. Although the actual stresses that

developed were well below that value, they are still important to understand since they do limit

the usable strength remaining in the structure.












Table 5-1. Material properties used in ABAQUS model for the plain weave composite RVE


E1 (Gpa)


E2 (Gpa)


G12 (Gpa)


a, (e/C0)


Weave (AS4/3501-6) at 1200C 138.00 5.64 2.48 0.30 0.59 26.90
at 20C 138.00 9.35 5.30 0.30 0.45 23.03
Matrix (3501-6) at 1200C 3.50 3.50 1.30 0.35 49.00 49.00
at 200C 3.50 3.50 1.30 0.35 41.30 41.30


E, Ell
(". crt.:7s
41 F. rJ21e-O3
*IL 82-1:-03
e -e C

1.300 D0
7 33 33. 3, ql ,


3I1:C. TZZ.I i *P f Eon II:e 10
2OE C"'a'l 3 AEi:L' mi~.i..l 6.4 1 The AUQ 2A 1u:1e:S, Ea

p 2. Enforce room r-p
n 1 Step Tim: 1.
e Dir: (i ef.a~ on .51e fact: 1 7-98.5-03


Figure 5-1. Simulated process induced residual strains in the 1-direction (s1) resulting from
cooling the RVE from Tref to room temperature and introducing it to atmospheric
pressure. Region A is the resin rich zone and Region B has a lower resin content.


a2 (LE/C)











E, E33
CAv- C -. i


_-Z.- I
_z .X6 oz_
2. S_ ;L
iL-:..-! s i- i


"'l_-130/2 -o,6 ;1.- q I-2 ;I. eV \' 2ode
2 ODB: CoclO.oB A&CV s.r.j. S. Q41. T. Aug 23 1.:1:.SE 1

9^rm gEcr ?. Ernfore irnnni rewi
1; Step Tin 1000
\mr.E ED33
Figu e5-2.e *her cnt Defuraaon Steale raFtei: +-7 9B5e-03


Figure 5-2. The contour plot of F1 after the simulated cooling cycle to room temperature.


[x10j
0.40

0.00

-0.40

-0.80

-1.20

-1.60

-2.00

-2.40

-2.80


-3.20 L
0.00


[x10-1
0.50

0.00

-0.50

-1.00


S I I I I V 1-1 -350 1 I I I I I I I
0.40 0.80 120 1.60 0.00 0.40 0.80 1.20 1,60
True distance along path 'Mid1' True distance along path TopEdge'


Figure 5-3. Plot of 1i along the length of the midline of the warp of the RVE in the 1-direction.
The data is taken on the surface of the matrix along the line shown in the subset
figures.












L








F


Figure 5-4. FEA S1 results for the RVE from the simulated curing cycle obtained

4000
3000
2000
1000
I
I I
-1000


-3000
-4000


Figure 5-5. Experimental phase shifted Pe resulting from the cooling cycle


Q

o,,, 7,5,

... .... V4
"1"'0 A'11 4,
Wk*TOAVAWYAY
A
jr 4"
11" 0~4%, 4 'A' I
I 'AUT % %
i4. '4 -.1 .4 I
X1.
".650",
f I








'OeAV-


J


1


I I


uI


I I


I I


r7I


I ---------------- I-


---------------------------------------- ---------------------------- ------------------------------


E, El1
(Ave. Crit.: 75;)
+1.821e-03
+1. 00e-03
+5. 00e-04
-2.328e-10
-5.00Oe-04
-1. O0e-03
-1. S00e-03
-2.000e-03
-2.500e-03
-3. OOOe-03
-3 500e-03
-4.000e-03
-4.500e-03
-5. 00e-03
-1.163e-02


n











[x10o
-1.00



-1.50


-2.00


-2.50


-3.00 i I I-
0.00 0.40 0.80 1.20 1.60

Plot of the residual strain, e1, along the resin zone in the 3-direction.


Figure 5-6.










I I


L











r


I I


ImI


I-I


I I


ILIJ


L.


Figure 5-7. FEA 83 results for the RVE from the simulated curing cycle


4000

3000

20ff

1000






-2000

-3000


Figure 5-8. Experimental phase shifted 83 resulting from the cooling cycle


-W 19
-11, 'A tyTA -U AW
A' JV 1000AW.WA
AU,



. ....... X
IW.W, 'W!.
"XI A
TAY P,
A 11 N
%IA
ORA All
""A6, %
4 4%1'
't X.
UT 7 T:T
f, ow. A '.-I w
U, 40 IAV&VATAT&TAvt
1
"V


I I


E, E33
(Ave. Crit.: 75<)
+4.178e-03
+1.000e-03
+5. 00e-04
-2.328e-10
-5.000e-04
-1.00 e-03
-1.500e-03
-2. 00e-03
-2. 500e-03
-3.OOOe-03
-3.500e-03
-4. 00e-03
-4.500e-03
-5. 00e-03
-2.489e-02















S, Sl S, 533
( e. Cr t. 75-, IAe, Crit.: 754)






























'..--'. 0.ocorM.V -VI ':.i'."j.-?
2 .P Mo.0 6.4-1 Sun Cct ZS 344!52 2 ODB ExpComp.oBt ABAQUS/Scsndard 6.4-1 Sun 0m 14 22l25!3Z

Ier, t1 SteP: Step-. .
3 ii.1ite, lrsep Tite 1,000 en Sep Time 1.000
hi..j Sl P y Vat: 3, i33
:. r U Defontmion Scale Factor: +9.738e-03 Deformed Var 0 Deformation Scale Factor: +9.888e-03


Figure 5-9. Stress contours for the 1- and 3-directions, (a) and (b) respectively, obtained using

experimental displacement data































0.40 0.80 1.20 1.60


-25-00 ,I IW
0.00 0.40 0.80 1.20 1.60
(b)True distance along path 'Bottom'


*10.00 [ I
\1 i I I I ,
0.00 0.40 0.80 1.20
(c) True distance along path 'Mid3'
[x10']
5.oo00 K- I I I '


(c) (d)


S-5.00 -







-15.00 I
0.00 0.40 0.80 120
(d) True distance along path 'Right'


Figure 5-10. The stresses are plotted along the fiber directions in along two different bundles the
directions of the plots are shown in the subset figure.


0.00

-4.00

-8.00
C)
* -12.00
CO
-16.00

-20.00


-24.00 -
0.00
































0.00 040 0.80 1.20 1.80
True distance along path 'ResinBottom'
[xlOj]
-4.40


4.80 -


-5.20 -


-&60


[x10]


-14.00


-16.00


S-1800-


-20.00


-22.00


-24.00
0.00 0.40 0.80 1.20 1.60
True distance along path 'ResinRight'


-500

c -550
Cn
-6.00

-6.50

-7.00


0.00 0.40 0.80 1.20 1.60
True distance along path 'ResinBottom'


0.00 0.40 0.80 120 1.60
True distance along path 'ResinRight


Figure 5-11. The stresses along the resin zones are shown along and transverse to the resin zone.
Figures (a) and (c) are plotted along line (a) in the subset, and figures (b) and (d) are
plotted along line (b) in the subset.









CHAPTER 6
CONCLUSIONS AND FUTURE WORK

Conclusions

Residual strains are usually inevitable to occur in composite materials due to the high

temperature manufacturing process required and chemical shrinkage. Because of this, it is

critical that those strains, and ultimately the stresses, can be determined so that the usable life of

the structure can be assessed. Many methods exist to measure those strains in most materials,

and even in unidirectional composites. However, only a handful of methods have been used to

measure the strains developing in woven composites, and none of those have been able to obtain

a full-field contour map of the strains. Cure reference method was the solution to that problem,

because with the combination of moire interferometry, strain measurements can be obtained

throughout an entire field of view. Then by processing the resulting fringe images with phase

shifting, a complete two dimensional map of the residual strains can be obtained for the U and V

directions.

The original procedure for producing the autoclave tool for CRM was very time intensive

and provided poor repeatability when attempts were made to reproduce that method. Therefore,

this work became focused on modifying and improving that method. By creating a master

autoclave tool, and then pre-curing a thin layer of epoxy onto the prepreg before cure, very

repeatable gratings were obtained that easily separated from the autoclave tool.

The resulting data showed a wide range of strain values occurring in a very small region of

the weave geometry. Because of the complex geometry exhibited by woven composites, it was

expected to obtain both tensile and compressive strains throughout the field of interest. Those

did occur, and the strains throughout the RVE fell between -3000[t, and 25001ae. The maximum

tensile strains were found to occur over the fiber bundles and the high compressive regions were









the resin rich zones between the bundles. The experimental displacement data was then input

into the ABAQUS model to calculate the residual stresses that were developing because of the

cooling process. It was found that the stresses were bound between approximately -24 MPa and

+5 MPa. Although the strength of the resin used in the prepreg is 82.7 MPa that level of stress is

not actually available to the end user because of the residual stresses that developed from the

curing cycle, this is especially try for cryogenic uses. Even though the values of stresses were

small with respect to the strength, if not correctly accounted for, failure could occur before it

would be expected because those residual stresses were neglected.

The finite element model was also used to validate and verify the data obtained through the

improved CRM. After comparing the strain plots obtained from the two methods it was apparent

that the improved CRM method was providing accurate data for the residual strains developing

in the specimens.

Future Work

Although the final CRM grating production procedure was significantly improved over the

original process, both by time and ease of production, some improvements could still be made.

The final method definitely showed the best repeatability as far as achieving smooth grating

surfaces that were easily separable from the autoclave tool. However, there are still frequent

occurrences in which sufficient epoxy is not covering the fiber bundles closest to the surface.

Whenever that occurs, the fringes are highly discontinuous so phase shifting analysis cannot be

completed in that region. The method to apply the last epoxy layer should be slightly modified to

resolve that issue.

For this research, a field of view was chosen on the specimen, and the full-field strain

information was calculated throughout that entire region. As was seen from each of the contour

maps created, perfectly uniform strains were not measured throughout the entire viewing region









on the specimens. Therefore, although the RVE is a repeating unit, there do appear to be small

discrepancies between each of those elements. It would be desirable to obtain statistical

information throughout a region of interest to determine the variability between different RVE's.

The variability that could produce different strain values would be caused by non-uniform resin

distributions throughout the prepreg. Having regions either rich or low in resin would produce

this type of inconsistency in the results.

For the finite element analysis to calculate the experimental residual stresses, the

deformations were input as periodic boundary conditions accounting for the total average

displacements that occurred for the entire RVE. Although that was a reasonable approximation

to predict the stress levels occurring throughout the region, it was unable to capture the full

displacement non-uniformity that occurs throughout the region. Therefore, future work should

include developing a method to input the experimental data directly from the phase shifting

analysis into the ABAQUS model. Once that is accomplished, a more accurate stress field can

be obtained through the finite element analysis.







APPENDIX
CRM IMAGES AND PHASE SHIFTED RESULTS


Figure A-i. Silicone Rubber and 3501-6 Resin, U Field


50-

100-


I


150 -


200-

250-

300-

350-


Figure A-2. 1i


F''I


'V


A


9


I ... t Ii
50 100 150 200 250 300 350
phase shifted results for silicone rubber and 3501-6 resin


1000
500
0
-500
-1000
-1500
-2000
-2500
-3000























Figure A-3. Silicone rubber and 3501-6 resin V Field
Figure A-3. Silicone rubber and 3501-6 resin, V Field


S-6000

50 5000
50 -
4000
100- 3000


150- am -2000

S-1000
200 'a : ....
0

250 -1000


300 -2000
-3000
350I
50 100 150 200 250 300 350
Figure A-4. 82 phase shifted results for silicone rubber and 3501-6 resin






































Figure A-5. High temperature epoxy spread and set on prepreg, U Field

I I I I I I I I


1000


500





-500


-1000


-1500





-2500
-2500


-3000


-3500


5n inn in 1 ?nn 5ln nn an An fn sn nn 5n


Figure A-6. S1 phase shifted results for High temperature epoxy spread and set on prepreg





































Figure A-7. High temperature epoxy spread and set on prepreg, V Field


50- 5000


1001 4000


150


200- 2000


250
1000
300
0
350

400 -1000



450 -2000


500- -3000


550 O-4000
50 100 150 200 250 300 350 400 450 500 550


Figure A-8. 82 phase shifted results for High temperature epoxy spread and set on prepreg









LIST OF REFERENCES


1. Gibson, R.F., Principles of Composite Material Mechanics, McGraw-Hill, New York,
New York (1994).

2. Post, D., Han, B. and Ifju, P.G., High Sensitivity Moire: Experimental Analysis for
Mechanics and Materials, Springer-Verlag, New York (1994).

3. Hahn, H.T., and Pagano, N.J., "Curing Stresses in Composite Laminates," J. Composite
Materials, 9, 91 (1975).

4. Mathar, J., "Determination of Initial Stresses by Measuring the Deformation Around
Drilled Holes," Trans. ASME, 56, 249-254 (1934).

5. Rendler, N.J., and Vidness, J., "Hole-drilling Strain-gage Method of Measuring Residual
Stresses," J. Experimental Mechanics, 6, 577-586 (1966).

6. Schajer, G.S. and Yang, L., "Residual-stress Measurement in Orthotropic Materials
Using the Hole-drilling Method," J. Experimental Mechanics, 34, 324-333 (1994).

7. Sicot, O., Gong, X.L., Cherouat, A. and Lu, J., "Determination of Residual Stress in
Composite Laminates Using the Incremental Hole-drilling Method," J. Composite
Materials, 37, 831-844 (2003).

8. Wu, Z., Lu, J. and Han, B., "Study of Residual Stress Distribution by a Combined
Method of Moire Interferometry and Incremental Hole Drilling, Part I: Theory," J.
Applied Mechanics, 65, 837-843 (1998).

9. Diaz, F.V., Kaufmann, G.H and Galizzi, G.E., "Determination of Residual stresses Using
Hole Drilling and Digital Speckle Pattern Interferometry with Automated Data Analysis,"
Optics and Lasers in Engineering, 33, 39-48 (2000).

10. Zhang, J. and Chong, T.C., "Fiber Electronic Speckle Pattern Interferometry and its
Applications in residual Stress Measurements," Applied Optics, 37, 6707-6715 (1998).

11. Hahn, H.T., "Residual Stresses in Polymer Matrix Composite Laminates," Composite
Mat., 17, 265-277 (1976).

12. Kim, R.Y. and Hahn, H.T., "Effect of Curing Stresses on the First Ply-failure in
Composite Laminates," J. Composite Mat., 13, 2-16 (1979).

13. Predecki, P. and Barrett, C.S., "Stress Measurement in Graphite/Epoxy Composites by X-
Ray Diffraction from Fillers," J. Composite Materials, 13, 61-71 (1979).

14. Benedikt, B., Kumosa, M., Predecki, P.K., Kumosa, L., Castelli, M.G. and Sutter, J.K.,
"An Analysis of Residual Thermal Stresses in a Unidirectional Graphite/PMR-15









Composite Based on X-ray Diffraction Measurements," Comp. Science and Technology,
61, 1977-1944 (2001).,

15. Benedikt, B., Rupnowski, P., Kumosa, L., Sutter, J.K. and Predecki, P.K.,
"Determination of Interlaminar Residual Thermal Stresses in a Woven 8HS
Graphite/PMR-15 Composite Using X-Ray Diffraction Measurements," Mech. Of Adv.
Mat. And Struct, 9, 375-394 (2002).

16. Daniel, I.M., and Liber, T., "Lamination Residual Stresses in Fiber Composites," IITRI
Rep. D6073-1, for NASA-Lewis Research Center, CASA CR-134826 (1975).

17. Daniel, I.M., Liber, T. and Chamis, C.C., "Measurement of Residual Strains in
Boron/Epoxy and Glass/Epoxy Laminates," Composite Reliability, ASTM STP 580,
Amer. Soc. Test. And Mat., 340 (1975).

18. Daniel, I.M., and Liber, T., "Effect of Laminate Construction on Residual Stresses in
Graphite/Polymide Composites," J. Experimental Mechanics, 17, 21-25 (1977).

19. Zewi, I.G., Daniel, I.M. and Gotro, J.T., "Residual Stresses and Warpage in Woven-
Glass/Epoxy Laminates," J. Experimental Mechanics, 27, 44-50 (1987).

20. Lawrence, C.M., Nelson, D.V, Spingarn J.R. and Bennett, T.E., "Measurement of
Process-Induced Strains in Composite Materials Using Embedded Fiber Optic Sensors,"
Proc. Of SPIE, 2718, p 60-68 (1996).

21. Lawrence, C.M., Nelson D.V., Bennett, T.E. and Spingarn, J.R., "An Embedded Fiber
Optic Sensor Method for Determining Residual Stresses in Fiber-reinforced Composite
Materials," J. Intell. Mater. Syst. Struct. 9, 788-799 (1998).

22. Anastasi, R.F., and Lopatin, C., "Application of a Fiber Optic Distributed Strain Sensor
System to Woven E-Glass Composite," TM-2001-211051, NASA, 2001.

23. Manson, J.E., and Seferis, J.C., "Process Simulated Laminate (PSL) Characterizing
Internal Stress in Advanced Composite Materials," J. Composite Materials, 26, 405-431
(1992).

24. Ifju, P.G., Niu, X., Kilday, B.C., Liu, S.C. and Ettinger, S.M., "Residual Strain
Measurement in Composites Using the Cure-referencing Method," J. Experimental
Mechanics, 40, 22-30 (2000).

25. Scida, D., Aboura, Z., Benzeggagh, M.L. and Bocherens, E., "Prediction of the Elastic
Behavior of Hybrid and Non-hybrid Woven Composites," Comp. Sci. and Tech, 57,
1727-1740 (1998).

26. Naik, R.A., "TEXCAD Textile Composite Analysis for Design," NASA Contractor
Report 4639 (1994).










27. Huang, X., Gillespie, Jr., J.W. and Bogetti, T., "Process Induced Stress for Woven Fabric
Thick Section Composite Structures," Comp. Struct., 49, 303-312 (2000).

28. White., S.R. and Kim, Y.K., "Process-induced Residual Stress Analysis of AS4/3501-6
Composite Material," Mech. Of Comp. Mat. And Struct., 5, 153-186 (1998).

29. Whitcomb, J.D., "Three-dimensional Stress Analysis of Plain Weave Composites,"
Composite Materials Fatigue and Fracture (Third Volume), ASTM STP 1110, 417-438
(1991).

30. Strickland, N.M., Yin, W. and Ifju, P.G., "Residual Strain Measurement Analysis of
Woven Composites Using Phase Shifting," Proceedings of the 2007 SEM Conference on
Experimental Mechanics, Springfield, MA, USA, June 4-6, 2007.

31. Karkkainen, R., "Stress Gradient Failure Theory for Textile Structural Composites,"
University of Florida PhD Dissertation (2006).

32. Carvelli V. and Poggi C., "A Homogenization Procedure for the Numerical Analysis of
Woven Fabric Composites," Composites Part A: Applied Science and Manufacturing,
32, 1425-1432 (2001).

33. Speriatu, L., "Temperature Dependent Mechanical Properties of Composite Materials and
Uncertainties in Experimental Measurements," University of Florida PhD Dissertation
(2005).

34. Berman, J.B., and White, S.R., "Theoretical Modeling of Residual and Transformational
Stresses in SMA Composites," Smart Material Structures, 5, 731-743 (1996).









BIOGRAPHICAL SKETCH

Nancy M Strickland was born April 12, 1983, in Pensacola, FL. She spent a significant

portion of her life living in Gulf Shores, AL, before moving to Jacksonville, FL, in 1995 where

she later attended Allen D. Nease High School. After graduating in the top five of her high

school class, she enrolled at the University of Florida in August 2001 where she began her

studies in mechanical and aerospace engineering. Although she had originally intended on

pursuing medical school, her love for the math and sciences drove her to engineering early in her

college studies. During her undergraduate career, she was a member of the Pride of the Sunshine

Marching Band for three years, as well as being very active with Tau Beta Pi and ASME.

In December 2005 she graduated cum laude with bachelor's degrees in mechanical and

aerospace engineering. The following spring she began pursuing her master's degree in

aerospace engineering under the advisement of Dr. Peter Ifju in the Experimental Stress Analysis

Lab (ESA Lab). During her time working in the ESA Lab, her research was focused on plain

weave composites and process induced strains until she graduated in December 2007





PAGE 1

RESIDUAL STRAIN MEASUREMENT OF PLAIN WEAVE COMPOSITES USING THE CURE REFERENCE METHOD By NANCY M. STRICKLAND 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 2007

PAGE 2

2007 Nancy M. Strickland 2

PAGE 3

In memory of my mother. Also to my family, since none of this would ha ve been possible without their encouragement and support. 3

PAGE 4

ACKNOWLEDGMENTS I would like to thank everyone who provide d support and assist ance throughout this research. First, I want to thank Dr. Peter Ifju for all of the advice and su pport he has offered. I would also like to thank my other committee members, Dr. Bhavani Sankar and Dr. Raphael Haftka, for their advice and expertise. Also, this work would have not been possibl e without the help and shared knowledge from Weiqi Yin, Tzu-Chau Chen and Diane Villanueva, my colleagues in the Experimental Stress Analysis Lab. Finally, I would lik e to thank my mom, dad and the rest of my family for all of their support and encouragement they have provided throughout the years. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................7LIST OF FIGURES .........................................................................................................................8ABSTRACT ...................................................................................................................... .............11 CHAPTER 1 INTRODUCTION ................................................................................................................ ..13Woven Composites .................................................................................................................13Residual Stresses and Strains ..................................................................................................14Research Objectives ........................................................................................................... .....152 LITERATURE REVIEW .......................................................................................................17Introduction .................................................................................................................. ...........17Destructive Techniques ........................................................................................................ ..17Hole Drilling Method ......................................................................................................17First Ply Failure Test .......................................................................................................18Non-Destructive Techniques ..................................................................................................18X-Ray Diffraction ............................................................................................................1 8Embedded Sensors ...........................................................................................................19Warpage Measurements ..................................................................................................20Cure Reference Method ...................................................................................................20Analysis Methods Used in Predicting Residual Stresses in Woven Composites ...................203 CRM DEVELOPMENT .........................................................................................................22Principles of the Cure Reference Method ...............................................................................22Moir Interferometry .......................................................................................................... ....23Specimen Preparation .............................................................................................................24Development Process ........................................................................................................... ...25Original Cure Reference Method ....................................................................................25Original cure reference method procedure ...............................................................25Results using original CRM .....................................................................................26Silicone Rubber Based Methods .....................................................................................28Silicone rubber grating using resin from prepreg .....................................................28Results using silicone rubber gra ting using resin from prepreg ...............................29Modifications attempted: Thin silicone rubber grating ...........................................29Modification attempted: Additiona l epoxy cured onto silicone rubber ..................30Results from silicone rubber and 3501-6 .................................................................31 5

PAGE 6

6 Kapton Polyimide Film Based Methods .......................................................................31Results using the Kapton film method ..................................................................32Modifications attempted to the Kapton method ....................................................33High Temperature Epoxy Based Methods ......................................................................33Pooling the epoxy onto the grating ..........................................................................35Spreading the epoxy onto the grating .......................................................................35Spreading the epoxy onto the prepreg before immediately applying to grating ......36Spreading the epoxy and letting it set before placing on autoclave tool ..................37Conclusions .............................................................................................................................374 RESIDUAL STRAIN MEASUREMENTS ...........................................................................51Phase Shifting .........................................................................................................................51Results .....................................................................................................................................52Conclusion .................................................................................................................... ..........535 FEM SIMULATION OF PROC ESS INDUCED STRAINS .................................................59Motivation for Performing FEM Analysis .............................................................................59Model Description ..................................................................................................................59Simulating the Curing Cycle ..................................................................................................60Using FEA to Validate the Experimental Method ..................................................................61Experimental Process Indu ced Residual Stresses ...................................................................62Conclusions .............................................................................................................................636 CONCLUSIONS AND FUTURE WORK .............................................................................73Conclusions .............................................................................................................................73Future Work ............................................................................................................................74 APPENDIX CRM IMAGES AND PHAS E SHIFTED RESULTS ............................................................76LIST OF REFERENCES ...............................................................................................................80BIOGRAPHICAL SKETCH .........................................................................................................83

PAGE 7

LIST OF TABLES Table page 5-1 Material properties used in ABAQUS model for the plain weave composite RVE ..........65 7

PAGE 8

LIST OF FIGURES Figure page 1-1 Representative volume element of the plain weave composite geometry .........................163-1 Schematic of a typical four-beam interferometer ..............................................................393-2 Fixture used to align the diffracti on grating with the autoclave tool. ................................393-3 Oven and vacuum lines used for the curing cycle .............................................................403-4 Vacuum bagging assembly ................................................................................................403-5 Cure cycle used for HMF plain weave prepreg .................................................................413-6 Tool used to separate master grating from the autoclave tool ...........................................413-7 Tools required to make Teflon device ...............................................................................423-8 Specimens made using Original CRM procedure. ...........................................................423-9 Specimen made using original CRM with successful transfer. .......................................433-10 The shapes of the diffraction gratings wh en the master grating is hard from a high temperature epoxy (a) and when a silicone rubber grating is used as the master grating (b)...........................................................................................................................433-11 Fiber reactions with different silicone rubber thicknesses. ..............................................443-12 Silicone rubber grating showing exposed fibers indicating that the grating surface would not be perfectly flat .................................................................................................443-13 Specimen made using thin silicone rubber grating ............................................................453-14 Grating production procedure for sili cone rubber with cured 3501-6 resin ......................453-15 Specimen made using the silic one rubber with cured 3501-6 resin ...................................463-16 Moir images taken from the sp ecimen shown in Figure 3-14. .......................................463-17 Stacking method used for the Kapton polyimide process ..................................................473-18 Specimen made using th e Kapton polyimide method ........................................................473-19 Moire images taken for the spec imen shown in Figure 3-17. ..........................................473-20 Specimen showing the locations of strain discontinuities occurring on peaks of fiber bundles (a) and the corr elated location within the RVE (b). .............................................48 8

PAGE 9

3-21 Various specimens manufactured wi th different polyi mide materials ..............................483-22 Procedure used to produce base gra ting for high temperature epoxy methods .................493-23 Specimens produced by spreading a thin epoxy layer directly onto the aluminized grating surface.. ..................................................................................................................493-24 Pooling the epoxy onto the prepreg as done in the high temper ature epoxy method ........503-25 Sample specimens made after allowing the epoxy to begin to set on the prepreg. ............504-1 Fringe images for Specimen 1 for the U-field(a) a nd V-field (b). .....................................554-2 1 strain contour map via phase shifting from Specimen 1 ................................................554-3 2 strain contour map via phase shifting from Specimen 1 ................................................564-4 Fringe images for Specimen 2 for the U-field (a) and V-field (b). ....................................564-5 1 strain contour map via phase shifting from Specimen 2 ................................................574-6 2 strain contour map via phase shifting from Specimen 2 ................................................574-7 The representative volume elemen t of the plain weave geometry. ....................................585-1 Simulated process induced residu al strains in the 1-direction ( 1) resulting from cooling the RVE from Tref to room temperature and introducing it to atmospheric pressure. Region A is the resin rich zone and Region B has a lower resin content. .........655-2 The contour plot of 1 after the simulated cooling cy cle to room temperature. .................665-3 Plot of 1 along the length of the midline of the warp of the RVE in the 1-direction. .....665-4 FEA 1 results for the RVE from the simulated curing cycle obtained .............................675-5 Experimental phase shifted 1 resulting from the cooling cycle ........................................675-6 Plot of the residual strain, 1, along the resin zone in the 3-direction. ...............................685-7 FEA 3 results for the RVE from the simulated curing cycle ............................................695-8 Experimental phase shifted 3 resulting from the cooling cycle ........................................695-9 Stress contours for the 1and 3-directions using experimental displacement data ...........705-10 The stresses are plotted along the fiber directions in along two different bundles. ...........715-11 The stresses along the resin zones are shown along and transver se to the resin zone. ....72 9

PAGE 10

10 A-1 Silicone Rubber and 3501-6 Resin, U Field ......................................................................76A-2 1 phase shifted results for sili cone rubber and 3501-6 resin .............................................76A-3 Silicone rubber and 3501-6 resin, V Field .........................................................................77A-4 2 phase shifted results for sili cone rubber and 3501-6 resin .............................................77A-5 High temperature epoxy spread and set on prepreg, U Field .............................................78A-6 1 phase shifted results for High temperat ure epoxy spread and set on prepreg ................78A-7 High temperature epoxy spread and set on prepreg, V Field .............................................79

PAGE 11

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 RESIDUAL STRAIN MEASUREMENT OF PLAIN WEAVE COMPOSITES USING THE CURE REFERENCE METHOD By Nancy M. Strickland December 2007 Chair: Peter Ifju Major: Aerospace Engineering Residual stresses develop in composite materials due to the high temperature manufacturing process that is re quired. As the structure cools from the curing temperature, those stresses develop because of a mi smatch between the coefficients of thermal expansion values of the two constituent materials of the composite. If the residual stresses are not accounted for when designing a composite structure, premature fa ilure can develop due to the lack of available strength remaining in the structure. A si gnificant amount of work has been focused on determining these stresses for unidirectional co mposites, but as the w oven geometries become more popular, the research must shift. Woven composites consist of a complex geom etry, so because of that, a full-field measurement technique must be used to obtain the stress patter ns throughout th e repeating unit of the geometry. The cure reference method was designed to measure surf ace residual strains of unidirectional composites with ve ry high sensitivity. That method is extended to the woven composite; however, to do so modifications are re quired to achieve optimal results. The final procedure that is chosen to empl oy for this research is a less time intensive and more repeatable process than the original method suggested fo r the original embodiment of the CRM. The experimental data show that significant strains develop thr oughout the repea ting unit and the 11

PAGE 12

12 strains alternate between highly tensile to highly compressive over a very small distance. The approximate range of strains that are occurring is between -3000 and +2500 A finite element analysis is also completed to verify the experimental method developed and to calculate the residual stresses resulting from the experimental residual strains. The verification analysis shows very good agreement be tween the analytical and experimental data. The tensile strains are occurring over the fiber bu ndles and the compressive strains are occurring in the resin rich zones. The range of strains that is o ccurring experimentally is larger than that predicted by the analysis but that is due to some assumptions made in the finite element model. The residual stresses that are developing are be tween -13 MPa and +4 MPa which are small with respect to the strength of the resin (82.7 MPa) but th ey cannot be ignored.

PAGE 13

CHAPTER 1 INTRODUCTION Composites are composed of two or more diffe rent materials combined to form a single macroscopic structural unit. A common reason for manufacturing a composite material is so that a combination of desired propertie s can be achieved yielding optimal results. It has been shown in studies performed by Griffith that materials in the fiber form yield higher strengths than they do in the bulk form [ 1 ]. However, it is also very difficult to use fibers structurally since they cannot maintain compressive loads and the propertie s transverse to the loading direction are very poor. This can be solved by embedding the fibers in a matrix, or filler material, so that many fibers bound together by the surrounding matrix ca n be used to carry the load. Ultimately, the structural advantages of fibrous materials, such as stiffness and strength, can be used in a low weight structure yielding a part with a high specific stiffness or strength. Composite materials are becoming more widely used by various industries. As the desire increases for advanced material properties that composites provide, more research needs to be continually done so that they can be fully characterized and understood. Woven Composites There are two categories of composites, plie s composed of unidirectional or randomly scattered chopped fibers (laminates ), or fiber bundles woven into a regular pattern (textiles). Unidirectional composites provide strength and stiffn ess properties in the dire ction that the fibers are aligned. By stacking multiple plies and aligning the fibers in different orientations, material properties can be enhanced in desi red directions for the part bei ng designed. The second class of fiber reinforced composites is the woven textile. Similar to the unidirectional composites, the weave pattern can be adjusted so that certain material properties are ach ieved, particularly if more strength/stiffness is desired in one dire ction as opposed to another. Additionally, the 13

PAGE 14

overall stiffness or strength of the part can be enhanced by stacking multiple layers to create a thicker structure. Several weave patterns exis t, including the plain weave and various satin weaves. The focus of this work will only be on the plain weave geometry. As opposed to the unidirectional composite laminates, woven composites exhibit a very complex geometric structure that results in pr operty trends very di fferent from many other materials. Most material classifications yi eld a uniform strain field during tension or compression tests as long as strain concentrations do not exist in the region of interest. In these cases, the average strain is a reas onable estimate of the local strain s. However, this is not the case in woven composites because of the complex ge ometry exhibited by the fibers. Because of this, even though the large scale av eraged strains may appear unifo rm, the strains on the scale of the representative volume element (RVE), shown in Figure 1-1 are not uniform. Typically, under most loading conditions a repeating pattern w ould occur in the strain field indicating very different strain values at various locations in the RVE. A characteris tic of woven composites that is distinctly different from most materials is that in a given loading condition, it could be possible for the structure to exhibit both tens ile and compressive stra ins throughout the RVE. For example, moir interferometry was used to measure the strains transverse to the loading directions on a plain weave composite during a uniaxial tension test. Repeating regions of tensile strain were measured across the specimen even though typically on ly compressive strains would be expected to develop because of the Poisson effect [ 2]. Residual Stresses and Strains The manufacturing process for many composite structures requires a high temperature heating cycle. Those high temperatures are required so that the resin can fully heat and complete its polymerization process as well as wet the fibe rs and finally cure into a hard structure. However, this necessary process introduces a si gnificant amount of residual stresses into the 14

PAGE 15

composite specimen after the heating cycle completes and the specimen cools back down to room temperature. The residual stresses in a composite arise from both chemical and thermal shrinkage [ 3]. The chemical shrinkage is due to polymerizat ion of the resin in wh ich the two monomers in the epoxy come together to form the final compound. The majority of the chemical shrinkage occurs during the initial heating period of the curing cycle and therefore does not contribute to the overall residual stresses. This is because stresses cannot exist before the composite fully cures and the resin transforms to the solid state. However, th ere is some additional chemical shrinkage that does contribute to th e final residual stress levels in the specimen that occurs after the solidification of the resin. On the ply level, a significant amount of the residual stresses develop because of the mismatch that naturally exists between the coe fficients of thermal expansion (CTE) of the two materials. The resin typically has a much higher CTE value than the fiber; therefore a larger amount of thermally induced shrinkage would develop in the matrix. However, the higher strength fibers restrict the cont raction of the matrix. That re striction is what introduces the residual stresses into the composite. Research Objectives The goal of this research was to employ an experimental method to measure the residual strains that develop in the plain weave composite due to the manufacturing process. Because of the complex nature of the material, a full-field ex perimental method was desi red so that a strain contour map could be obtained for the representative volume element of the weave pattern. The process induced residual stresses were to be obtained via finite element modeling by using the experimental displacement data as inputs into the analysis. Additionally, the finite element model will be used to perform a compar ison study to the experimental data. 15

PAGE 16

16 Figure 1 1. Represe n n tative volu m m e element of the plain weave com p p osite geom e e try

PAGE 17

CHAPTER 2 LITERATURE REVIEW Introduction Many methods have been used to experiment ally measure or anal ytically predict the residual stress levels of composite materials. The experimental methods can be divided into destructive and non-destru ctive techniques. Destructive Techniques Hole Drilling Method The hole drilling technique is one of the ol dest and most widely accepted methods for measuring residual stress levels of most materi als. It was first disc overed by Mather in the 1930s that by removing a section of ma terial in a stressed structure, a change in stress state in the remaining material will occur so that static equilibrium can be maintained [ 4]. Doing so results in the material surrounding that hole to relax th erefore altering the dimensions of the hole which correlates with the surrounding strain field. Much advancement to the hole drilling method has been made since the origin of the technique. The development of the hole drilling strain gage rosette ha s allowed more accurate measurements of the area directly surrounding of the blind drilled hole [ 5]. Also, as technology has improved, the gages have become smaller giving the ability to better cap ture the large strains that are developing close to the hole edge th at cannot be measured due to the significant averaging larger gages must perform. The holedrilling method has been applied to orthotropic materials by developing intensiv e analytical relations between the stress and strain fields [ 6,7]. Because of the extensive calculations required for the application to orthotropic materials, either many assumptions must be made to simplify the calculations or a finite element analysis must be performed to gather the coefficients required by the analytical relations. By introducing 17

PAGE 18

interferometric methods to the hole drilling pr ocedure, the strain fi elds surrounding the hole could be measured as opposed to a averaged values close to the hole [ 8,9,10]. However, even those methods have difficulty accurately measur ing the large stress c oncentrations around the hole. Although much advancement has been made in this technique, ther e is a major drawback associated because of the lack of full-field information it provides. First Ply Failure Test The residual stresses that exis t in a composite can also be f ound by the first ply failure test [ 11,12]. A [0/90] specimen was loaded until failur e, and throughout the load ing the strains were recorded. Using elastic stress-s train relations, the approximate lo ad was calculated at which the first ply failed in the test. That value was then compared to the ultimate strength of the ply, and the difference was called the residual stress. Non-Destructive Techniques X-Ray Diffraction X-Ray diffraction was introduced into the co mposites community as a possibility for measuring residual stresses by Predecki and Barrett in 1979 [ 13]. By embedding metallic particles between prepreg plies, the residual stresses of the part iculates could be obtained using X-Ray diffraction. The stresses of the particul ates could then be re lated to the surrounding matrix material to determine the stress valu es of the matrix. This methodology was then extended to measuring the residu al strains of the composite and using numerical modeling via the Eshelby method to determine the residual stresses [ 14]. After having success with the unidirectional laminate, X-Ray diffration was used for measur ing the interlaminar residual strains of woven composites [ 15]. A drawback to this method is that it requires embedding a foreign object into the matrix which could ultimately modify the material properties. 18

PAGE 19

Embedded Sensors Two different types of sensors are commonly embedded in com posite materials to measure residual strains: electrical re sistance strain gages and fiber optics. By embedding the sensor between layers of the composite laminate, the deformation e xperienced by the specimen is recorded by the sensor and afte r using the necessary correctio n factors the strains can be obtained. Electrical resistance strain gages were firs t embedding within laminated composites to measure the residual st rains in the 1970s [ 16,17,18]. The gages were placed between plies of the laminate and on the surface of the specimen dur ing the manufacturing process prior to cure. By doing this, any strains that developed duri ng the curing process would be measured by the strain gages. Daniel made sure that the ga ges and required wiring were chosen so that the measurement equipment did not introduce a ny local thickness differences to limit the introduction of errors in the strain results. The method of embedding strain gages within the composite specimen has also been used to help determine the warpage developed in asymmetric woven laminates combined with laminate theory and shadow moir [ 19 ]. The ability to measure residual strains by embedding fiber optics within composite laminates was developed under the same principle as embedding electrical strain gages, and was first used by Lawrence [ 20,21]. Both an extrinsic Fabry-Pe rot interferometer and a fiber Bragg grating were used to measure the process induced strains in a composite laminate. The fiber optic is laid between plies before cure and is nulled at room te mperature. Assumptions are made that the optic bonds well to the epoxy and that it does not structurally re inforce the specimen. Work has also been done to extend the use of fiber optics to woven composites by using a distributed sensor system [ 22 ]. 19

PAGE 20

Two main drawbacks of using embedded strain sensors are that a foreign object must be placed within the specimen and that only loca lized strain values can be obtained. The introduction of the foreign object is always assumed not to reinforce the material or alter the properties in the area of interest, however, some errors a lthough minimal are bound to develop. Warpage Measurements When assymetric composites are manufactured warpage will ultimately result. Much investigation has been done on determining the warpage of composites and relating that to the residual stresses [ 18,23]. However, this method is not desirable because it is seldom that an assymetric composite is desired and due to th e inherent symmetric nature of plain weave composites, measuring warpage is not feasible. Cure Reference Method The Cure Reference Method (CRM) was develo ped at the University of Florida as a nondestructive technique to measure full-field residual strains of composite materials [ 24]. By applying a diffraction grating to th e specimen at the cure temperature, which is assumed to be a free stress state, moir interferometry can be us ed at room temperature to measure the residual strains that develop from the curing process. This technique is the basis of this work, so an in depth description will follow. Analysis Methods Used in Predicting Residual Stresses in Woven Composites For traditional laminate composites, the classical laminate theory (CLT) can be used to calculate the stresses in a structure once the material properties and st rain data are known. However, in advanced composites, such as woven textiles, the complications of the material geometry no longer allow these straightforward calcu lations. Both analytical and finite element models have been used to predict the process induced strains in woven composites. Several groups have worked to develop analytical models based off of CLT that account for the complex 20

PAGE 21

21 geometries contained in woven composites [ 25, 26, 27]. Both TEXCAD and MESOTEX have been extended to predict the process induced thermal stresses for the plain weave composite. Other work has been done by using finite elemen t models to predict the residual stresses in woven composites, however; some of these mode ls are still either simplified twoor threedimensional models that are not able to comple tely predict the stress field occurring throughout the geometry [ 28, 29].

PAGE 22

CHAPTER 3 CRM DEVELOPMENT The focus of this research was to modify a nd improve the original process proposed for the Cure Reference Method [15]. This technique wa s then to be used to measure and obtain the residual strain field of a plain weave composite. This chapter will go into the details of all methods that were attempted in modifying this experimental technique and explain what was desired when creating a grating to be used for the cure reference method. Principles of the Cure Reference Method The cure reference method (CRM) works in conjunction with moir interferometry to measure the residual strains that develop in a co mposite material because of the high temperature curing cycle required for fabri cation. During the curing cycle, the composite undergoes a transformation process from its original liquid stat e as a prepreg to the solid state once the resin has cured. Until the point it transforms to a solid, it can be assumed to be in a stress-free state. By attaching a diffraction grating to the com posite surface at the point that transformation occurs, any changes in the frequency of the grati ng will only be attributed to the deformations that develop as the specimen cools to room temperature. The diffraction grating was replicated directly onto the specimen from the autoclave tool, which was a piece of ultra low expansion glass so to ensure that the grating frequency would be the same at the cure temperature as it was at ro om temperature. The two main assumptions that this method employs were that the grating bonds perfectly to the specime n surface and that the bonding undergoes the same process as the solidifi cation of the resin in the composite. Also, because it is assumed that the grating adheres to the specimen at the tran sformation point to the solid, the exact time the adhesion occurs does no t need to be known, since the room temperature measurements will always be in re ference to the free-stress state. 22

PAGE 23

After the specimen was given the appropriate time to cure, dict ated by the curing cycle, the composite was separated from the autoclave tool before the cooling cycle began. This allowed the specimen to undergo free contraction as it co oled and ensured that the frequency change measured in the grating was due to the residual strain occurring from the cure temperature. Moir interferometry is then used after the specimen completely cools to room temperature to measure the residual strains. Moir Interferometry Moir interferometry is a laser based optical method that is used to measure full-field strains with extreme sensitivity [ 2]. The sensitivity of this method is 0.417 m, and it provides a contour displacement map throughout the area of in terest which is constructed via interference from light diffracting off of the specimen. A di ffraction grating is adhered to the specimen before tests are conducted and it has an initial frequency matchi ng the reference, or master grating. As the specimen deforms, the frequenc y of the grating changes and that change is directly proportional to the deformation that oc curred. A four-beam interferometer was used for this work and is shown schematically in Figure 3-1 To measure the changes of freque ncy that occur, the interferom eter has to be tuned to the initial, or reference, grating. This is done by directing the light source s towards the specimen so that no, or very few, interference fringes can be observed. Then when the specimen replaces the reference grating in the interferometer, interferen ce of the light sources will occur because of the changes in the grating frequency. As it is applie d to CRM, after the composite cools, the residual strains that develop due to the cooling process can be measured by comparing the specimen grating to the reference grat ing in the interferometer. 23

PAGE 24

The interferometer used in th is work was a self-contained, po rtable moir interferometer (PMI) in the Materials Characterization Lab of th e Civil Engineering Department. It was a fourbeam, 2400 lines/mm interferometer that consisted of a 628nm He-Neon laser. Specimen Preparation The material tested for this research was a plain weave composite manufactured by Cytec Industries. The typical test specimen was a la y-up composed of four 3.0 x 3.0 layers. While stacking the prepreg layers careful attention was directed towa rds aligning the fiber bundles with each other so to obtain minimal misalignment. It was critical to obtain the best alignment possible so that the strains measured via moir interferometry corres ponded with the strains along the fiber directions. Also, prior to baggi ng the specimen and autoclave tool, the alignment of the grating needed to be de termined with respect to the au toclave tool. Using the fixture shown in Figure 3-2 a laser source was reflected off of the grating surface and the diffracted beams were aligned with the markings seen in the figure by adjusting the position and rotation of the autoclave tool with respect to the optical bench. The specimens were required to be cure d in a vacuum oven that is shown in Figure 3-3 Irregardless of the CRM production method used, to prepare the specimen for the vacuum curing cycle it was bagged according to the process de scribed below and shown schematically in Figure 3-4 1. A circular hole was cut into a piece of nonporous release film large enough to expose the grating on the autoclave tool. That piece of film was then laid direc tly onto the autoclave tool leaving the grating exposed. 2. The prepreg was then placed over the release film so that the fiber bundles aligned with the diffraction grating and so that the grat ing was exposed to the composite surface. 3. The composite was then covered with a porous release film, anothe r non-release film layer and then a breather clot h that was used to pull an even vacuum over the entire specimen. 24

PAGE 25

4. The autoclave tool and respective layers were th en placed inside a sealed air tight vacuum bag that was connected to the vacuum line of the oven and checked for air leaks before beginning the cure cycle. After bagging, the specimens were cured accord ing to the cure cycle stipulated by the manufacturer shown in Figure 3-5 The critical step when implementing the CRM is separating the specimen from the autoclave to ol after it has fully cured, and th at point is also shown in the figure. However, that separation must be done be fore the cure cycle reaches the cooling process so that the entire cooling regime can be measured in the deformations. If the specimen is not separated from the tool, it is not allowed to freel y contract and any stra ins that are measured would be from an unknown temperature. Development Process When work initially began on this research, a few CRM gratings were available that were produced using the process described in the orig inal CRM method. However, after those were used, new gratings needed to be produced for more experiments to be conducted. During that replication it was determined that the existing procedure needed to be modified so that it could be applied to woven composites and so that it would require a less time demanding procedure to produce a single grating. Original Cure Reference Method Original cure reference method procedure The original fabrication method proposed fo r CRM was a time intensive process that required approximately sixty hours, over a total of seven steps, to produce a single autoclave tool [ 24]. After the four available gratings were us ed to manufacture test specimens, new gratings had to be produced. The following procedur e was used to replic ate that process. 1. A silicone rubber grating was replicated at room temperature onto a piece of ultra low expansion glass (3.0 x 4.0 x 0.5) from a Photoresist master diffraction grating. The silicone rubber was a two-part mixture, GE 615RV. 25

PAGE 26

2. The master silicone grating was used to re plicate the diffraction grating onto another piece of Astrosital (3.0 x 4.0 x 0.5) using a high temperature epoxy (Epon 862 with curing agent W). This would become the in termediate grating. The parts were mixed together using a mass ratio of 100:26.4 and then heated for five minutes at 130C. Afterwards it was centrifuged for five mi nutes to remove any bubbles before being applied to the grating for repl ication. The epoxy was then allowed to cure for ten hours in a 130C oven before the gratings could be separated. The gratings were separated using the device shown in Figure 3-6 3. Two aluminum layers were deposited onto th e grating surface us ing vacuum deposition with a film of Photo-flo used as a parting agent between th e two layers. The Photo-flo was allowed to sit for twenty-four hours before the second layer of aluminum was applied. 4. The grating was then replicated onto the autoclave t oo, a piece of Astros ital glass (5.0 x 5.0 x 0.5) using an epoxy resin, 3501-6. The resin was heated for ten minutes at 175C before a vacuum was also applied to the ove n for an additional fifteen minutes. While the resin was heating, the inte rmediate grating and the auto clave tool were also placed into the oven to preheat. Af ter liquefying the resin, it was then pooled onto the autoclave tool and the intermediate grating was lowered onto it for replication. The resin was allowed to cure for ten hours at 175C before being separated. 5. A single layer of aluminum was applied to the autoclave tool over the resin layer. 6. A piece of Teflon was tightened over a pie ce of glass using the device shown in Figure 37. The device was then placed into the ove n for an hour at 175C and the Teflon was tightened at several points in that time to ensure a smooth surface. After the Teflon was tightened, the 3501-6 resin was heated and vacuumed using the same procedure as in Step 4. Afterwards a small amount was pooled onto the aluminum surface of the autoclave tool and the Teflon device was car efully lowered down onto it so that no air bubbles were trapped. The autoclave tool and Teflon device were then heated in a 175C oven for ten hours. After that point the Te flon device was unscrewed and the Teflon was peeled off of the epoxy at a steep angle so that the epoxy does not separate from the aluminum. Results using original CRM After following the above procedure, grati ng production was approximately a thirty percent success rate. The method was extremely sensitive to the th ickness of the final layer of resin applied to the autoclave tool. If that layer was allowed to be too thick, as it cooled it would crack at several locations in the grating. In seve ral instances, as the e poxy layer cracked it would 26

PAGE 27

chip the glass. When this occurred, that gra ting was no longer usable and the six step process previously described needed to be completed again. Four specimens were made using the pre-fabr icated gratings, and fringe images were captured. The number of fringes was drastically lower than expected and the fringes were indicating almost zero strain. It was thought that the epoxy on those gratings was too stiff, therefore not allowing any shrinkage to occur. Therefore, new gra tings were fabricated using the described procedure and those would be used for testing. When specimens were produced using the gra tings made from the procedure explained above, several issues arose when transferring th e grating onto the composite and separating the composite from the autoclave tool. Some exampl es of specimens produced from these gratings are shown in Figure 3-8 In most gratings that were produc ed, the grating was either severely cracked, as in Figure 3-8 (a), making it impossible to obtain us able fringe images or the grating transfer was very incomplete as in Figure 3-8 (b). It was believed th at the reason for the poor grating transfer was due to insufficient resi n available in the prepreg. The grating did successfully transfer in a couple tests. Ho wever, due to the low resin volume present, discontinuous fringes developed at the locations which should be resin rich zones. As can be seen in Figure 3-9 (a), those areas are completely void of resin in many locations, and because of that, discontinuous fringes developed as shown in Figure 3-9 (b). However, even though a few gratings were successfully produced, because of the difficulty and time required for the procedure it was deci ded that efforts should be targeted towards modifying the process to one that would be more repeatable and less time demanding. The following sections describe the methods attempted in achieving that goal. 27

PAGE 28

Silicone Rubber Based Methods As opposed to other methods that were tested, this technique initially required only one grating replication. A silicone rubber grating was replicated on to the autoclav e tool and was used to transfer the CRM grating onto the specime n. Philosophically, the s ilicone rubber grating would provide two distinct advantages towards improving the CRM procedure. Firstly, the use of silicone rubber would provide easy separation between the autoclave tool and the composite specimen. By improving the separation it helps to ensure that the grati ng surface will separate cleanly so that smooth and continuous fringe inform ation could be obtained. Also, if the gratings are difficult to separate, some bending of the specimen would inevitably occur which could introduce erroneous information into the diffraction grating. Secondly, the production process for the grating would be drastically reduced, by using silicone rubber, th erefore shortening the time required for one test. Silicone Rubber Grating Using Resin from Prepreg The first attempt at this method assumed that there was sufficient resi n in the prepreg. By making this assumption, as the composite cured, th e resin within the prepreg would replicate the diffraction grating without any additional epoxy be ing applied. The detailed procedure follows the steps below: 1. A diffraction grating was replicated using a two part silicone rubber onto a piece of Astrosital ultra low expansion glass at room temperature using a Photoresist master grating. The silicone rubber (GE 615RV) wa s carefully mixed and centrifuged before use. Three days were allowed for the silicone rubber to fully cure at room temperature. 2. The prepreg was then placed directly on the silicone rubber grating before being placed into the autoclave. When the resin of the prepreg cured, the diffrac tion grated would be formed in the resin of the composite surface. 28

PAGE 29

Results using silicone rubber grat ing using resin from prepreg When silicone rubber was used to replicate the grating onto the autoclave tool, the resulting fringe image was not able to be obtained using the interferometer because of the lack of a smooth reflective surface on the specimen. This occurred because the silicone rubber was not hard enough to prevent the fibers in the prepreg from indenting the grating. Ideally, the grating surface should appear similar to that in Figure 3-10 (a) so that all light directed towards the grating is reflected directly b ack towards the camera. However, when a soft medium, such as silicone rubber, is used to transfer the grating, th e fibers of the composite cannot remain perfectly rigid and eventually sink in to the rubber grating as shown in Figure 3-11 Therefore, the resulting grating surface appe ars similar to that shown in Figure 3-10 (b). As can be seen in the figure, because the grating su rface is not completely smooth, the light is redirected when it reflects away from the surface providing an inco mplete and inaccurate fringe image at the camera. This can also be seen by visually inspecting the grating on the specimen, Figure 3-12 as one can see individual fibers in the grating area which indicate d that the grating surface was not perfectly smooth. Modifications attempted: Thin silicone rubber grating An attempt was made to create a thin silicone rubber grating that would behave stiffer so that a flatter grating surface could be created on the specimen. It was believed that if the thickness of the silicone rubber was of a similar magnitude to that of the fiber diameter, significant indentation would be inevitable yielding a grating surf ace similar to that achieved in Figure 3-11 (b). However, if the thic kness of the silicone rubber was held to a magnitude much less than that of the fiber diameter as in Figure 3-11 (d), it was hypothesized that the grating surface would flatten out and yield a more comp lete fringe image. As can be seen in Figure 329

PAGE 30

13, making this modification did not improve the grating surface and a moir image was still unattainable. Modification attempted: Addition al epoxy cured onto silicone rubber The next modification was to add an additiona l layer of epoxy between the grating and the prepreg during the lay-up process. This a dditional epoxy was pooled onto the silicone rubber grating immediately before the prepreg was laid down. By doing this, it was thought that better replication could occur between the reference gr ating and the specimen due to the addition epoxy available to form the grating. However, this pr oved not to be the case after attempting this procedure after using both 3501-6 epoxy-resin and a two-part epoxy, Epon 862 with Curing Agent W. Pooling the epoxy directly onto th e grating before immediately prior to the cu re cycle did not provide desirable resu lts. Therefore, the next modificatio n was to cure a layer of epoxy onto the silicone rubber grating, and then at the transformation point in the cure cycle that grating would adhere to the specimen under the same pr inciple that was used in the original CRM. Because of the very low viscosity of the Epon 862, continuous gratings could not be produced because of the repellant nature of the silicone rubber (GE 615RV). Therefore, the epoxy was changed to the 3501-6 resin. The procedure used for this process is described below and is shown graphically in Figure 3-15 : 1. A silicone rubber grating was replicated at room temperature onto a piece of Astrosital glass (5.0 x 5.0 x 0.5) using a two part mi xture of GE 6428. That silicone rubber had a short working life which allowed for quick er replication. The silicone rubber was allowed to cure for one hour before the gratings were separated 2. A Teflon device was prepared using the same pro cess as in Step 6 in the previous section. After the Teflon was tightened on the device, a small amount of 3501-6 was heated for ten minutes before a vacuum was also applied for additional fifteen minutes. After the resin had been sufficiently vacuumed, it wa s the pooled onto the si licone rubber grating and the Teflon device was carefully lowered down so as to not trap any air bubbles within the epoxy. 30

PAGE 31

3. After allowing the epoxy to cure for ten hours in a 175C oven, the Teflon was carefully removed by peeling it away from the grati ng surface at a sharp angle. The exposed epoxy grating would later adhere to the prepreg during the cure cycle. Results from Silicone rubber and 3501-6 After producing several gratings by this met hod, the diffraction grating was found to crack after it cooled to room temperature, similar to the problem that was encountered when the 3501-6 resin was used in trying to reproduce the or iginal CRM method. However, the cracking was much less severe, so the gratings were still capable of being used from this method. A test specimen produced using this CRM method is shown in Figure 3-15 and the resulting moir images are shown for the U and V fields in Figure 3-16 (a) and (b) respective ly. An example of the cracks occurring in the diffraction grati ng can also be seen in the images of Figure 3-16 as there is a crack visible along the bo ttom and left side of both images The existence of that crack leads to discontinuous fringes along that area. The poi nt defects that can also be seen are due to air bubbles that could not be completely rem oved from the epoxy before it was cured. The presence of those defects introduces significant strain concentrations that do not correctly describe the actual strain state of the specimen. A slight repeat ing pattern was evident in the fringes shown in the moir images of Figure 3-16 ; however, it was not as prominent as was expected. Therefore, because of the inconsistent diffraction gratings that were being replicated and the low quality fringes, new me thods needed to be attempted. Kapton Polyimide Film Based Methods Because many of the problems encountered when attempting the silicone rubber based methods were due to the inability of replicating a perfectly flat grating onto the specimen, it was thought that if a stiffer medium wa s used to replicate the grating, those issues could be resolved. If a material with a glass transition temperature (Tg) lower than the curing temperature of the composite was used, it would adhere to the specimen at cure. Then by assuming perfect 31

PAGE 32

adhesion to the specimen, the deformations occurring in the composite would transfer to the medium as it cooled down. This is similar in principle to the original CRM; however, instead of transferring the displace ment information through an epoxy la yer, it was transferred through a polyimide medium. When selecting the medium to use for this technique, the strength needed to be lower than that of the resin in the composite so the deformations coul d be transferred and the Tg needed to be lower than the cure temperat ure. The production time was very minimal using this procedure and required approximately thirty minutes prior to manufacturing a specimen. The process is described below and the stacking sequence is shown in Figure 3-17 1. A master silicone rubber grat ing was replicated from a ma ster Photoresist diffraction grating onto a piece of Astros ital ultra low expansion glas s (5.0 x 5.0 x 0.5) using a two part silicone rubber (GE 615RV). 2. A thin layer of Epon 862 with curing agent W was pooled di rectly onto the silicone rubber grating at room temperature. Th e high temperature epoxy was mixed using a 100:26.4 mass ratio of epoxy to curing agent. It was then heated for five minutes to reduce viscosity and then centrifuged for five minutes. The purpose of this layer was to replicate the grating from the autocl ave tool onto the Kapton film. 3. A small piece of the Kapton film (1.0 x 1.0) was carefully lowered onto the epoxy pool so as to not entrap ai r bubbles within the epoxy. 4. An additional layer of the high temperature epoxy (Epon 862 with curing agent W) was pooled directly onto the Kapton film. This la yer was carefully spread into a thin layer so as to not intr oduce any air bubbles. 5. The prepreg was then lowered onto the layer of uncured epoxy before being cured in the autoclave. This method was originally completed wit hout performing Step 4. However, due to insufficient resin in the prepreg, the additiona l layer of epoxy was added in hopes that better adhesion would occur between the specimen and the Kapton film. Results using the Kapton film method An example specimen using this method is shown in Figure 3-18 and the fringe images are shown in Figure 3-19 for both the U and V fields. As can be seen from the moir images, 32

PAGE 33

discontinuous fringes are developing at very regular locations throughout the field of view. These locations correlate with th e areas in the weave that are cl osest to the surface as illustrated in Figure 3-20 The discontinuities were occurring because the epoxy that was pooled in Step 4 wicked away from fiber bundle peaks and concentrat ed in the resin pockets. That produced areas in the film that could not fully adhere to the specimen which ultimately produced small air pockets between the composite and the film. Therefore, because the displacement measurements in those areas were only record ing the deformations caused from the air pockets, extremely high density fringes developed resulting in discontin uities within the overall fringe patterns. Modifications attempted to the Kapton method Different combinations of epoxy and polyimide films were used to modify this method in attempts of achieving a replicable grating. Th e epoxies used were Epon 862 with curing agent W and Tra-Con F230. After attempting various materials, the resulting specimens produced gratings very similar to those shown in Figure 3-21 The discontinuities could not be avoided after attempting several combinations of e poxies and polyimide films. Although highly repeating fringes were developing, the images coul d not be correctly analy zed due to the extreme strain concentrations at so many locations. High Temperature Epoxy Based Methods The high temperature epoxy based methods tried to correct the main problems that arose from the previous attempts. It was decided that the features that needed to be included in the grating production were: easy separation of the specimen from the autoclave tool, the master grating had to be hard and epoxy needed to be able to cover and fill the weave pattern. In some of the other methods attempted, separation of the specimen from the autoclave tool was difficult. This resulted in incomplete grating transfer and it introduced bending in to the specimen during separation which could give inaccura te results in the fringe maps. It was decided that the best 33

PAGE 34

separations had occurred when two layers of aluminum were applied onto the base grating with a layer of Photo-flo between them. Through the co mbination of aluminum layers and a release film between the prepreg and th e autoclave tool, separation shoul d become significantly easier. After using silicone rubber as the base grating material for th e autoclave tool, it was apparent that it was too soft to yield a su ccessful grating transfer. Theref ore, it was preferable that the base grating was to be made from a hard high te mperature epoxy. The initial four steps were the same for all of the attempts made using this method. Those steps are described below and shown schematically in Figure 3-22 : 1. A silicone rubber grating was replicated at room temperature onto a piece of ultra low expansion glass (3.0 x 4.0 x 0.5) from a Photoresist master diffraction grating. The silicone rubber was a two-part mixture, GE 615RV. 2. The master silicone grating was used to re plicate the diffraction grating onto another piece of Astrosital (3.0 x 4.0 x 0.5) using a high temperature epoxy (Epon 862 with curing agent W). The parts were mixed toge ther using a mass ratio of 100:26.4 and then heated for five minutes at 130C. Afterw ards it was centrifuged for five minutes to remove any bubbles before being applied to the grating for replication. The epoxy was then allowed to cure for ten hours in a 130C oven before the gratings could be separated. 3. Two aluminum layers were deposited onto th e grating surface us ing vacuum deposition with a film of Photo-flo used as a parting agent between th e two layers. The Photo-flo was allowed to sit for twenty-four hours before the second layer of aluminum was applied. Several modifications were made for the re mainder of this technique which included applying additional epoxy between the prepreg and grating so that the voids and surfaces were sufficiently filled. In each method that was attempted, both the Tra-Con F230 and the Epon 862 with curing agent W were used to test the proced ure because the two epoxies have very different viscosities. The four methods that were tried involved pooling the epoxy onto the grating, spreading the epoxy onto the grating before appl ying the prepreg, spreading the epoxy onto the prepreg before immediately applying to gra ting and spreading the epoxy onto the prepreg and 34

PAGE 35

allowing the epoxy to begin to set before applying the prepreg to the grating. Those processes and reasons for doing so are descri bed in the following sections. Pooling the epoxy onto the grating The philosophy behind this attempt was that additional epoxy needed to be applied between the grating and the prepreg, and the simp lest method to achieve this was to pool the epoxy directly onto the grating. Th e procedure is described below: 1. A thin layer of epoxy was pooled directly on to the aluminized grating. Epon 862 with Curing Agent W, Tra-Con F230 and 3501-6 resin were all used during different attempts at this procedure. 2. The prepreg was then carefully lowered onto the pool of epoxy so as to not introduce any air bubbles. After producing several specimens through this technique, several flaws were discovered. By placing the prepreg onto a single pool of epoxy, a uniform epoxy distribution was difficult to achieve. With the majority of the epoxy localiz ing in one location, warpage was introduced into the specimen at the pool location. All three of the epoxies tried in this approach resulted in some localized warpage and pooling. Because of that, any strain measurements that could have been taken would not have been true cooling induced deformations. In correct information would have been recorded because the resul ting strains would be strongly influenced by the epoxy instead of the fiber weave. Spreading the epoxy onto the grating It was thought that the coverage of the epoxy onto the prepreg could be improved if the epoxy was initially spread into a thin layer on the aluminized surface prior to laying down the prepreg. This would also prevent areas with significant concentratio ns of epoxy which would introduce warpage into the specimen as was obt ained by the previous method. The procedure used was the same presented in the previous s ection; however, before lowering the prepreg onto 35

PAGE 36

the grating, the epoxy was carefully spread into a thin layer so that air bubbles were not introduced. This method did produce slightly better resu lts; however, the gratings on the specimens were not fully continuous. There were many in stances that the resin pockets did not have sufficient resin, therefore producing a discontinuity in th e grating surface. Another problem was that there were several locations where there was insufficient epoxy where the fiber bundles were closest to the surface which resu lted in discontinuous fringes to develop. Some examples of specimens made using this method are shown in Figure 3-23 Spreading the epoxy onto the prepreg be fore immediately appl ying to grating Many issues seemed to be caused from having insufficient epoxy in th e resin pockets and a lack of complete coverage over the fiber bundles in the previous two techniques. Therefore, to attempt to improve the coverage by the epoxy onto the specimen, the epoxy was spread into a thin layer on the prepreg before it was placed do wn onto the grating. The prepreg was prepared for curing as described in the following steps: 1. A small amount of epoxy was pooled onto th e center of the prepreg, as shown in Figure 3-24 (a). 2. Using a flat edged spatula, that pool of epoxy was spread in to a thin layer so that it covered the area of inte rest on the specimen ( Figure 3-24 (b)). This was done carefully so that air bubbles were not inadvertently introduced which would cause discontinuities in the fringes. 3. The specimen was then lowered onto the alum inized autoclave tool so that the epoxy region matched up with the grating area on the glass. The grating surfaces did show improvement usi ng this method. However, there were still regions yielding discontinuous fringe s caused either by flawed areas at the resin pock ets or at the surface contact zones. It was thought that this was still oc curring because although the epoxy 36

PAGE 37

was spread onto the composite fil ling those voids, when the prepreg was pressed onto the grating the epoxy wicked away from those areas, still leaving voids. Spreading the epoxy and letting it set before placing on autoclave tool By examining the gratings on the specimens, it appeared that the epoxy layer intended to fill the resin pockets and cover the fiber bundles was not remaining there after the prepreg was placed onto the autoclave tool. The next attempt allowed that layer of epoxy to begin to cure so it would not easily wick away from the critical locations. The pr ocess followed was identical to that in the previous section ex cept that after the layer of epox y was spread onto the prepreg it was allowed to sit and partially cure before it was placed onto the autoclave tool. Examples of specimens made using this procedure are shown in Figure 3-25 As can be seen from the figure, the transferred gratings were repeatedly smoot h and continuous. Also, all of the specimens produced from this technique were easy to separate from the autoclave tool. The fringe images, which will be included in Chapter 4, also yielded areas within the region of interest of continuous fringes with minimal defects which wa s ultimately the main goal so that continuous strain contour maps could be obtained. Conclusions The original CRM process that was developed was a time intensive method that was found to be very difficult to repeat. In addition, becau se of the complex geometry of the plain weave composite, the original method was not able to re peatedly separate from the autoclave tool to produce a continuous grating. Theref ore, the focus of the research had to be shifted to modifying that technique to be applied to woven composites. Several methods were attempted using differe nt philosophies to tr ansfer a diffraction grating from the autoclave tool to the specimen at the point of the liquid to solid transformation 37

PAGE 38

point of the resin. However, most methods yiel ded undesirable results in either the separation abilities or the quality of the final grating and resulting fringe images. The last method attempted, allowing the spre ad epoxy to begin to cure on the prepreg before being placed onto the autoclave tool, pr ovided the most repeatable and highest quality results. Therefore, for the experimental testi ng, that method will be used to measure the process induced residual strains developing in the specimen. Those results are examined in Chapter 4. 38

PAGE 39

Figure 3-1. Schematic of a typi cal four-beam interferometer Figure 3-2. Fixture used to align the diffraction grati ng with the autoclave tool. 39

PAGE 40

Figure 3 Figure 3 3. Oven a n 4. Vacuu m n d vacuum li n m bagging as s n es used fo r 40 s embly r the curing c c ycle

PAGE 41

Figure 3-5. Cure cycle used for HMF plain weave prepreg Figure 3-6. Tool used to separate ma ster grating from the autoclave tool 41

PAGE 42

Figure 3-7. Tools required to make Teflon device Figure 3-8. Specimens made using Original CRM procedure. Cracking along resin lines is shown in (a) and incomplete separation is shown in (b). 42

PAGE 43

Figure 3 Figure 3 9. Specim e seen in ( a 2. The sha p temperat u (b). e n made usi n a ) and the re s p es of the di f u re epoxy (a ) n g original C s ulting disc o 43 f fraction gr a ) and when a C RM with s u o ntinuous fri u ccessful tra n nges for the n sfer. Regi o U-field can o ns void of r be seen in ( r esin ( b) a tings when t a silicone ru b t he master g b ber gratin g g rating is ha r g is used as t h r d from a hi g h e master g r g h r ating

PAGE 44

Figure 3-3. Fiber reactions with different silicone rubber thickne sses. Figures (a) and (c) show the initial conditions for di fferent silicone rubber thic knesses. During curing the fibers compress into the silicone rubber as seen in (b) and (d) Figure 3-4. Silicone rubber grating showing expos ed fibers indicating that the grating surface would not be perfectly flat 44

PAGE 45

Figure 3-5. Specimen made using thin silicone rubber grating Figure 3-6. Grating producti on procedure for silicone r ubber with cured 3501-6 resin 45

PAGE 46

Figure 3-7. Specimen made using the si licone rubber with cured 3501-6 resin Figure 3-8. Moir images take n from the specimen shown in Figure 3-16 The U and V fields are shown in (a) and (b) respectively. 46

PAGE 47

Figure 3-9. Stacking method used fo r the Kapton polyimide process Figure 3-10. Specimen made us ing the Kapton polyimide method Figure 3-19. Moire images take n for the specimen shown in Figure 3-19 The U and V fields are shown in (a) and (b) resp ectively. Fringe discontinuitie s can be seen at several locations in both images. 47

PAGE 48

Figure 3-11. Specimen showing the locations of strain discontinuities occurring on peaks of fiber bundles (a) and the correlated location within the RVE (b). Figure 3-12. Various specimens manufactur ed with different polyimide materials 48

PAGE 49

Figure 3-13. Procedure used to produce base grating for high temperature epoxy methods Figure 3-14. Specimens produced by spreading a thin epoxy layer directly onto the aluminized grating surface. Several lo cations on both gratings can be noticed that include many discontinuities in the grating surface. 49

PAGE 50

50 Figure 3-15. Pooling the epoxy onto the prepreg as done in the high temperature epoxy method (a) and spreading the epoxy into a thin layer that fills the resin pockets and covers the fiber bundles (b). Figure 3-16. Sample specimens made after allo wing the epoxy to begin to set on the prepreg.

PAGE 51

CHAPTER 4 RESIDUAL STRAIN MEASUREMENTS Phase Shifting Moir fringe patterns develop by interference caused when two light sources reflecting off a common location. Within that interference there is a corresponding in tensity and phase associated at each location within timage as can be seen from he fringe Equation 1 where s is the Equation 1 intensity and is the phase. Therefore, the fringe images can be analyzed using either the intensity or the phase information. Traditionally, the intensity based techniques have been used. To calculate the strain at a desired location, the number of fringes in that area, Ni, and the exact locations between those fringes, xi must be measured. Then using Equatio 4 the strains can be calculated. n 2 Equation 2 Equation 3 Equation 4 However, this method is not always desired as it is rather time intensive and errors in the results are inevitable because of the uncertainty involve d in choosing the exact location for a fringe. The second method uses the phase information from Equation 1 to calculate the strain information. However, because that expression uses a cosine relationship, it is impossible to determine whether the phase is positive or negati ve from a single fringe image because both have 51

PAGE 52

the same intensity value. This problem is solved by introducing a know n carrier fringe pattern, or phase ramp into teas i qu h equation s seen in Eation 5 Equation 5 By capturing several fringe images that are shifted by that known phase ramp; the complete phase information can be obtained for the entire region of interest. The strains can then be calculated using Equations 6-8. Equation 6 Equation 7 Equation 8 For this research an automated analysis to ol was designed for Ma tlab by Weiqi Yin (a student in the experimental stress analysis lab at the University of Florida) [ 30]. By inputting four sequentially shifted fringe images into the program, full-field di splacement and strain information could be obtained. This was critical for the research since full-field information was desired so the strain trends th roughout the RVE could be determined. Had an intensity based method, or another non-automated technique been used, the ability to gather full-field information would have been drastically reduced. Results The fringe images are shown in Figure 4-1 and Figure 4-4 and the phase shifted residual strain results for two specimens are shown in Figures 4-2 4-3, 4-5 and 4-6 The maximum repeating tensile strains that were developing were approximately 2500 and the maximum 52

PAGE 53

repeating compressive strains were approximately -3000 in both the U and V fields. Additional specimens and their resulting strain contour maps are included in Appendix A. The strain pattern that developed can be explai ned by relating the trends to the geometry of the weave. As can be seen in Figure 4-7 there are two main regions within the RVE, the resin and the fiber regions. The resi n zone expands the entire leng th of the RVE between fiber bundles. However, along that length, the resin de pth from the surface changes as it crosses over transverse bundles. The fiber regions have much lower surface resin content levels since they are much closer to the surface. The regions of high tensile strains seen in th e figures were occurring along the tops of the fiber bundles in the fiber regions in both fields of view, wherea s the compressive strains were developing throughout the resin zone s. The CTE value for the resi n is significantly larger than that of the fiber, so as the composite cools to room temperature, a considerable amount of residual strains develop w ithin the resin because the fiber bun dles restrict free contraction in those areas. The tensile regions that developed along the fiber bundl es were a direct result from the compressive resin regions. To maintain static equilibrium, the thin re sin zones that cover the fibers must undergo tension to accommodate for the large co mpressive areas. It can also be noted that th e strain patterns and values ar e very similar for the U and V fields. That was expected to occur because of the symmetry of the weave geometry. Conclusion The improved CRM was used to measure the pr ocess induced residual strains for the plain weave composite specimen. By using an automated phase shifting analysis method, the strains throughout the entire region of in terest were calculated. Havi ng that information allowed the strain trends throughout the RVE to be presented. Significant strains were found to develop within the RVE, both in compression and in tension. The displacement information that was 53

PAGE 54

obtained from that data will be used in Chapter 5 to numerically determine the residual stresses that were occurring via finite element method. After examining the strain maps for severa l specimens, including those produced by other CRM processes in Appendix A, it can be seen that fairly repeatable strain values were obtained in the extreme strain regions. Although all of the methods tested throughout this research were not able to produce strain contour maps with the definition that was obtained using the final method, the overall strain values obtained were fairly constant between all methods. Therefore, the results that were presented in this chapter do appear repeatable. Analy tical validation will be shown in Chapter 5 which discusses the simula ted cooling cycle that was completed within ABAQUS. 54

PAGE 55

Figure 4-1. Fringe images for Specimen 1 for the U-field(a) and V-field (b). Figure 4-17. 1 strain contour map via phase shifting from Specimen 1 55

PAGE 56

Figure 4-3. 2 strain contour map via phase shifting from Specimen 1 Figure 4-4. Fringe images for Specimen 2 for the U-field (a) and V-field (b). 56

PAGE 57

Figure 4-5. 1 strain contour map via phase shifting from Specimen 2 Figure 4-6. 2 strain contour map via phase shifting from Specimen 2 57

PAGE 58

58 Figure 4-7. The representative volume el ement of the plain weave geometry.

PAGE 59

CHAPTER 5 FEM SIMULATION OF PROC ESS INDUCED STRAINS Motivation for Performing FEM Analysis Because of the complexity of the woven geometry, finite element modeling was the preferred analysis method due to the full-fi eld capabilities it provides and because most analytical models are highly computation inte nsive and require many si mplifications. There were two ultimate goals for performing the finite element analysis. Those were to compare the results obtained experimentally by simulating the curing cycle and to determine the experimental residual stresses using the disp lacement data. Although experimental validation could be performed using a unidirectional laminate and comparing against pr evious tests, the same could not be done for the woven composite. Therefore, the validation was important for this research to ensure that the data collected was reasonabl e, taking into account both the mechanical and thermal properties of the material. Model Description The model used for the analysis was one deve loped by Karkkainen of the representative volume element for the plain weave composite [31]. The geometry of the model was taken from a literature source that had documented the dimensions requ ired when constructing the weave pattern and had taken measurements of a typical plain weave fabric us ing SEM and microscopes [ 32]. Two materials were created within the model, that of the matrix and that of the weave. The weave was assigned the properties of a uni directional laminate, AS4/3501-6, taken from experimental data [ 33]. However, because of the undulation that occurs in the weave, the properties could not be a ssigned directly to the entire fibe r bundle. Instead, when building the model Karkkainen assigned a local material coordi nate system to each element so that when the 59

PAGE 60

laminate properties were applied they would correctly align with th e local direction of the fibers. By doing this, only one material needed to be de fined for the weave, and it could be applied for the warp and fill yarns. The temperature depe ndent properties for the matrix, 3501-6, were taken from the literature [ 34]. All of the material propertie s used in the model are shown in Table 5-1 Simulating the Curing Cycle The application of temperature and pressure fields was required to model the cooling process to simulate both the change in temper ature and the releasing of the vacuum surrounding the specimen. When the temperature dependent CTE data was input into the model, the reference temperature was chosen to be 130C so that any temperatur e field applied would simulate cooling from the cure temperature. The room temperature conditions were modeled by applying a full-field temperature of 22C and a hydrostatic pressure load of 101.3kPa. To prevent rigid body motion, while allowing full deform ations of the model, one node was fixed in all three translational directions. The results are shown in Figure 5-1 and Figure 5-2 Plots were also created along the fiber bundles to illustrate the change in sign that occurred between the tension and compressive regions. As can be seen in Region A of Figure 5-1 large compressive strains are developing within the resin rich z ones of the RVE. Conversely, the strains are highly tensile on top of the fiber bundles where less resin is present. Those tensile strains are being induced from the surrounding compressive resin rich zones. As th e resin rich zones contract, the thinner resin zones are forced into tension to satisfy equilibrium. Because of this phenomenon, undulating strain patterns develop as a function of the amount of resi n in that region. Figure 5-3 shows this occurring, as 1 varies between tensile and compressive values along the length of the fiber bundle in the 1-direction. Those trends were expected to occu r because of the significantly 60

PAGE 61

larger CTE value of the resin as opposed to the fiber bundles and the strain values show good agreement to the experimental data. Using FEA to Compare to the Experimental Method Because it was not feasible to use experimental verification for the developed CRM procedure, it was essential to simulate the cooling process associated with the cure cycle to model the residual strains that developed. By performing a comparison between the two data sources, the validity of the experi mental results could be obtained. It would not be expected for the results to match perfectly as the finite mode l did not account for any chemical shrinkage that would be occurring in the actual specimen. Also, the geometry of the model and the actual weave geometry used for this research were not exactly the same. The geometry for the model was taken from a literature source when the model was developed. However, the actual geometries of the materials tested were close to those used in the model, so the results provide good approximations for the true values. Figure 5-4 shows the 1 results for the RVE calculated in ABAQUS with the schematic of the unit cell overlaid to show how the strain values relate to the weave geometry. For comparison, the contour plot of th e experimental data obtained for 1 using the CRM is shown in Figure 5-5 By comparing the two plots for the U fiel d data, it can be seen that good agreement exists between the two methods. The overall trends throughout the RVE are the same. The values differ slightly in some re gions. One case in particular, is that along the fiber bundles of the experimental data, extremely high tensile strains develop, approximately 2500 verse the maximum strain developing in the fin ite element model was approximately 500 One reason that could cause such a significant difference woul d be the amount of resin in the specimen as opposed to that in the model. Through the differe nt procedures that had been tested it became apparent that the resin level was low in the pr epreg. Having low resin content over the fibers 61

PAGE 62

would yield higher strains due to the lack of material to s upport the surrounding compressive strains. Therefore, it does seem reasonable th at such a large discrepancy would occur in the values for the maximum tensile strains. The undulating compressive strain values along the resin zones in th e 3-direction shown good agreement between the two plots. Both plots show that the strain s in those areas are remaining between approximately -3500 and -1000 The actual repea ting pattern can be better seen in Figure 5-6 Similar plots were created for 3 and those are shown in Figure 5-7 and in Figure 5-8 for the FEA results and experimental data respectively. Like the plots for 1, good agreement occurs between both sets of data. Again, there is a si gnificant difference between the maximum tensile strains occurring in the two plots. However, th e reason for this occurri ng would be the same as the case for the 1 data. Experimental Process Induced Residual Stresses The displacement data obtained from the expe rimental results was used in the model to determine the residual stress contour fields. When temperature fields are applied to a material, strains develop due to the coefficient of thermal expansion as the change in temperature increases. However, those deformations are not due to an applied force or stress, so they have to be modeled as virtual loads. The phase shifting analysis provides strain and displacement information at every pixel location within the area of interest. Theoretica lly then, one could use that full-field data and input the full-field displacement data into a finite element model to obtain the resulting full-field stress data. However, due to the large numbe r of data points from the experiment and the number of nodes in the model, it was decided to input the displacements through the periodic boundary conditions (PBC). 62

PAGE 63

When Karkkainen developed the ABAQUS mode l, a dummy node was created to control the PBCs. Therefore, in order to take advantag e of the PBCs, the overall displacements for the U and V fields needed to be determined from th e experimental data. Because the displacements constantly vary throughout the RVE, the average overall displacements were used for each direction. Using this method, the displacements that were used for the model in the U and V directions were -6.13x10-3 mm and -6.06x10-3 mm respectively. Those values were achieved by taking the average displacement occurring in eac h direction along the length of the RVE. The stress contours are shown in Figure 5-9 (a) and (b) for 1 and 3 respectively. Also, the overall trends for the stresses are plotted in Figure 5-10 and Figure 5-11 The stresses follow the same trends that have been shown for the stra ins. Peak tension valu es are occurring over the fiber bundles and the peak compression values are within the resin rich pockets. The peak tensile stress was approximately 5 MPa and that was located in th e very center of the RVE over the fiber bundle where the resi n content is lowest. The p eak compressive stress was approximately -24 MPa and was located at al ternating resin zones throughout the RVE. Conclusions Finite element analysis was used for two asp ects within this research. The cooling cycle was simulated to obtain a prediction for the strain pattern and values th at should be expected from the experimental work. The strain patter n was described by relating it to the geometry of the RVE and the different CTE values of the two di fferent materials. The strains fell within the approximate range of a maximum tensile strain of 1820 and a maximum compressive strain of -3700 By comparing the contour plots obtained thr ough the experimental a nd analytical methods, reasonable verification of the method develope d was completed. Because other experimental techniques do not exist that are capable of meas uring the full field strain s of woven composites, 63

PAGE 64

data was not available to perform experimental validation. Therefore, performing the simulation within ABAQUS was the best method available to verify the experimental results. Even though the results cannot be exact between the two mode ls due to geometry and chemical shrinkage issues, very good agreement occurred. Therefore, the results that were obtained through the improved CRM procedure were those expected. After using the deformations obtained from th e experimental data, the residual stresses along the surface for the RVE were calculated. A lthough the values are relatively low, those are the stresses within the re sin on the surface. According to th e data sheet provided by the prepreg manufacturer, the tensile strength of the resin is 82.7 MPa. Although the actual stresses that developed were well below that value, they are still important to unde rstand since they do limit the usable strength remaining in the structure. 64

PAGE 65

Table 5-1. Material properties used in ABAQUS model for the plain weave composite RVE E1 (Gpa) E2 (Gpa) G12 (Gpa) 12 1 ( /C ) 2 ( /C) Weave (AS4/3501-6) at 120C 138.00 5.64 2.48 0.30 0.59 26.90 at 20 C 138.00 9.35 5.30 0.30 0.45 23.03 Matrix (3501-6) at 120C 3.50 3.50 1.30 0.35 49.00 49.00 at 20 C 3.50 3.50 1.30 0.35 41.30 41.30 Figure 5-1. Simulated process induced re sidual strains in the 1-direction ( 1) resulting from cooling the RVE from Tref to room temperature and introducing it to atmospheric pressure. Region A is the resin rich zone and Region B has a lower resin content. Region B Region A 65

PAGE 66

Figure 5-2. The c ontour plot of 1 after the simulated cooling cycle to room temperature. Figure 5-3. Plot of 1 along the length of the midline of the warp of the RVE in the 1-direction. The data is taken on the surface of the matr ix along the line shown in the subset figures. 66

PAGE 67

Figure 5-4. FEA 1 results for the RVE from the simulated curing cycle obtained Figure 5-5. Experimental phase shifted 1 resulting from the cooling cycle 67

PAGE 68

Figure 5-6. Plot of the residual strain, 1, along the resin zone in the 3-direction. 68

PAGE 69

Figure 5-7. FEA 3 results for the RVE from the simulated curing cycle Figure 5-8. Experimental phase shifted 3 resulting from the cooling cycle 69

PAGE 70

Figure 5-9. Stress contours for the 1and 3-dire ctions, (a) and (b) respec tively, obtained using experimental displacement data 70

PAGE 71

Figure 5-10. The stresses are pl otted along the fiber directions in along two different bundles the directions of the plots are shown in the subset figure. 71

PAGE 72

72 Figure 5-11. The stresses along th e resin zones are shown along and transverse to the resin zone. Figures (a) and (c) are plotted along line (a) in the subset, and figures (b) and (d) are plotted along line (b) in the subset.

PAGE 73

CHAPTER 6 CONCLUSIONS AND FUTURE WORK Conclusions Residual strains are usually inevitable to occur in composite materials due to the high temperature manufacturing proce ss required and chemical shrinka ge. Because of this, it is critical that those strains, and ultimately the stress es, can be determined so that the usable life of the structure can be assessed. Many methods exist to measure those strains in most materials, and even in unidirectional composites. However, only a handful of methods have been used to measure the strains developing in woven composites, and none of those have been able to obtain a full-field contour map of the st rains. Cure reference method wa s the solution to that problem, because with the combination of moir interferometry, strain measurements can be obtained throughout an entire field of view. Then by processing the resulting fringe images with phase shifting, a complete two dimensional map of the re sidual strains can be obtained for the U and V directions. The original procedure for producing the auto clave tool for CRM was very time intensive and provided poor repeatability when attempts we re made to reproduce that method. Therefore, this work became focused on modifying and im proving that method. By creating a master autoclave tool, and then pre-cu ring a thin layer of epoxy onto th e prepreg before cure, very repeatable gratings were obtained that easily separated from the autoclave tool. The resulting data showed a wide range of stra in values occurring in a very small region of the weave geometry. Because of the complex geometry exhibited by woven composites, it was expected to obtain both tensile and compressive strains throughout the field of interest. Those did occur, and the strains th roughout the RVE fell between -3000 and 2500 The maximum tensile strains were found to occur over the fi ber bundles and the high compressive regions were 73

PAGE 74

the resin rich zones between the bundles. The experimental displacemen t data was then input into the ABAQUS model to calculate the residual stresses that were developing because of the cooling process. It was found that the stre sses were bound between approximately -24 MPa and +5 MPa. Although the strength of the resin used in the prepreg is 82.7 MPa that level of stress is not actually available to the end user because of the residual stresses that developed from the curing cycle, this is especially try for cryogenic uses. Even though the values of stresses were small with respect to the streng th, if not correctly accounted for, failure could occur before it would be expected because those re sidual stresses were neglected. The finite element model was also used to vali date and verify the data obtained through the improved CRM. After comparing the strain plots obtained from the two methods it was apparent that the improved CRM method was providing accura te data for the residual strains developing in the specimens. Future Work Although the final CRM grating production procedure was significantly improved over the original process, both by time and ease of production, some improvements could still be made. The final method definitely showed the best repeatability as far as achieving smooth grating surfaces that were easily separable from the autocl ave tool. However, there are still frequent occurrences in which sufficient epoxy is not co vering the fiber bundles cl osest to the surface. Whenever that occurs, the fringes are highly di scontinuous so phase shif ting analysis cannot be completed in that region. The method to apply the last epoxy layer s hould be slightly modified to resolve that issue. For this research, a field of view was chos en on the specimen, and the full-field strain information was calculated throughout that entire region. As was seen from each of the contour maps created, perfectly uniform strains were not measured throughout the entire viewing region 74

PAGE 75

75 on the specimens. Therefore, although the RVE is a repeating unit, there do appear to be small discrepancies between each of those elements. It would be desirable to obtain statistical information throughout a region of interest to dete rmine the variability between different RVEs. The variability that could produ ce different strain values would be caused by non-uniform resin distributions throughout the prepre g. Having regions either rich or low in resin would produce this type of inconsiste ncy in the results. For the finite element analysis to calcul ate the experimental residual stresses, the deformations were input as periodic boundary conditions accounting for the total average displacements that occurred for the entire RV E. Although that was a reasonable approximation to predict the stress levels occurring throughout the region, it was unable to capture the full displacement non-uniformity that oc curs throughout the region. Th erefore, future work should include developing a method to i nput the experimental data dir ectly from the phase shifting analysis into the ABAQUS model. Once that is accomplished, a more accurate stress field can be obtained through the fi nite element analysis.

PAGE 76

APPENDIX CRM IMAGES AND PHASE SHIFTED RESULTS Figure A-1. Silicone Rubber and 3501-6 Resin, U Field Figure A-2. 1 phase shifted results for si licone rubber and 3501-6 resin 76

PAGE 77

77 Figure A-3. Silicone rubber and 3501-6 resin, V Field Figure A-4. 2 phase shifted results for si licone rubber and 3501-6 resin

PAGE 78

Figure A-5. High temperature epoxy spread and set on prepreg, U Field Figure A-6. 1 phase shifted results for High temper ature epoxy spread and set on prepreg 78

PAGE 79

79 Figure A-7. High temperature epoxy spread and set on prepreg, V Field Figure A-8. 2 phase shifted results for High temper ature epoxy spread and set on prepreg

PAGE 80

LIST OF REFERENCES 1. Gibson, R.F., Principles of Compos ite Material Mechanics McGraw-Hill, New York, New York (1994). 2. Post, D., Han, B. and Ifju, P.G., High Sensitivity Moir: Experimental Analysis for Mechanics and Materials Springer-Verlag, New York (1994). 3. Hahn, H.T., and Pagano, N.J., Curing Stresses in Composite Laminates, J. Composite Materials, 9, 91 (1975). 4. Mathar, J., Determination of Initial St resses by Measuring the Deformation Around Drilled Holes, Trans. ASME, 56, 249-254 (1934). 5. Rendler, N.J., and Vidness, J., Hole-drilling Strain-gage Me thod of Measuring Residual Stresses, J. Experimental Mechanics, 6, 577-586 (1966). 6. Schajer, G.S. and Yang, L., Residual-stress Measurement in Orthotropic Materials Using the Hole-drilling Method, J. Experimental Mechanics, 34, 324-333 (1994). 7. Sicot, O., Gong, X.L., Cherouat, A. and L u, J., Determination of Residual Stress in Composite Laminates Using the Incrementa l Hole-drilling Method, J. Composite Materials, 37 831-844 (2003). 8. Wu, Z., Lu, J. and Han, B., Study of Re sidual Stress Distribution by a Combined Method of Moir Interferometry and Incremental Hole Drilling, Part I: Theory, J. Applied Mechanics, 65, 837-843 (1998). 9. Daz, F.V., Kaufmann, G.H and Galizzi, G.E., Determination of Residual stresses Using Hole Drilling and Digital Speckle Pattern Interferometry with Automated Data Analysis, Optics and Lasers in Engineering, 33 39-48 (2000). 10. Zhang, J. and Chong, T.C., Fiber Electronic Speckle Pattern Interferometry and its Applications in residual Stress Measurements, Applied Optics, 37, 6707-6715 (1998). 11. Hahn, H.T., Residual Stresses in Polymer Matrix Composite Laminates, Composite Mat., 17, 265-277 (1976). 12. Kim, R.Y. and Hahn, H.T., Effect of Cu ring Stresses on the First Ply-failure in Composite Laminates, J. Composite Mat., 13, 2-16 (1979). 13. Predecki, P. and Barrett, C.S., Stress M easurement in Graphite/Epoxy Composites by XRay Diffraction from Fillers, J. Composite Materials, 13, 61-71 (1979). 14. Benedikt, B., Kumosa, M., Predecki, P.K., Kumosa, L., Castelli, M.G. and Sutter, J.K., An Analysis of Residual Thermal Stresse s in a Unidirectional Graphite/PMR-15 80

PAGE 81

Composite Based on X-ray Diffraction Measur ements, Comp. Science and Technology, 61, 1977-1944 (2001)., 15. Benedikt, B., Rupnowski, P., Kumosa, L., Sutter, J.K. and Predecki, P.K., Determination of Interlaminar Residua l Thermal Stresses in a Woven 8HS Graphite/PMR-15 Composite Using X-Ray Di ffraction Measurements, Mech. Of Adv. Mat. And Struct, 9, 375-394 (2002). 16. Daniel, I.M., and Liber, T., Lamination Resi dual Stresses in Fiber Composites, IITRI Rep. D6073-1, for NASA-Lewis Resear ch Center, CASA CR-134826 (1975). 17. Daniel, I.M., Liber, T. and Chamis, C.C., Measurement of Residual Strains in Boron/Epoxy and Glass/Epoxy Laminates, Composite Reliability, ASTM STP 580, Amer. Soc. Test. And Mat., 340 (1975). 18. Daniel, I.M., and Liber, T., Effect of La minate Construction on Residual Stresses in Graphite/Polymide Composites, J. Experimental Mechanics, 17, 21-25 (1977). 19. Zewi, I.G., Daniel, I.M. and Gotro, J.T., Residual Stresses and Warpage in WovenGlass/Epoxy Laminates, J. Experimental Mechanics, 27, 44-50 (1987). 20. Lawrence, C.M., Nelson, D.V, Spingarn J.R. and Bennett, T.E., Measurement of Process-Induced Strains in Composite Mate rials Using Embedded Fiber Optic Sensors, Proc. Of SPIE, 2718, p 60-68 (1996). 21. Lawrence, C.M., Nelson D.V., Bennett, T.E. and Spingarn, J.R ., An Embedded Fiber Optic Sensor Method for Determining Residua l Stresses in Fiber-reinforced Composite Materials, J. Intell. Mater. Syst. Struct. 9, 788-799 (1998). 22. Anastasi, R.F., and Lopatin, C., Application of a Fiber Optic Distri buted Strain Sensor System to Woven E-Glass Composite, TM-2001-211051, NASA, 2001. 23. Mnson, J.E., and Seferis, J.C., Process Simulated Laminate (PSL) Characterizing Internal Stress in Advanced Composite Materials, J. Composite Materials, 26, 405-431 (1992). 24. Ifju, P.G., Niu, X., Kilday, B.C., Liu, S.C. and Ettinger, S.M., Residual Strain Measurement in Composites Using the Cure-referencing Method, J. Experimental Mechanics, 40, 22-30 (2000). 25. Scida, D., Aboura, Z., Benzeggagh, M.L. and Bocherens, E., Prediction of the Elastic Behavior of Hybrid and Non-hybrid W oven Composites, Comp. Sci. and Tech, 57, 1727-1740 (1998). 26. Naik, R.A., TEXCAD Textile Composite Analysis for Design, NASA Contractor Report 4639 (1994). 81

PAGE 82

82 27. Huang, X., Gillespie, Jr., J.W. and Bogetti, T., Process Induced Stress for Woven Fabric Thick Section Composite St ructures, Comp. Struct., 49, 303-312 (2000). 28. White., S.R. and Kim, Y.K., Process-induced Residual Stress Analysis of AS4/3501-6 Composite Material, Mech. Of Comp. Mat. And Struct., 5 153-186 (1998). 29. Whitcomb, J.D., Three-dimensional Stress Analysis of Plain Weave Composites, Composite Materials Fatigue and Fractur e (Third Volume), ASTM STP 1110, 417-438 (1991). 30. Strickland, N.M., Yin, W. and Ifju, P.G., R esidual Strain Measurement Analysis of Woven Composites Using Phase Shifting, Pr oceedings of the 2007 SEM Conference on Experimental Mechanics, Springf ield, MA, USA, June 4-6, 2007. 31. Karkkainen, R., Stress Gradient Failure Th eory for Textile Structural Composites, University of Florida PhD Dissertation (2006). 32. Carvelli V. and Poggi C., A Homogenization Procedure for the Numerical Analysis of Woven Fabric Composites, Composites Part A: Applied Science and Manufacturing, 32, 1425-1432 (2001). 33. Speriatu, L., Temperature Dependent Mechan ical Properties of Co mposite Materials and Uncertainties in Experimental Measurements, University of Florida PhD Dissertation (2005). 34. Berman, J.B., and White, S.R., Theoretical Modeling of Residual and Transformational Stresses in SMA Composites, Smart Material Structures, 5, 731-743 (1996).

PAGE 83

BIOGRAPHICAL SKETCH Nancy M Strickland was born April 12, 1983, in Pensacola, FL. She spent a significant portion of her life living in Gulf Shores, AL, be fore moving to Jacksonville, FL, in 1995 where she later attended Allen D. Nease High School. After gradua ting in the top five of her high school class, she enrolled at the University of Florida in August 2001 where she began her studies in mechanical and aer ospace engineering. Although sh e had originally intended on pursuing medical school, her love for the math and sciences drove her to e ngineering early in her college studies. During her underg raduate career, she was a member of the Pride of the Sunshine Marching Band for three years, as well as being very active with Tau Beta Pi and ASME. In December 2005 she graduated cum laude with bachelors degrees in mechanical and aerospace engineering. The following spring sh e began pursuing her masters degree in aerospace engineering under the advi sement of Dr. Peter Ifju in th e Experimental Stress Analysis Lab (ESA Lab). During her time working in the ESA Lab, her research was focused on plain weave composites and process induced stra ins until she graduated in December 2007 83