<%BANNER%>

Design and Implementation of a 6 DOF Parallel Manipulator with Passive Force Control

Permanent Link: http://ufdc.ufl.edu/UFE0011000/00001

Material Information

Title: Design and Implementation of a 6 DOF Parallel Manipulator with Passive Force Control
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011000:00001

Permanent Link: http://ufdc.ufl.edu/UFE0011000/00001

Material Information

Title: Design and Implementation of a 6 DOF Parallel Manipulator with Passive Force Control
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011000:00001


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 E20101130_AAAAAC INGEST_TIME 2010-11-30T05:27:08Z PACKAGE UFE0011000_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 1053954 DFID F20101130_AAABMF ORIGIN DEPOSITOR PATH zhang_b_Page_013.tif GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
9c239f6a32f9770c202434d745771d6a
SHA-1
5043caab373b4eb5cefc65dd55840ffc7e3ba081
25271604 F20101130_AAABMG zhang_b_Page_091.tif
58d0b941f21c895bf8205e6f847ae3b9
ec4635977d96ef7844aa020c25bf2580a2dc3af1
6394 F20101130_AAABMH zhang_b_Page_078thm.jpg
9fa69a57743528ddb8b08cab2d0cf62e
7ffa29ea66d57ae73e2f26ddc7ce31e6c9b528ab
6612 F20101130_AAABMI zhang_b_Page_019thm.jpg
a36005b3d132e22313e587e9df78f98d
9e79bf8529648d089c17bb719d10908df0a4c3da
5344 F20101130_AAABMJ zhang_b_Page_113thm.jpg
749dd48e18e5b938344136a4d099543e
5e69d1a30962b40793deb5be0944c3c36f915d06
93876 F20101130_AAABMK zhang_b_Page_028.jp2
bc3d429b0c5565bcfb79b640c32206ed
06649bc06c0b3a0a6cfda88cd5ec88f0aa97aa37
78834 F20101130_AAABML zhang_b_Page_128.jpg
2dc58f0398c355bbe35d44201a242992
a726437e6ca779837fde77676bc22a9d07b07f03
F20101130_AAABNA zhang_b_Page_061.tif
344465b2143a6fee06d6b18a0de532ff
223bdd3f66e5abd1a0322c7d60bb3093834c0025
23219 F20101130_AAABMM zhang_b_Page_001.jpg
d580a49579be90a822ecc7e03cb49a96
fffd656206582deadbb98373eca16c4d5fe5fd43
F20101130_AAABNB zhang_b_Page_081.tif
96aa9d484acc2b98fb3817fa3a2d4775
36c081f66d8f5b91d1f0264b10a5e4baba2bb4e1
6023 F20101130_AAABMN zhang_b_Page_039thm.jpg
b04cea7e4c20f0c7c0ba48bcf059bd1b
f8d0d5c7fe9dc44067be75518f9aeaa95f609544
18344 F20101130_AAABNC zhang_b_Page_047.QC.jpg
5b4bb7db9f47417621baf555d2033ae9
b3dd29fbc4554926bd4b27bd6af337c22497d622
612286 F20101130_AAABMO zhang_b_Page_095.jp2
c772d707017ab13599141de67a31c976
672beb2a3a1799350eceea63a7d9e29e4e7ff288
1031156 F20101130_AAABND zhang_b_Page_082.jp2
26ee27399a4e1b7f9cf134c3d3441967
f47a8af52b30658d7f51afec6e7f0f23cfeb4470
104103 F20101130_AAABMP zhang_b_Page_014.jp2
f97f0c25aec80cbe26d3594da0006cf9
77bd1cc5ee3372a15ec6defd43279e9f531a576d
F20101130_AAABNE zhang_b_Page_018.tif
68395cdbe1ad85c86bef6f91ca8e4600
282b024ea69b46f177a6fa9177d48b1e5dc70be5
109829 F20101130_AAABMQ zhang_b_Page_018.jp2
976fabe9737c9103f01fb906ce43aa78
d444581b120503923447151b3c657f7c9a960a4e
28328 F20101130_AAABNF zhang_b_Page_012.jpg
de6d174b0bd111faaa270a5fbfc629de
20b492111e44ee21daf46418e2aff45d2b8c47b1
578475 F20101130_AAABMR zhang_b_Page_085.jp2
2407097d099632804c07d60626e4a28b
298d942a10023ac14ad2a9b90b2c35997a269294
73081 F20101130_AAABNG zhang_b_Page_024.jpg
b60a1eb70ec16cab441310fccf641122
b941a791d19a9543d1c6b5b09ec7f2807abb2419
F20101130_AAABMS zhang_b_Page_109.tif
9a2ba81e8a96399095a8218483706901
cdde1e277cc3936ee4e66cffadb666fa9d58ca86
15932 F20101130_AAABNH zhang_b_Page_044.QC.jpg
9e23d454f020a89fa0b6ba5aad79931e
e27b2deadb9b2b3bf1a4462cd8a42f1e73231deb
16658 F20101130_AAABMT zhang_b_Page_029.QC.jpg
c6ce32a86a193975d531145f2166e6c3
b4f108e523bbaa6285c2a74bc5f399f255bc662b
21086 F20101130_AAABNI zhang_b_Page_097.QC.jpg
e0c05ff8ce11c4022712667bbb57776d
ffe6e85d1ef73496fd1783059300cc6e94ffb9a2
5012 F20101130_AAABMU zhang_b_Page_070thm.jpg
ca2cb9fafbe04c0c31cdc26ff98f19bf
c613ebf7ec9796dde87f0ba76808a488a59c010c
11873 F20101130_AAABNJ zhang_b_Page_067.QC.jpg
2974f1b779cf0f4899a88a6d2ee9897e
bc1af2526ec59aef95242ee8e31d6170397ee427
56273 F20101130_AAABMV zhang_b_Page_111.jpg
ce81ddfabb533357a74a92a4ecf163a2
330d4256470baea307b98f8d2999cfef383c8c2e
22751 F20101130_AAABNK zhang_b_Page_014.QC.jpg
33f6f3e3d8c877fd4a39ffbdf94c1d17
61fab9853ed40457812a939fa39674a05c254db3
24733 F20101130_AAABMW zhang_b_Page_127.QC.jpg
e087b34b7420baed17d1b60f63c7d4cc
646ec5a0b97289df3e69ad8fe7c2efe0f60dcc1f
56601 F20101130_AAABNL zhang_b_Page_055.jpg
3a61d054573d9c5e036650690108b85a
c8fd93f13e2b92067264b703f9fda8db4e430be0
15271 F20101130_AAABMX zhang_b_Page_084.QC.jpg
9e68488031b753555aade81d2e08bdd5
c4a982f18c6a6064f567d11f9874d03395b04eaa
63302 F20101130_AAABOA zhang_b_Page_063.jpg
e4def89fdfa76058e995f6136c0a9128
ca0044080d2e490e2551a0d4e189a868e315fcd9
58715 F20101130_AAABNM zhang_b_Page_040.jpg
87a05c66d7aae27c7149fc4283cca420
1b2a7253848c1b04aaeda69fcb461f9d971947b9
16885 F20101130_AAABMY zhang_b_Page_096.QC.jpg
3dc1a66a46b6e32aac48be18f8e52d08
1a7450660c13d8d24d680715061f9668b4ac1b17
6772 F20101130_AAABOB zhang_b_Page_022.QC.jpg
f62340363d8cfbf9ccb6557f25df9353
8817bbbbcdb186fdeed75868a33780f75bf60fb6
25265604 F20101130_AAABMZ zhang_b_Page_121.tif
65e5f2272a0e4d6ed8c4bda052f23cc9
54b01bc254fa0360bc1e41ca30424ea4ccaeb52d
F20101130_AAABOC zhang_b_Page_092.tif
4d3df00808ca9ecc119958ba2c7c700f
4452cb0e162af7d6eaa8d8d2ed1c7efc9ee92fcb
55772 F20101130_AAABNN zhang_b_Page_096.jpg
6acb10079524bca3d72e2dacd2ee00f8
3e0de7bf08c5c1169fc2fcba439e4ec34a75d2e2
14734 F20101130_AAABOD zhang_b_Page_021.QC.jpg
340772083fdcc5d5c7055ea2bb0c97b2
25a4983ec016edc334c9ebbcebb34d94b74fcb64
5892 F20101130_AAABNO zhang_b_Page_103thm.jpg
75af267ef5af889493904a9cdbcb352c
1c46eec919a63cc7690337353ba54d46c6112372
9573 F20101130_AAABOE zhang_b_Page_109.QC.jpg
5444209daf82fb72fd26535a62ce2de7
ca559bf628b486a0c3f162628279c9e7a814679e
6241 F20101130_AAABNP zhang_b_Page_082thm.jpg
a77ddfd73089c32968a1fb1f18e2ff2a
83f8f6288c65eb4bb6c5657a2c44c82d3e694e83
82160 F20101130_AAABOF zhang_b_Page_100.jp2
e2e21716f1da10db4e6be51f1189e98f
e9e3efd17f71f570139870ff98ee3a4e14a2aab1
5814 F20101130_AAABNQ zhang_b_Page_037thm.jpg
b73b2a95b34800439793335d9a84eb35
629baada7290021d12d1d0172cea389ac5414c79
11606 F20101130_AAABOG zhang_b_Page_005.QC.jpg
35571fd27df03953db2bba7829d4f14b
bee9fbfc147de017f7c07b8a2013afe0d76979aa
13840 F20101130_AAABNR zhang_b_Page_031.QC.jpg
7478dfc55cae3d5228ebb424bc196d74
afbd6595e44165eb0f31fb7914e3d20f7d2930c0
87265 F20101130_AAABOH zhang_b_Page_127.jpg
f0ee2c61a4c8682235a015ac987b4209
dcc6eb22c4dd8076f9bcfdf57273ec138210b325
F20101130_AAABNS zhang_b_Page_034.tif
93067cbc756a5cda7ae33967bab84363
ffc3e045dc4b0c4d14cccf3a2d9884e5753c90ea
75046 F20101130_AAABOI zhang_b_Page_112.jpg
3eaf30518038e043fed9286b953e9b66
0ae8099b28b4127ae10fda20a222e478f4588523
36638 F20101130_AAABNT zhang_b_Page_086.jpg
2a631d2df7dae7f0069a483e491fa88f
c88dddf59f339768e504f08a55f15381fb40ca85
F20101130_AAABOJ zhang_b_Page_011.tif
92d6587a0a15cf17f7fb0596edc06b4f
825bd70a7e76937e5045101d14c3169180160a82
14740 F20101130_AAABNU zhang_b_Page_108.QC.jpg
6f9e3904e4e09f8f4dc12aa5fea10ecc
db95d2fdc12117a73e3b4600cc8071846cac13c8
10632 F20101130_AAABOK zhang_b_Page_125.QC.jpg
9eee78d5dacab48b7cadd0f8cba92c1d
ce1e18c0d258e7b17eaef8a4f39f00dade0a938e
F20101130_AAABNV zhang_b_Page_120.tif
9fba6d03c965ccc60bc2e0cedb058efa
c3fe10c8f9ff13161ebff3cb036e39b07bafc10a
F20101130_AAABOL zhang_b_Page_112.tif
13b9c09d62f68ac06fbf55ae64f2e5ec
2dc975f598df1d06de4a0ce537a829ac68dfe700
70623 F20101130_AAABNW zhang_b_Page_075.jp2
1368bb974bc6dbe7860516967783acb3
6aba14331ec3bb54841fde3d7e3acd614a92857a
64980 F20101130_AAABOM zhang_b_Page_105.jpg
ade022feb88286aec9e19c6cce47c5e7
cebd0af3af5db612929ee678918e873fa43ca106
F20101130_AAABNX zhang_b_Page_075.tif
9e20b69194bda28a9e6d1cc4106d3174
ce4c8ded71958c91cfc5fadf027d8dff31824a40
F20101130_AAABPA zhang_b_Page_030.tif
00d65249da88be17260157e2ca359c2d
7ed81ccc91aa44ebbdc05f70c79c0291e47aa2a9
83633 F20101130_AAABON zhang_b_Page_115.jp2
1e1162a2f8fa5a24257272d8e0bdcf6c
587ec291bc118c80613857df3b0120eedcd6e754
33932 F20101130_AAABNY zhang_b_Page_095.jpg
482147855f2f7070c7e5e1a963618c08
8ff760869427cd41e988ad7671b0d2a9553f8e22
20614 F20101130_AAABPB zhang_b_Page_022.jpg
4a8212d42f9d0ce32885f6f393089701
91d36acf10418bb24e77ed339074a5c88f00ce3e
5681 F20101130_AAABNZ zhang_b_Page_114thm.jpg
cdc8ad5d17b2a162a71380189f338fa7
71216b9d9b1786b4c03262c2c3805ab719dc777b
20406 F20101130_AAABPC zhang_b_Page_020.QC.jpg
09bb7345084908bbcece51c946ea201e
d762b07977fcc045383d9c53c6a63ec19d499356
F20101130_AAABOO zhang_b_Page_062.tif
b821699d913d76c817a0e4c7d1c47d83
6f2463b63ed937fd44f626bdc45fa9eb5c54770c
56858 F20101130_AAABPD zhang_b_Page_115.jpg
c0d86012a1d9b34befb1e0c7ac4fb799
57f178e81003a66c6409d7730ca80677319acff1
17590 F20101130_AAABOP zhang_b_Page_062.QC.jpg
36687ca5cb20dfcf8e477ee6aab0e4ed
0c11bd0bf6a5fb119c73e4213bf2ca4811505eb7
F20101130_AAABPE zhang_b_Page_047.tif
394f90e4d75b45779f2f8204acd50190
721dbf4611be8c0bf61b52971a89a0ee58da173f
74702 F20101130_AAABOQ zhang_b_Page_019.jpg
0d9c151ab3f9b2256f220b2e4c6a870e
17a08379ba12b6ced492f398c93062b1afe202d0
F20101130_AAABPF zhang_b_Page_010.tif
cf40f072c7f5d7d9f2b5dcb4e000f5af
448621fe2c3c864979aba3fe79da7f1a338406f6
5130 F20101130_AAABOR zhang_b_Page_056thm.jpg
e5d393ef1fdb2951834ac9c2e0594855
8b1f0b6cedaad6dc07316a4a018a19cce0061370
6064 F20101130_AAABPG zhang_b_Page_050thm.jpg
d5f28711206f6fc5de7307591d9326a1
a01145a58b6758d679a8bcc615db1b88191d618e
94277 F20101130_AAABOS zhang_b_Page_023.jp2
c5dcc62a79029974b1712263d4082cde
dc781543043f6835f05905c9e97accb038da0b2f
46859 F20101130_AAABPH zhang_b_Page_061.jpg
fa8c69aebaa98deca79b2cb7a26f5f57
19780ec1c4bffb8f3052c612a9d05be92a60a766
F20101130_AAABOT zhang_b_Page_103.tif
d7d0cfd3a8583f8acd56a07748f50269
3170bff3cd21dd648da55096565de96fd30b9183
23380 F20101130_AAABPI zhang_b_Page_099.QC.jpg
6dfc6e36218a1ef8accdbd8d2def1a27
e8674d7dd1ad970fb31843bb7998725be7a85efc
388860 F20101130_AAABOU zhang_b_Page_007.jp2
b568d996384155f7798a7403720fa31a
f46bfb4a45b945bdb0c7fae854aecd595c174565
78602 F20101130_AAABPJ zhang_b_Page_047.jp2
06f47b827b9009abdf105bdca7500606
95d8720f73e383d3706ab2332ab58796881b19b0
56333 F20101130_AAABOV zhang_b_Page_053.jpg
16634bf9e29bd19b9167f9545f3d1586
9021e16d5f407fbe9f3eb6c5d1a42def71ca583e
62248 F20101130_AAABPK zhang_b_Page_089.jp2
976f97ffcf0a9b0e8ea2d42d3dd77dfb
a18fcdb5865fe3b5bb772e370f5c645f968a842f
55620 F20101130_AAABOW zhang_b_Page_107.jpg
82ca62c98f3042846e1cb57c41a410cb
207d1acec0b5c41b5b262396a596032ea1270e1f
55132 F20101130_AAABPL zhang_b_Page_043.jpg
6c4599e0eb27c7bf91cc2b8cb8cb9d3f
9cb6981f854f76ba4adb58ea67d9a815bd49c22b
55928 F20101130_AAABOX zhang_b_Page_011.jpg
22767139014bdfbdc4d6dc8c3bb83c78
dc5a23f76457b180ecdca05a9ef58fcaaa63f14b
6445 F20101130_AAABQA zhang_b_Page_097thm.jpg
9c8538d104dc4bdfe2461d7204205124
7e638fa32c75d6964cb954fbbc7e0cbe52575850
21289 F20101130_AAABPM zhang_b_Page_026.QC.jpg
b486c574d6a20fb812b40cdc170dee6e
5abf85a0093ee125eb22ada9b01310030c328411
31841 F20101130_AAABOY zhang_b_Page_124.jpg
a2718d2d0b8889964066504677a6939c
402f01efd92a7f784325016838d32298056654b7
F20101130_AAABQB zhang_b_Page_006.tif
ae35da7e88f6063f7b4c593ec45c5996
39753dfedb3f4e78dcf55448e6361020cfecd897
F20101130_AAABPN zhang_b_Page_031.tif
543288ea54318e2c49521249ea25ad3a
926ac3400c3cd76e7514e0a93deb5dcbc5fdf2b1
F20101130_AAABOZ zhang_b_Page_025.tif
059a7ba9d6ecea7f26b301a85f20e244
dd97f9772c2cbffac32a27bf15237d204dd633ba
17832 F20101130_AAABQC zhang_b_Page_055.QC.jpg
94f009d45698654476c079bbfd05abb8
2050f59c025d518d4c41e72059dab95f7d6c37d4
F20101130_AAABPO zhang_b_Page_021.tif
9f43deb040a99a39f11188543f0eeba0
44a26f642da8439636a60bf43209f2f6845246d2
F20101130_AAABQD zhang_b_Page_040.tif
961049d6bda7602b28cbb84c1527e985
271319f98306b8dffd12a1fa317418980b32190a
F20101130_AAABQE zhang_b_Page_027.tif
86802d2cfdd17db5deedde6747b5086f
4151ad8cf061a0f0594249ad78735001d2e95a2a
48823 F20101130_AAABPP zhang_b_Page_102.jpg
da1add3cddd6fb65372f7cfdac9fe118
12732d544d11f96d6b8383bfa28a2d78c8b9575d
F20101130_AAABQF zhang_b_Page_119.tif
e7094a065b9173d21a075df7df8ab878
9667f1ba9d202d0c7cf23fc6b1db2254fe1e1640
3206 F20101130_AAABPQ zhang_b_Page_002.QC.jpg
8748057becbeefccf5c484feb72ae25b
2c99f0e82fe783f542e303c7b2bf142db2bc3fce
20050 F20101130_AAABQG zhang_b_Page_105.QC.jpg
5f1af868da140884acd1f2835cd44ab3
68cdc35177304a5eefb9957d80c89237f3419410
23723 F20101130_AAABPR zhang_b_Page_024.QC.jpg
27ceb74933d18bd29eab8eca8dff9b1a
a7d3eda3ba8b16c69ac4e580b4c01aadd68ae916
71319 F20101130_AAABQH zhang_b_Page_110.jp2
6158499290b6f32ac7e4b4b8c74746d0
64c631e96fae4edd543f3263b91ea259fc56a817
F20101130_AAABPS zhang_b_Page_087.tif
aeb971ae5e509d9a43e5767995759ae3
a88bbefa2522be6ba77f66a577a5a8f3494556ed
55946 F20101130_AAABQI zhang_b_Page_048.jpg
823520fda61ec59d962f8077083fcd10
7a96625cea9fd61acabed8748617c94207e025e8
5463 F20101130_AAABPT zhang_b_Page_111thm.jpg
b48bf71694a1fa3d37311efc62488720
ff1d679b3169834066222594189fce6d69c0fc24
20964 F20101130_AAABQJ zhang_b_Page_100.QC.jpg
7a0befdb13cb0fae3e2d6db86f7f56a1
882153e1ad6e31c54479a272054ba57097420429
3133 F20101130_AAABPU zhang_b_Page_045thm.jpg
d438b6a102bcb13546bbffc3b917b821
f8f0d2f6a309671e912adc3f65752e09a7c9955b
3425 F20101130_AAABQK zhang_b_Page_091thm.jpg
17224f7e96f0d39fb975e5946570d446
aa0f8c6e35535d262928cb9ebb8b940af2226b41
F20101130_AAABPV zhang_b_Page_128.tif
8301b8afac479861357b245a57ddc989
e5bff3e28e3343e8654e48942acc3996a60c5f1b
10125 F20101130_AAABQL zhang_b_Page_095.QC.jpg
0f845274a42e78ca1eb3b851eec44d53
88ff2d1c4b22c830d1876cb042c27c963bf7b10d
F20101130_AAABPW zhang_b_Page_098.tif
36e999c0e30f21081386eb89a3ff74d8
6e496fa9e42efb79ceb41902c1cb8a8f11ee466d
F20101130_AAABRA zhang_b_Page_130.tif
eec0a3decc044d33900681e2a32ffe5f
1053a2e59c12c1253d60fb977265ce61d8f455ab
734457 F20101130_AAABQM zhang_b_Page_015.jp2
c4492afbc940d58f25b06719b9213e61
b312d11a1ee9d2beeaca88c8cc82f6001e814d44
29324 F20101130_AAABPX zhang_b_Page_122.jpg
395b151095d5af8ce852001d3b997b80
2e94a0d1174ca76fdb3b44a6d315000a4b7a9879
4701 F20101130_AAABRB zhang_b_Page_004thm.jpg
dcbba04a4cf681c0ca33f662e1186d81
ea213d0502c74ecc2983d26e7896f4dc3efcadd0
F20101130_AAABQN zhang_b_Page_123.tif
cd99550bf575ca07e095323139bf87c2
62706a35aa4b89e002460ed53df903e537182bd8
127676 F20101130_AAABPY zhang_b_Page_010.jp2
c012a42f292f31bbfc5e55b8a69ccfa1
327779fcf75ac899b072cd9d9d1f176338e2765c
5820 F20101130_AAABRC zhang_b_Page_104thm.jpg
78a698346ce8c99c0d40c5cd5932b130
ff10a4f0619f519cd94c051979565e80c15d0a8f
57630 F20101130_AAABQO zhang_b_Page_059.jp2
eb937b7a2bf5403c80ac5655de45920b
21ed7c47ea7f015dc933ae6ededec7c1dcb764b1
487244 F20101130_AAABPZ zhang_b_Page_122.jp2
bd4c652c10230770f6d7736637d7f55f
7d206536ed8fb126ba78301180faba84939a568d
75795 F20101130_AAABRD zhang_b_Page_072.jp2
1cc4746b0572b8936d9a3ec4b3c83d1c
c62865cbfea91b816d5b0e7ff33f3415811c195f
F20101130_AAABQP zhang_b_Page_076.tif
432cce258c3ddf44ebc1896f4a6d6adf
42bc702f60a1bc4df1c55908af9d827795ad2cc7
35583 F20101130_AAABRE zhang_b_Page_091.jpg
360437a9090e93e89aa4c4682ae750ec
99ae5275175b5fcc03adadc853ffca646c17e103
F20101130_AAABRF zhang_b_Page_026.tif
b01f7bd313912ce02034ce3f790f16f3
9417caf8c5e5f8cb319b76f08759634ea3d35d60
79128 F20101130_AAABQQ zhang_b_Page_046.jp2
27167da565ec13692de2cf8c9cf8549a
88ea2f9891c6a7f510411c7c3e7e3d7328993f89
54473 F20101130_AAABRG zhang_b_Page_034.jpg
e6979b1bea58cc9a4f1a86b8ec77aac6
b1e98bc81b5a139b3b1004d72b9fd7b80427fef9
602580 F20101130_AAABQR zhang_b_Page_042.jp2
97b48b364f996e8c07236e46cfb3397f
b666f1caad64625b67ff9370da8546231ed1eab2
F20101130_AAABRH zhang_b_Page_049.tif
7bc11c8aac0bbfbbd4f16347dbd6f6d4
297d0217ef9621dc6f816f927b042fd4d83d55ca
59339 F20101130_AAABQS zhang_b_Page_031.jp2
b69016c0279de4b56c435abb1d9b8bff
f0eb5a72959588116729fe9c2d6d52af0a3fecc2
10513 F20101130_AAABRI zhang_b_Page_091.QC.jpg
5f6ed4573e0e7b775156a088a3d2dd61
46dfdf4af3696469cccf68e0761c444a44885900
5184 F20101130_AAABQT zhang_b_Page_049thm.jpg
63033e5e2cf0dd1a7e31f00800f61b04
1649c972c16dae233922a3b43431e0abc8481144
69409 F20101130_AAABRJ zhang_b_Page_080.jp2
82daf81d2a3a4ea58070b1eecbb32b90
d0c1ac81c17a383ba3dce4cc11e949e90421cb91
50429 F20101130_AAABQU zhang_b_Page_044.jpg
ad3c2c5cd9e2a55ab7dfb316da486a43
db40e1fd1384acf65cd7ec3df1218204efd236ad
5706 F20101130_AAABRK zhang_b_Page_038thm.jpg
ac58c75cb16bfee69a6026af736a28a6
8c5fc80cf8aeaf826ce14a5466d949f2565de591
F20101130_AAABQV zhang_b_Page_099.tif
ed95db8d9049952a0df434e0a627e2d4
c04e136bd99e906dde7f98275ffcf036cc918c2a
50445 F20101130_AAABRL zhang_b_Page_073.jpg
a63c4ac62385b77303054296e0ad702c
7ccd5d371e4049c541bdbad407cd017e3f3707da
39617 F20101130_AAABQW zhang_b_Page_117.jpg
76fc0ee01d1ed540591a53b65152cb08
5063b7ccbe9d9e66df747517aab51b7889f8333d
1353 F20101130_AAABRM zhang_b_Page_002thm.jpg
eea38fd38534e2ce539015aa1ae2dcbf
2fcded3726d61f4268e7fb1a30ace9a8211170e4
639770 F20101130_AAABQX zhang_b_Page_079.jp2
6704c34045b25b656654ea903a31a219
f4dfa45ad3114e16f761ca4da21f4247fb90da71
22483 F20101130_AAABSA zhang_b_Page_069.QC.jpg
eb05d4ddbc3855f7db4952bc9f9bdfdc
116b72961badab04263f5018118a5fe0055cab2f
5769 F20101130_AAABRN zhang_b_Page_100thm.jpg
b81ba11a27ac4f4e92be6fdbf6f02f27
62e870f312f6ee68e94e7cfe376536d755fd65ed
9538 F20101130_AAABQY zhang_b_Page_012.QC.jpg
5d15d4ae876a24b84ce4a77c9819c66b
90f02c74671185f99cb0fa8069f8c23c2230f6a7
79440 F20101130_AAABSB zhang_b_Page_106.jpg
8b190f1f0b94d042409efb92499ac98c
47f7507d07be9d8e78911f53a7adc146886f5680
5325 F20101130_AAABRO zhang_b_Page_015thm.jpg
85bc76fb2301a18a52092ff08603b09c
a2c682e5be5e7bdfb0d7f6914c6e814bd99b21ef
5724 F20101130_AAABQZ zhang_b_Page_040thm.jpg
a2c68b41d62b01d627d5208230c950ca
9ef36b9a1fe09dad464f24779fc008177b56f83c
111832 F20101130_AAABSC zhang_b_Page_126.jp2
40963f9c009b21be8c4566347e294b44
f2b456982e861487f69f6586a4d1c46cb80e6df5
5316 F20101130_AAABRP zhang_b_Page_048thm.jpg
2d068fd5c996072b765bbccb65c23216
e1ff6fc5f371658429f1473af7bdbecacf4bae26
5673 F20101130_AAABSD zhang_b_Page_020thm.jpg
7ce116db6f176c50348d063d2a201ac6
924ce862c2b5115a236735fb5433d4fbaada0878
9488 F20101130_AAABRQ zhang_b_Page_121.QC.jpg
e3da62bcc77324ee92f1c204753fc8a4
8b72ac38aa19a3c78dac2dfd3c74de4a500554db
F20101130_AAABSE zhang_b_Page_094.tif
11d0098e9c758f98db30e5a23426d7eb
5377e9bca86b14e87c3095122c21c7e42b370cae
5756 F20101130_AAABSF zhang_b_Page_028thm.jpg
e0dae5e439b129d48e0583b0cf5a3a4f
2d2c76643de8692d47e3b026509794e403ce18c6
F20101130_AAABRR zhang_b_Page_114.tif
d809305efe3ea61b81d57d70afb72f73
17607b77d723bf871ec55e69c27bf799d996f902
71731 F20101130_AAABSG zhang_b_Page_060.jp2
71a46ec6e74e804de826b8a78f8ee542
2cc4f5a7995fe9db0a588f80e52dd2d2567ba09a
93830 F20101130_AAABRS zhang_b_Page_027.jp2
ccab0728c4db0e159de7c832013caf10
0949c456cee4cc0a7390df156ef8dd1e40277da9
F20101130_AAABSH zhang_b_Page_085.tif
f8da795c2eb732ae27894f838601da7c
aaaf1c294242ed6edb69f348c7d6b52c9a5b1c7c
6195 F20101130_AAABRT zhang_b_Page_118thm.jpg
1b7729d7dc735d3278f5742f6b8ceae1
1e248fa2d3c4ddefe7c26594494a5f2899bd88ab
3024 F20101130_AAABSI zhang_b_Page_041thm.jpg
2c7484470ffa9efd2ff55a3ef06bc717
63acbc53134b046cdfba7732476a9b63a6596c63
82027 F20101130_AAABRU zhang_b_Page_055.jp2
9e25a80416945bd979928b5fcbbb8a8c
fb6e45c02255d8599f08f792eda9607c4bdc5ab4
16583 F20101130_AAABSJ zhang_b_Page_004.QC.jpg
5655752d932323478c6964c8d6af3a4e
556b0df55ad608ddd5cd6efebcf8afbdd2576d17
22130 F20101130_AAABRV zhang_b_Page_030.QC.jpg
696b34d4b35e49d65e7bad7e69bf0cb0
b2c5d91ff13653fde97c807f2269e21438a3d435
21133 F20101130_AAABSK zhang_b_Page_063.QC.jpg
876e6599efc324b57f8b6bc1b9a73265
174da5308bc63a31e73563d47f100c140f0571ea
3284 F20101130_AAABRW zhang_b_Page_095thm.jpg
d43f270c9ca89ba3f7cf66858e8aa2f4
68631369e2caa78c175e04335328959d16661607
5998 F20101130_AAABSL zhang_b_Page_036thm.jpg
952ef029ae8b9d6836b19b97dee0283a
a57ba2f486bd882e1e65325bc44ef58feda855f5
17643 F20101130_AAABTA zhang_b_Page_065.QC.jpg
c9bf5cb7aa81f753ac354a89f93b6bf5
d73be6ad12a21522570b04a3a433da40ff792559
F20101130_AAABSM zhang_b_Page_077.tif
e3c8141a38fe5b8154244ce287e11f77
b1b2c9c131eb4040c38b6f023c16306ceea69f2b
18754 F20101130_AAABRX zhang_b_Page_035.QC.jpg
fc559ea5d7708df2070d31ac95ee4228
d6679da50d3808ce9b7284767f01368719aac046
19354 F20101130_AAABTB zhang_b_Page_129.QC.jpg
bc9cfe6cc8cdc9494c4728712690906e
94b7b238f85f61b98207ccfdbeddca0edd7c9e7e
F20101130_AAABSN zhang_b_Page_110.tif
8fa268ef1f86c62b01201b0f23129d2e
10d39693d634c608022b4655932e1ab3f40f1d50
3599 F20101130_AAABRY zhang_b_Page_010.QC.jpg
090b7e58f4ab5ea0c4992d4efabbd7a6
19104ee96ce6e91bb55e6514208640ed5b69a483
F20101130_AAABTC zhang_b_Page_054.tif
2dfa0b0afca1fe6b3a289bf3470fcdb5
ce5a5d20dc68302f22e6bfc663afbf1894cd3424
15585 F20101130_AAABSO zhang_b_Page_052.jpg
523e5c6ed4e3893a5c0352ba4ea0b3f1
320076a562c87bdf3a781ce1ac23cc3e0175e813
1051927 F20101130_AAABRZ zhang_b_Page_062.jp2
eb00d903ebca6b3cda0ce4738d2ab62c
74968530f31f4a0cc997917b6174d0edc1ac0862
40752 F20101130_AAABTD zhang_b_Page_059.jpg
dab930659da9ca680ff0ff6d90d36d15
917e9324eeebe72c3d450609ecd184d88d7b5bb9
66059 F20101130_AAABSP zhang_b_Page_077.jpg
691d7d7f1d4f59e9fba32436fd863832
4663e1ac0c6f0390b32fce7009bf6a82c67ba3d2
6015 F20101130_AAABTE zhang_b_Page_033thm.jpg
ca599b0e42158116f2ba392ed742c33a
5b5917c0b2ed4be3b4285b839e5c2e2eefefe0a3
5402 F20101130_AAABSQ zhang_b_Page_013thm.jpg
1197711828af8ba3e19e301b5bc744d9
f6aca6ef20558e53a5ac304d06f5e95219980b58
F20101130_AAABTF zhang_b_Page_050.tif
0440687344625d163e3e761a641393d5
46e73b3017f3089f07a8c26863697c72500ee2df
7561 F20101130_AAABSR zhang_b_Page_119.jp2
de2499f94ae3cc6958d76b5ee435de21
8c901ea1ba8970cb0fef5e0fe7a9c2982754baa9
63030 F20101130_AAABTG zhang_b_Page_069.jpg
239cc076f77f003580b05a1d9f57c82c
3492928dfd865a1daa8055a9633a173b3625b4ea
109009 F20101130_AAABTH zhang_b_Page_024.jp2
c597505bab6f8fac59415c710855c6e5
419931db5ae2f3856f22a3ed888c0b8cf29d4140
67895 F20101130_AAABSS zhang_b_Page_030.jpg
5248cc76b4840d05efed2f1ebf483cec
aa283aa4fc1f2c3b084aed5f8d0f10f7a1616e66
10342 F20101130_AAABTI zhang_b_Page_076.QC.jpg
174c7b0cc55ce9069dc11caa43851d02
c9365b9e4b9a7ed1777ee64f342945a3a3de6637
17906 F20101130_AAABST zhang_b_Page_073.QC.jpg
78a9fa45f4e1eb34a3f355eb3cbbfc33
d45d5805841a824ffc121fba7729e08817f7f144
F20101130_AAABTJ zhang_b_Page_001.tif
a045628a23413768693a94bb8a209f14
b1e4ef709147b9fc835a2b7ad7d9ea892d496328
74839 F20101130_AAABSU zhang_b_Page_126.jpg
ae5002c2309eb534bbc58b53e6559112
c86ff38c625262473edd8ce0ab9e1f452305ff79
96136 F20101130_AAABTK UFE0011000_00001.mets FULL
87b2600e94a715b91c588e306e755031
19b9e96ae8992160a4f153e38a644739c8eec08c
4207 F20101130_AAABSV zhang_b_Page_067thm.jpg
d31c546f6469f7eae03fcf7edce72751
efc1e57ebc71ebd7b634d74c181a4ee6eb2bec30
14957 F20101130_AAABSW zhang_b_Page_089.QC.jpg
f0b2ed7e6755d868d75e4949d3d7606d
0e060fd62ebaa10ccc915a858757b8f0af88b2d3
19179 F20101130_AAABSX zhang_b_Page_050.QC.jpg
309296ce60292fc7049fa72e9118a8b2
00cc96d58eae48606252f6d51c4d0938144e8b08
72961 F20101130_AAABUA zhang_b_Page_018.jpg
1211655689ba9e78474738ddc553ad82
d74dac8acdeb74f389d40e5197378c80ca43e31a
10067 F20101130_AAABTN zhang_b_Page_002.jpg
a4261fe03c8ada2c0aaf2cc1f0d85505
981569fc63f35264265b78d032fd22aea981b29e
61509 F20101130_AAABSY zhang_b_Page_058.jpg
bc97d70bc7afd5324628918020d882b4
f1be97c326ee4fb42fbd4d7626763475d3143080
64108 F20101130_AAABUB zhang_b_Page_020.jpg
79d269a5808df6d024c67a2bd803c5a6
67d33326e54a27bb4847276a81e3b8a3360aef36
47226 F20101130_AAABTO zhang_b_Page_003.jpg
93cd3f38df2a661669085da9f3cbca08
bbf73dd908786f2ab2f9c1aa86a63f72eb8b0a72
60330 F20101130_AAABSZ zhang_b_Page_039.jpg
d883299349c5b95489b134e1d1baa0de
90ad9b44037c4388261c98e979162b73235233f4
42719 F20101130_AAABUC zhang_b_Page_021.jpg
018559795c3f6053f1ee65835e659ce8
ecc40a3495a2a99c967f4afd3f4062327d500b0a
58894 F20101130_AAABTP zhang_b_Page_004.jpg
de89e804648a096299583d3a362b5324
b85c38d1402420a2ec7b4f5b617b86a0d14aa133
63511 F20101130_AAABUD zhang_b_Page_023.jpg
349a08c86081668426109fbd8706d138
112c9bea48e67c173acf6ce386742d8fd741ae16
40715 F20101130_AAABTQ zhang_b_Page_005.jpg
c6e0c4e692352592a5a2f72ab481a358
196d6dc0caf0fc4bac24483878abd462079672f0
65920 F20101130_AAABUE zhang_b_Page_025.jpg
8a8abcedf6fe5b763c7bd855ecbb2e77
9b09643a0ebb6e35c06a6bb659b56fd6d56b2862
20569 F20101130_AAABTR zhang_b_Page_007.jpg
5851ef0b8d58e3018ee183ab829ed10d
74defd8ae00821197433d5009c662676f5497501
65035 F20101130_AAABUF zhang_b_Page_026.jpg
55fd0ff2e8256c5dbb956dc2cf8b4892
e2228aa762956d0ca53af328cf29c986076d713a
60596 F20101130_AAABTS zhang_b_Page_008.jpg
9220bde1947b613f4d9a4f7bd0646a05
cd50fb08473605b974245c1d01b8862ee671bbf2
77380 F20101130_AAACAA zhang_b_Page_107.jp2
68a9673bd1d8289ea2ff0d31ad30f8a0
1ee6bf152b9798617196922aa91f118e0f077d49
62925 F20101130_AAABUG zhang_b_Page_027.jpg
452a0a74da74520e116c42e0f9c6b945
e96a54ef989c9cbb27ba11f98b0baf661667fd44
66451 F20101130_AAACAB zhang_b_Page_108.jp2
3baacf607e839552a716bb9325b54e22
fbae4f603d90e9748af56094172448d8fc086a28
63416 F20101130_AAABUH zhang_b_Page_028.jpg
1c8e1717d06e150cfe414944fe0bbb07
4d6d8c8f4801a3dd6b848a70714a69413d35dbab
67299 F20101130_AAABTT zhang_b_Page_009.jpg
96a67bc37d68ad75a0be656c0ede49a4
04b2762899af4c006e34ed58333383544256137a
38496 F20101130_AAACAC zhang_b_Page_109.jp2
3ff1dfe8e252792b57310188171126e2
58b24bceebca81afb8867fb635eef29636d85adc
54070 F20101130_AAABUI zhang_b_Page_029.jpg
decb6a8f3c8f3f719dab114ef480911c
8e533009ede452409cc32ad8ebb3033a9327315b
12410 F20101130_AAABTU zhang_b_Page_010.jpg
cfc28a66cfb661c860f0054ed0bb51cf
c66ceac64939079530ef2fea9a4b56cef38a03de
81091 F20101130_AAACAD zhang_b_Page_111.jp2
64eedd1ef8485994ab61ffbc9acfb231
542ed3f89c2b0759b9759f4b494d48cf42bb3b16
42283 F20101130_AAABUJ zhang_b_Page_031.jpg
839614690ae2d8fd2445c77139ecaf2c
75fb0a01e9c0bc447e24de961ab10825bc6160e4
59801 F20101130_AAABTV zhang_b_Page_013.jpg
402685192b0a759d9f397fd93fcff524
804b2405f1747272b768e473af8d527ba304e13b
104712 F20101130_AAACAE zhang_b_Page_112.jp2
f08822586b779cf96e064243c38ad127
e8e67e172409c66caee9f48f2d2be41ebb5ad1e8
57084 F20101130_AAABUK zhang_b_Page_032.jpg
e39efb028d78cff4a896cf0e31f8153f
072fa43fcd55091d7256fa07ba76a0cd61d610f3
67973 F20101130_AAABTW zhang_b_Page_014.jpg
ac9f7f9d5707bba8fffa832d31812cf0
99a840ee50b5337e4562cbc6ca934653ee0438ea
73054 F20101130_AAACAF zhang_b_Page_113.jp2
36c2d6fdba4ab61b701c8d13d9b1f091
3bb65ec62c72cc803004383258f8a95b2219b520
66029 F20101130_AAABUL zhang_b_Page_033.jpg
6698e10842b3bdd6337cd2564d65dc2d
8bfeacac1f8acac08597b66a67448d2d826f8ba1
54154 F20101130_AAABTX zhang_b_Page_015.jpg
52f33d1e8c163eeb1e0f5ae77161fd5a
0d093c787be7d79e391b19fa60a24e5a03095fd7
85824 F20101130_AAACAG zhang_b_Page_114.jp2
1a8f7c32c5c98b31f5c38d8f9deaae08
a307963ec9354277e8285cec87f2b8c96a7b2f18
53164 F20101130_AAABVA zhang_b_Page_057.jpg
65c4ce1cd239037cd3d62922fcd4d83b
a05bab6ffcc153ddeb9e9b5707bc7a93983dc148
56222 F20101130_AAABUM zhang_b_Page_035.jpg
ff27ccac899a5d935bfbeb6e1d9cfcdc
924827e875a47525cbad81b8bd94be6b6e902aea
62544 F20101130_AAABTY zhang_b_Page_016.jpg
3bce00446d25db2087b3cb5b02875eab
5e037faa5c2c35c42c4a4144c81d95ef1bd7baa7
67717 F20101130_AAACAH zhang_b_Page_116.jp2
7c14abf98c453434d5b2a4ffa857ec12
4e10f82a9f486a6aea8ff848fe4f615200beba10
49317 F20101130_AAABVB zhang_b_Page_060.jpg
417b006d0f22a547ef53cd03fc75b2dd
074b3eebeeff35dcdbacfd746a541be5514c65d5
62150 F20101130_AAABUN zhang_b_Page_036.jpg
1bae92c2d9770232b25b6c688b6b52ca
5990334aa7661c7812519b7017f340c58d9cfdb8
55639 F20101130_AAABTZ zhang_b_Page_017.jpg
b7f13fc3c452d7c55581b5f5ca660774
8e84e09bdde12cb78e4428b36526da667f4942f8
53065 F20101130_AAACAI zhang_b_Page_117.jp2
c5edb10fdf8eca3c07902e9b1bc2cd45
fa1addc0b73af2066e3740c2d29c473fa6a93c3a
54491 F20101130_AAABVC zhang_b_Page_062.jpg
f1e9c13b8fdaad0d65746a9e711ee1ff
f76e463bfeedeb2b303aceff621e7081a4a4c8d2
60589 F20101130_AAABUO zhang_b_Page_037.jpg
8bd1f49bf742a9ec4f1e4e7c827a845f
7797367e72fa052cd5c20d7cdb1bebb546c01cd3
100379 F20101130_AAACAJ zhang_b_Page_118.jp2
23ace12d222778f384cda24efd5cadb6
09f9fc3098d1a3d54c7b87b861875f24aed6af28
49632 F20101130_AAABVD zhang_b_Page_064.jpg
8420f96362dfa8a0237116f828cd9601
05e7ac99043e354c1a0cc0caaec365cc692bf657
57262 F20101130_AAABUP zhang_b_Page_038.jpg
be0eef2ee76cc3b024a474126f647e12
028aad270ecd1251792fbfd1dd4dd1224f055743
422202 F20101130_AAACAK zhang_b_Page_120.jp2
8f09e70b2072b216b755fe908d4d3d20
74d72ff0b765b92b7053a4fab56e0e47a29d65e6
52720 F20101130_AAABVE zhang_b_Page_065.jpg
76decca2fa872431e1174b57075d76e4
360a8f73852cf9f9421de23ddca6d0337d3f0956
26448 F20101130_AAABUQ zhang_b_Page_041.jpg
6f28d2b5a1ebc82b52091c370b0f2dc0
8a8db16c46f269c2e3b444c8b9a18f19d8825fef
470157 F20101130_AAACAL zhang_b_Page_121.jp2
44cd48d4aef2ada28626a3c731d59b7d
dce7777530f72b0b4a87039d13a53676aa8521e3
39201 F20101130_AAABVF zhang_b_Page_066.jpg
e085723b86de2bcf3f581ab9bb1b7e56
ce9100f0d453d5f4f1c619359301714655f7cbd1
44061 F20101130_AAABUR zhang_b_Page_042.jpg
0f8cbe1e4e8ab78d0fd1cbdf0982c92d
ef770f9adf736db7d1897d33d3c3d40e99896c1e
715222 F20101130_AAACAM zhang_b_Page_123.jp2
c89801e286894edbf854a9f4586258e0
e8a777818622c1b2b55c43b9506031814232dac6
34608 F20101130_AAABVG zhang_b_Page_067.jpg
8a6a4124e70e02757c8e8657bcb15298
c09df60d700431bab7dcfac26d4b5e39c758d92b
31331 F20101130_AAABUS zhang_b_Page_045.jpg
4adee360970d988a0b3b239b83e43b1b
fe61d169601bed8ce03e813b9f71538a2e555b0e
F20101130_AAACBA zhang_b_Page_012.tif
05bbf1d80bf17edab71c5d435ee7f477
047236bc09041444ca81d18cd8e76b65698aebe2
630778 F20101130_AAACAN zhang_b_Page_124.jp2
59703b9a6b727ab64494d829ede34c06
a2a246d96624c3646ab9a426c7aab717caa95c09
66056 F20101130_AAABVH zhang_b_Page_068.jpg
8ae59f25acc72dcaf8c0c97bc4d53f37
b9aec460af3c0dcb7ce10b1fb4a720299daaedc9
54453 F20101130_AAABUT zhang_b_Page_046.jpg
b3813cb3cbe3ec896a71e3cb38beff17
cd89d4d274b0f6ab1f4d68d19b8e462f4b12e621
F20101130_AAACBB zhang_b_Page_014.tif
54c0eba40cabbcedd19f5d454d3a125c
865e22b5d41509070c51358737764db386ed04f2
663296 F20101130_AAACAO zhang_b_Page_125.jp2
b8d3a4286012c640e2dcef806cf29421
f23bbef2456c2e1ca891c76c43788a431b1365f3
48989 F20101130_AAABVI zhang_b_Page_070.jpg
2688ad15881a3d24520d3f706163d827
c675db3d5a601e7fe580dfa8f3d4cb5dbfc19eb7
F20101130_AAACBC zhang_b_Page_015.tif
c228cfb34d8dc6187931719d5d62e1bb
652751086fd39b208186f99fd1952f2bc639e2a2
122727 F20101130_AAACAP zhang_b_Page_127.jp2
7c4b1755c5d9cac438a4c88b2133a8a0
c2dc8f2a6066824c1b126eb22d013c00ab79b144
52273 F20101130_AAABVJ zhang_b_Page_071.jpg
a359e7151d49268e82f599a3415cbb56
d53e0294a00de7cb4d13fa7e65adf6abcfdff6f7
53788 F20101130_AAABUU zhang_b_Page_047.jpg
e264071c66ce2b017e0d1204c5d19d3d
4fd8c7dcefe3d80cd2d864acac7483861ecf7e08
F20101130_AAACBD zhang_b_Page_016.tif
080818ab3ade4291955d6fddd12a877a
23981e6393cd79c8f719cfbb2318125da7e4fc5a
119912 F20101130_AAACAQ zhang_b_Page_128.jp2
638c1d9af91d67f7672d394062a5edb0
3a913cdb67dec715f9da5397ff8ed18c7e8757fd
53147 F20101130_AAABVK zhang_b_Page_072.jpg
8acc8b10ded4cb3dcb23c50b4b1982cb
e1bf8631626136df9d7a0f7bbeee8454d2f14152
55178 F20101130_AAABUV zhang_b_Page_049.jpg
d1d1611cb40ee2eb03d314ced16d754f
c6c71a10af0c0e6dc5581b6fb75a27b6b94bc870
F20101130_AAACBE zhang_b_Page_017.tif
521cbec7ec7216dbe935de8d5ce844eb
2df1a5ee0448c820fa57ca0231dd77fee80c8d62
92863 F20101130_AAACAR zhang_b_Page_129.jp2
ea0803f2d3fe0a10bab7933d1fe5f8f7
29cb3fb3a706103c6e302d17a371886b0353539e
67504 F20101130_AAABVL zhang_b_Page_074.jpg
c57a5bbccfcdb1f31d1399b9b68a52a3
bfeea0341dbb3aff5a6baa38ea38375a3eba6143
59835 F20101130_AAABUW zhang_b_Page_050.jpg
2e0b2d58c8490ffca554b306a63483b3
09e2b13bc4170820271be4cc3a8b73a678030c5d
F20101130_AAACBF zhang_b_Page_019.tif
f2bcc5e4a7e0e50bc418f665710d9a1d
d8f8acb39669455dab4a6dfe8255241029b7157f
61514 F20101130_AAABWA zhang_b_Page_092.jpg
cbdc96bc753be4f1b996a07be944a4f7
a9757120e7a4c0fdf5c0aa0b36be078716425504
41754 F20101130_AAACAS zhang_b_Page_130.jp2
907add1b06e5f484a31b23fd923a2f40
a31329a087fe42c2d9c07bafdaf5dc4ee2dd73c2
48460 F20101130_AAABVM zhang_b_Page_075.jpg
144d63ee51bdd3931bcdc966df1328e6
f5dd185548aa28489e36b28e73699074bdf16198
57200 F20101130_AAABUX zhang_b_Page_051.jpg
658721b04f67ef5f006e9e9117be10e6
6aabec5476e66bf1a91ba047890b9b1fa077a516
F20101130_AAACBG zhang_b_Page_020.tif
fbd3042bc2147fa77bff87597dc7482e
2a0c5897b28b7c304eb9256d0bd51f3478e90b85
38267 F20101130_AAABWB zhang_b_Page_093.jpg
2ad2639551cc86bf063a1db40488b6ce
aa45469b921cfd2dc5fa52259d68d96762144738
F20101130_AAACAT zhang_b_Page_002.tif
ef8dca9e65736b8ae0e4784d1c7a0127
b9c647497f31bc4abe98922106670de74e0a02c5
31489 F20101130_AAABVN zhang_b_Page_076.jpg
e21e69c393cc44a46f8c2d54182203f7
01335a4e001ffa14cf8de13b1d1a374730e7168f
40358 F20101130_AAABUY zhang_b_Page_054.jpg
fb92299056ba126f4fa29dbeb50cf1ae
56996b12417c7ce965c57ecda0e8898a6786c932
F20101130_AAACBH zhang_b_Page_022.tif
af58d3a888e3c905185bd63f6cb2e672
7e59bb4ac2f6fc254bba08c19c19917218cd223e
33713 F20101130_AAABWC zhang_b_Page_094.jpg
6d5188006cd830c8ebb867f65671866a
f783210ac3ca679da8f516b24750826dc3b3d15b
F20101130_AAACAU zhang_b_Page_003.tif
86723465bcc70e52595c524c5682bac9
4edd9e64527bb881b0ae48dc671822a42b66a2e4
70206 F20101130_AAABVO zhang_b_Page_078.jpg
3e66591d4470853dc9da0787608d061d
fffc0e1c12093f5a1a8b5cf703d01140851ffb09
50563 F20101130_AAABUZ zhang_b_Page_056.jpg
289a6d304f6095cb23c7414089a1c1ce
abb5fe8151587b6b6a09ed3d887556b4bb6b6884
F20101130_AAACBI zhang_b_Page_023.tif
4087e496006dfaab96ba6cc11e143305
a8d6de49ed1752d81e161622c8a0d7d8a596e68c
64506 F20101130_AAABWD zhang_b_Page_097.jpg
1f6eb558f3a2fc58a387e082507a2c94
810e9d64e6c35b96d15632e60703be83b911b67e
F20101130_AAACAV zhang_b_Page_004.tif
ffb523a86381779016ab57fd241499b7
647152fe42bedb6247ef48ba8891e6ad0650b7c1
39274 F20101130_AAABVP zhang_b_Page_079.jpg
1b1293f676411f863cd3d91f5c0c7977
f1e9d8e71ef44025410d8f2e7783f909ddf4687b
F20101130_AAACBJ zhang_b_Page_024.tif
e57f7132396aeaedf2aec6ed3b5b0024
877855ec3ce4f1d84e660f2f4b780ecea0395c4d
80412 F20101130_AAABWE zhang_b_Page_098.jpg
d9b83d6d5e8179083cc85eee7df75cf3
b0845ddba54e2f6bff29aa93d8b5ef39c5099b63
F20101130_AAACAW zhang_b_Page_005.tif
21e86e4df2db7b6f6b221b68488e92ca
5e8031dd38e383da07b83820933c9133fa42c5a7
49144 F20101130_AAABVQ zhang_b_Page_080.jpg
8663b3b383afae3da6b660d829268ed4
f0ff9b6cbb34b387595b7381de84314e27d30942
F20101130_AAACBK zhang_b_Page_028.tif
fc4e531bc5d5d88929578a57e55bd59c
156a56bfb2dbb640b52c4dcf5a3fc8da761b37e5
73602 F20101130_AAABWF zhang_b_Page_099.jpg
a5a7f10d699845058b557a9c2cece57a
1fde40c5f85469f23478376a20c5687fdb3b8472
F20101130_AAACAX zhang_b_Page_007.tif
d9f963199e77c0a152738febb4b6ef19
821b42af4187f28f7f8bf3272b9c5abcabd4c347
49885 F20101130_AAABVR zhang_b_Page_081.jpg
db7034e75004e9537973f8d31acb0acf
fc95a7d675ae78467f4eedf1ab5ed19003fbb5af
F20101130_AAACCA zhang_b_Page_051.tif
f6a2c4619199776afe94267fa9118a9a
43a8dcd71974a2872d80ed7337c1a0d8a8b69e28
8423998 F20101130_AAACBL zhang_b_Page_029.tif
70ceb15adbad02b6f2edf4520af7e1a2
4faf9ff398665252e7a80d432f771de04b9ce2c6
59444 F20101130_AAABWG zhang_b_Page_100.jpg
a1b8e6d126123ce6499f350d93e8c486
0226a9002258008bf92cec02db57c05bdd7787f0
F20101130_AAACAY zhang_b_Page_008.tif
d044a7a971a7d02a9cfa73d8b55edb1b
e9596184bc27179518099e0e5eedee13e5f9c626
65114 F20101130_AAABVS zhang_b_Page_082.jpg
384cb2a514e36f7001d612194355a128
e6a2ffaa308d8c803992b83a00cf5c96126298ba
F20101130_AAACBM zhang_b_Page_032.tif
6d1cfcb376978c7e5daa60a8dbbec202
acde6c03ff888103f86c1fdd5cd09b56c73a5d8d
68726 F20101130_AAABWH zhang_b_Page_101.jpg
33cbd8db7c19c9550c07d3ee99cee82f
1046c95b432eaf87e0f610db4ca6cc635120e402
F20101130_AAACAZ zhang_b_Page_009.tif
fc7d0ede6795edacdf8ec2891ff1c4de
800f7142c952d47bf209cea0893f8dd5f1b4cf25
61276 F20101130_AAABVT zhang_b_Page_083.jpg
68abccd571d9c95bed28478e55206061
a092616af72f0ce7f2554707494553defba1c6a0
F20101130_AAACCB zhang_b_Page_052.tif
1df2be315bf237f03c42be3d8737862d
812883ef93ca37208f79d6f521cface5041508d8
F20101130_AAACBN zhang_b_Page_033.tif
5b610b227fe6ef81b6ec2142147cb744
92f4f081b4828ad9d96dca7d099084de026bcd52
60854 F20101130_AAABWI zhang_b_Page_103.jpg
11fe922aae911c7df071989072e9a42f
139866473ef64919dc1b0ea7e09b7a23ff0ddd58
48662 F20101130_AAABVU zhang_b_Page_084.jpg
e54464870931384c892d945d6b742945
0ec34178d5d54f83c76555142f427f88fc381690
F20101130_AAACCC zhang_b_Page_053.tif
384a05442a978698afa6d0293307b7b5
8974fdb285d129e0e3b6592c15b0c118b4824e96
F20101130_AAACBO zhang_b_Page_035.tif
f25bdcec0b89e31d0653324a382fa41a
a9c8f4bd2801ca58dd74346f94c1c9e36fbe0462
61256 F20101130_AAABWJ zhang_b_Page_104.jpg
b2b9ece3a1edf5c4109ccc4a16d8897c
4bc5f3d03a21fab6e5918a9c9011f1560535283e
F20101130_AAACCD zhang_b_Page_055.tif
6d2856fe56d4055aab1fa8fdaa726ded
2caba5badd04f6c6fa0b4866df4544c7257660a6
F20101130_AAACBP zhang_b_Page_036.tif
17d47809d7301d5ebadefcb5a560ae71
e45d65edd4d8b46ae6757aca197d73ab99cf6e52
45829 F20101130_AAABWK zhang_b_Page_108.jpg
044f4940c943961bf3e2f3b205770442
99a0121d3afa964fced1c9a0c2cef87c711f5d05
43456 F20101130_AAABVV zhang_b_Page_085.jpg
a6794b8ea7b7527dc8487bff2a1ece75
f72a9e1ca737fe1acad923d88fae9e80c59f1066
F20101130_AAACCE zhang_b_Page_056.tif
d92e462d62eba6698d0b14e900d20436
6b2405bd6adf1bff93ef845c72dc953fa3dca573
F20101130_AAACBQ zhang_b_Page_037.tif
4f1c25f4d8c4a3ff3f6acc408be0783f
6ed6c911177851ea1465dfe23fdedc8188353a4e
30324 F20101130_AAABWL zhang_b_Page_109.jpg
fde028e09a97b35fec8e4e0bd9197ff3
4f6852682ab93a78fc02aeea695db4037ae49eb6
50285 F20101130_AAABVW zhang_b_Page_087.jpg
b81d86fc8240ea6444db3676b1f20741
693110aea7faab1f9d38eeb66a7a9a57ffa5e459
F20101130_AAACCF zhang_b_Page_057.tif
b2ad3b2328b5d6a11204384bd897f53a
4e5a898d6b1b9af80ec7fcfe7be8b0c0bebacd3f
F20101130_AAACBR zhang_b_Page_038.tif
c06c2f8ae8eba82370be7b5d610a2ab2
16b7cd996e5f08fb2974c4fc81cb56f19d1dd720
49905 F20101130_AAABWM zhang_b_Page_110.jpg
1808486aa00862b4b12c66934718744d
17c0f78ac60c1d01cb09eabab91982d0abd6e0d5
51786 F20101130_AAABVX zhang_b_Page_088.jpg
86a2f747f718a83ea783b3bfde459a9f
f06ae635c726a711923fff2e155645eb1fb544ce
F20101130_AAACCG zhang_b_Page_058.tif
737d904bf5f0680071694bc02a44284e
4a4001abff82718ddae06a74cca6b8388fc6e09d
67713 F20101130_AAABXA zhang_b_Page_003.jp2
dbe1ab9011c5dc1e799f5160be15e4b8
e66fef83de43e6c1ee428b931963549b5ac126c2
F20101130_AAACBS zhang_b_Page_039.tif
3679af40d5347d6828d30bdd2352eefc
793eeb96e321fe35e18ed3219f61c6a27d37580f
55183 F20101130_AAABWN zhang_b_Page_113.jpg
57133762b67edc0cf5ea06e4f9d43480
0d11df04243d2c5007631da53fbae4e98ec376fd
45038 F20101130_AAABVY zhang_b_Page_089.jpg
5a29e2c0708dc1c1e51160a08a03df13
2539bd097bdbc08fdad439f8b49201c35170212a
F20101130_AAACCH zhang_b_Page_059.tif
d182faa17a60973701e93fbdd08e6157
b3133000a225ea03e952666496c865859a339665
1051979 F20101130_AAABXB zhang_b_Page_004.jp2
bb38fb2f281bbb5f347a517ea9130dae
316c724f887367eaefa9a296173e59202dccf128
F20101130_AAACBT zhang_b_Page_041.tif
ac8a06f6ea3babbdaadd2c2da7949c02
093666b92eae8605c6af71f1a0e02612daedc7bb
58561 F20101130_AAABWO zhang_b_Page_114.jpg
edfe955f0a3d543d7eba0a8459336929
76fa4f88b113ab2ddc26758da3d8a316b570d0cf
55749 F20101130_AAABVZ zhang_b_Page_090.jpg
619bd419e41192b2229ba7d3cf86ce92
e285e0ecc28fae6fe77bffc7f5dd11ebce9b155b
F20101130_AAACCI zhang_b_Page_060.tif
49cab5cf92a47baae78a12238f6f6ecc
bd85b34b378bd26096490f15ea8f67499e30c41d
1024666 F20101130_AAABXC zhang_b_Page_005.jp2
af231892001c3fc5041718b0bf1cccb3
7958544d8ded20e9d868f607728e377a8254d3ec
F20101130_AAACBU zhang_b_Page_042.tif
97c1957ecc0778404613aab533f9cb29
cc4684c5bf155def6e56d3c6a6277b34df881841
47862 F20101130_AAABWP zhang_b_Page_116.jpg
2331feb65201239f27eb43df00e77371
82ddb68cfa551d856897dc02d2c4758fda5ee8dd
F20101130_AAACCJ zhang_b_Page_063.tif
4de0b521231153ed1f0563fd6c40ecd1
fcb9e21b87cddfd005b6f8a76ae1e31c8be654c6
1051968 F20101130_AAABXD zhang_b_Page_006.jp2
3b846a638d31031999c088964d1e67b6
e57ac33e1797cf479c853ef10495fa457caa872a
F20101130_AAACBV zhang_b_Page_043.tif
cd37bc32afaaa21e7b2add83ca86c0a7
833db1b1094c721be63e63b0332d58b49c50998f
67788 F20101130_AAABWQ zhang_b_Page_118.jpg
87e33de3d3db4ab717c1f62d34f18f68
0ffad5ad61ae668c37c2c4b1e3e7bcdff6d93087
F20101130_AAACCK zhang_b_Page_064.tif
0f87385a2ba247101d23531f2a6b686e
b35ed6ff68ae7646d48d57ee081edab9a82bf185
1051966 F20101130_AAABXE zhang_b_Page_008.jp2
c8ed7e94657007de8c6cd54f5d94ab37
9a4d2abf215bb960408d6e751865918824b0c751
F20101130_AAACBW zhang_b_Page_044.tif
bb8aa25bb690dd93ea792382eb80f2e7
04b13008c1ca4e01457937cb78147f38660b4c7d
11488 F20101130_AAABWR zhang_b_Page_119.jpg
04a2fdb2ee28508fcf2570aa1bf473d4
f144bbc2152cec59c0cbde991d6dcc06ac421dce
F20101130_AAACDA zhang_b_Page_084.tif
0b0801cb1e52474d1bd764e27b14159c
7cdde9f0ea7c5d769d2a1858effaf630112fe332
F20101130_AAACCL zhang_b_Page_065.tif
e5a68f3f53563697cae76e939aa995d4
ccaa2c59b5227a2ce9b5c579436f8ce88f850d4d
1051984 F20101130_AAABXF zhang_b_Page_009.jp2
6ee15ee3ace54b2977e694c6260c9e01
393e635d9895006edcb148579cdf99b57f51a8fc
F20101130_AAACBX zhang_b_Page_045.tif
e79e6ef314688e058106ee809845f360
4f90b12fc7640b50df820821c7093c8fdc3d0ec2
24242 F20101130_AAABWS zhang_b_Page_120.jpg
68cc311a986fa302d479089fa9eebb24
7ca5c673f0c17edd0e327d96c2c89ab9c64012bb
F20101130_AAACDB zhang_b_Page_086.tif
5b1f447330b9b9467c8ba657798b2b84
6215501e96e7ac3968eadb6589bdef3e7b24e45b
F20101130_AAACCM zhang_b_Page_066.tif
58246a0ee20958eba5dd2c83c7883d35
e8f01dc2bf58e55f9262bfd76eab8cd285d38722
79789 F20101130_AAABXG zhang_b_Page_011.jp2
0c7d074a8cfbbbdea560b6cfe27d0cb7
ebb5102bc319c50c1cb67984d763b1646cd1994f
F20101130_AAACBY zhang_b_Page_046.tif
bee08382d827179a63ab3eea2d81c5f8
997f05e00d96ea7f75521f35d98385c7575d4f56
27912 F20101130_AAABWT zhang_b_Page_121.jpg
8f7bae31d25b63d9c12db40a056b492c
4e3b086a0582ffd3d138e9109710525266232e2b
F20101130_AAACCN zhang_b_Page_067.tif
6a99046f10c6f8cd1b3f294b51854e63
357d3f969ca530e69a20ee655faec673e8ba0e4c
38675 F20101130_AAABXH zhang_b_Page_012.jp2
0314df9943295172aa4a99662195f168
d04db3ac241866823e74d1ac1c56245149cbebc0
F20101130_AAACBZ zhang_b_Page_048.tif
6855987984d5507fdff4afa0790ee78e
9949f3c6b7a7e2ee84cc21c163d422498a24f685
37313 F20101130_AAABWU zhang_b_Page_123.jpg
571db87db16ffb53ff9e9c993a1d329b
de347d4a44656a1aacd8e358550fb9388e3bdcc8
F20101130_AAACDC zhang_b_Page_088.tif
6659379fa4cd86c7f508078fad544773
b6cd1c3712ae7b968bf2266631adcfd2161252b4
F20101130_AAACCO zhang_b_Page_068.tif
208abf6f88182eb1d5c5fe10ee2069f2
b215662dfb13b1e808d9bdb0c99199d664650e08
87823 F20101130_AAABXI zhang_b_Page_013.jp2
b23386b53b0b2a58c6c0d423b5d5e451
b8d173ebc10bc7a68d54d3d0f8b8cbb01603434c
F20101130_AAACDD zhang_b_Page_089.tif
3eab0f34e4c147a264cd23dc9b81cacd
f663de26837b2da0f4720cd0ffa607dea1442834
F20101130_AAACCP zhang_b_Page_069.tif
947a8062234d6d19c61b7ffb86a399da
75f1cb4626e8e2764594664e219530299a679cda
901242 F20101130_AAABXJ zhang_b_Page_016.jp2
5e2de92139e7f1d07a4b7751b6259f87
7ba8c522ad01d8c273d950f6244215b25e5b909e
33957 F20101130_AAABWV zhang_b_Page_125.jpg
84b13e03df3d4c19dec2677cb8d51f55
327b0b39a0efe626779d36dfe7cdb4975b5ac5d9
F20101130_AAACDE zhang_b_Page_090.tif
4800ba068f4e17ee6099d293542a9576
6eefac07ac3e1bbdcd436b0a2bdb3c1d027bd26c
F20101130_AAACCQ zhang_b_Page_070.tif
cf30b57d5345ab1615a3e9534e350f53
00c527bb8f3b30b1f009e628eba69c49e40a214a
855676 F20101130_AAABXK zhang_b_Page_017.jp2
329d7b00abfff68d53ae46971df07471
8f0d50cf2a7c429ba04d01e3e52b0099be288fce
F20101130_AAACDF zhang_b_Page_093.tif
faf9768b147f8e4e7ea0366ff057ba4f
cd13db36573a902e38238ccbdae40d1ace464376
F20101130_AAACCR zhang_b_Page_071.tif
a1e67c1164f796323b21b2d4bb0f9b48
1b04f964c150947359c325863d1b39d4f14fee0f
116101 F20101130_AAABXL zhang_b_Page_019.jp2
c479be4db71d1334ec42036222a9a0ed
3fda6adc0783f143461d8d5b930aeae87792aab1
64275 F20101130_AAABWW zhang_b_Page_129.jpg
be0a6ebf0130aeeb6d629ef13e5b0a45
0a35905c1921bd1f2df73e3e6acee61fb993f55b
F20101130_AAACDG zhang_b_Page_095.tif
562110bf971cce59e0658f064624272b
f6ca33e07c387cfa3bf80c02e4edd2115de29256
971030 F20101130_AAABYA zhang_b_Page_039.jp2
32b7646b98c25418d1c7f831a7b8f3fc
fe3279d2e0353a36b4c036e5205e9584cc75bd5f
F20101130_AAACCS zhang_b_Page_072.tif
9a229e80f3b37e386b380c353c2fd931
732746421296e8a0d863099b9675c04b3deddb4d
95233 F20101130_AAABXM zhang_b_Page_020.jp2
0462c8fb2a9338bb595c3e23b27e4c3f
21f8c5455b326d1811200086701a25ffbfb78e2c
32093 F20101130_AAABWX zhang_b_Page_130.jpg
9a760699308d3d937181fe12c4835941
b7b465a9090d34350cfd3167745fb8fc478d745f
F20101130_AAACDH zhang_b_Page_096.tif
cbc7b17bc29e0cc2d1c9349ba182aba3
b42f8b3a056dad9787a949b71d9f6051869e05ab
86638 F20101130_AAABYB zhang_b_Page_040.jp2
b4ff01a8be097919358918ff1905b0aa
a204ce2d4d3698ada74ddae40a84f939c6feffbd
F20101130_AAACCT zhang_b_Page_073.tif
faffb62e5bb448edcd60a266c52b0365
fa43fe22d0123112e0eb626c8b3ee1d227e9d227
685286 F20101130_AAABXN zhang_b_Page_021.jp2
b38cab6a58fb1a2b14b2284f48b09abb
63d60f5262e2a1c3e6f494c1418f7dbeac883715
24156 F20101130_AAABWY zhang_b_Page_001.jp2
eab752abc69cb6c385e874ac31d8779b
1919b93f0b7a4db2841053afe48e47ee1f5d2a3e
F20101130_AAACDI zhang_b_Page_097.tif
fc19a2a8bc769f0705b0be6324c916c0
7a77f0f16d65c5cc60415bad09a5c61e095eaf3e
324531 F20101130_AAABYC zhang_b_Page_041.jp2
4294d1099679fbf4e5a672a2b5bceb5d
7b25df895a74c2a3996b6e433fc7a8236aea9ac3
F20101130_AAACCU zhang_b_Page_074.tif
1b46662aa56acb222ce03a7a8991c44b
aa1268948a20e80f688577aa1df815387996bca4
23683 F20101130_AAABXO zhang_b_Page_022.jp2
7b59052649d209f0d4f667ae9df2a4c4
d2be3a93e27a245627954363b14c051429c16c81
5306 F20101130_AAABWZ zhang_b_Page_002.jp2
bc99b83fd9b786f06ce28be661ca09fb
e4c95f5b7729a0c4e92c6838103326d630aa1c22
F20101130_AAACDJ zhang_b_Page_100.tif
2ea7de0cbcdad47edccd1fbd6e9f6266
9894662c752350d83917255db50941869a4995e5
81872 F20101130_AAABYD zhang_b_Page_043.jp2
55fdf41a3d4770f23e0a3d0baf0e16a7
1b20213dd9425dc6c76b323607c51f2b99159ca4
F20101130_AAACCV zhang_b_Page_078.tif
05cf5cac0513eb30dc9330a10c9f906e
7427ce3d85e7a662d429380cf984b6127b675bc4
100079 F20101130_AAABXP zhang_b_Page_025.jp2
bc787ea4697fbfd5fb813aab81d79e53
6a9d12371b393ea30460f939adf7f2d49ffe3504
F20101130_AAACDK zhang_b_Page_101.tif
1046b5d920a89f66369b9905dedd5dbd
fef0822b68eed774ae2ed59e4b6eeec5caf51025
1051958 F20101130_AAABYE zhang_b_Page_044.jp2
4ba3445d01146499e55f8ae74be3f591
924b57980acf41a294d869514b8734f5d430d694
F20101130_AAACCW zhang_b_Page_079.tif
2f8d97b240597ce434ecfc3b6ea83644
1e1f2006eb1b86e9e186188ac574627d0488cbfa
97679 F20101130_AAABXQ zhang_b_Page_026.jp2
8c4a19ad67c122fd6f0f9496784c2827
2639236af77fb1a81373046cfc5092fdfced0be0
F20101130_AAACDL zhang_b_Page_102.tif
d5decf75103bb5dc937e68afbb7b07c8
e583eb0c0ba862ad3e48175a26ff252be49a46c7
868427 F20101130_AAABYF zhang_b_Page_045.jp2
d24cc389d51ce3ae7be37bdb2e73d441
391677bb59a148d2ae531dfac50044e7737cb3bc
F20101130_AAACCX zhang_b_Page_080.tif
5bf051e66c42cc731cca356ad2ba8811
df4a1ec2ed85f8b10e9eca3afb00fb089a6b7082
1051982 F20101130_AAABXR zhang_b_Page_029.jp2
fb1dd34b2f62f787a62276b5e95b2f99
aea395e8d2a835b86dedf43884cca32447c8597e
F20101130_AAACEA zhang_b_Page_127.tif
1e32735825db6805802126ec06d7a1ce
d81e4f60448caecf5a0a1ed169dae82ea74a008d
F20101130_AAACDM zhang_b_Page_104.tif
41f54685327098c6cee8d55561bc5c82
547852e9960c81dd7dd4acc961b3d899f444aaf8
745538 F20101130_AAABYG zhang_b_Page_048.jp2
245e835277f9e824f2ee3715687d8f37
c7b98b896daca9df57f0bc65c62d73e4229c0fff
F20101130_AAACCY zhang_b_Page_082.tif
f3a2902e1289e2a05eff7571010758ca
35c1b49d6f34d7ba0532da84214ec4a21768c0ef
101997 F20101130_AAABXS zhang_b_Page_030.jp2
f7d9b1e40a0f912060c165d2c4f2ab65
01f81458c10d60c8f89a0c4951ae644ff1d500ee
F20101130_AAACEB zhang_b_Page_129.tif
0758be958bb5374386ed9d06025e92ac
0c4c48676582111295c2109baf0162ae848f5a2b
F20101130_AAACDN zhang_b_Page_105.tif
dc5445959aca0c275b9a6316b0e20082
df6ba36b4312fed8a3266cb76c114a36cfe6fe82
769385 F20101130_AAABYH zhang_b_Page_049.jp2
7a8b08042930d171b69be0434d1d80f1
1a4977556cfa3041a12bb6da44380fa346fcc073
F20101130_AAACCZ zhang_b_Page_083.tif
29277422a66748b5d35d21516a5e1afc
2091e22d0f13d09d4bf8b26827d79663cb7dc99f
85415 F20101130_AAABXT zhang_b_Page_032.jp2
bd792f601eab3f6eb9e2bb2b75dc2ebc
4a5e75dfefe050c261f035f65af06cf8d41c0ec6
2390 F20101130_AAACEC zhang_b_Page_001thm.jpg
3e598fcb7fd121a8e56d263b32540eb4
01a51e9db860c79e32e907a1c3430ed5ea5ec4a3
F20101130_AAACDO zhang_b_Page_106.tif
38238f8e395556f9c2d889ed927a4d50
7375bce56186393f243a0db6caa8f8f42097c817
87878 F20101130_AAABYI zhang_b_Page_050.jp2
60a7f9f34805ba58fa22020b4a400ee6
3c9dbbc683b3977eff38f6869b2ef6f1d68bc20b
94266 F20101130_AAABXU zhang_b_Page_033.jp2
f606f557ff2b919c414dc48ef50386ef
8e935e4e69176e1222c4b4444f0ac6cf3857791c
F20101130_AAACDP zhang_b_Page_108.tif
8570c6dcaa30a0721913ce159f5b6d8a
144cc4b358eb0c58cf6f942a6dd9d7cf8593b6a2
938302 F20101130_AAABYJ zhang_b_Page_051.jp2
4a0901fdd01ff25faa90dc3cdea1c723
069c442b99104ae8a6cdf80f5bc5856d46a2c169
74068 F20101130_AAABXV zhang_b_Page_034.jp2
76adcbf7d6552a663de31e3a3ec93754
4b6df6605008db8098c9b2cad2b75183b2e37194
3398661 F20101130_AAACED zhang_b.pdf
dd0c14880567823a96d3381658c2d8ee
6b6dd14d90455fa6932eccd7d9902552cd08e309
F20101130_AAACDQ zhang_b_Page_111.tif
5445e02930f8cbea115427118ddc8d6c
99cb2e49c4014e150b1523f4363efe71374f7d78
15768 F20101130_AAABYK zhang_b_Page_052.jp2
767c29307c2486530ae6b3d5de81021e
629dab5f072fac968f52598dbc33bb638c0b3112
823344 F20101130_AAABXW zhang_b_Page_035.jp2
d35172adba746420877799ed4d392232
eff69c0219ea4d694fca1f5353a1c9c3ae0b268a
5945 F20101130_AAACEE zhang_b_Page_063thm.jpg
665a5b354f3e64c5831ef29f6b76fe02
82f478664eed704a8f90df73623d79a7bd6a8510
F20101130_AAACDR zhang_b_Page_113.tif
2e1e58586cf27bd617253ed0a7ee5938
4403e72b031c41fa1fe45b75413523cd30545a54
84498 F20101130_AAABYL zhang_b_Page_053.jp2
c9119f8983e8ce88a5607680b414f36a
f0bd1d0a9e68643cde2d6ba9020cfae80fb7d01e
9711 F20101130_AAACEF zhang_b_Page_124.QC.jpg
2a43ce6b5bb584dc2609d7d46927aef6
fcfa4e9ba65bcfac02d7460bc011b4983999869e
58799 F20101130_AAABYM zhang_b_Page_054.jp2
7fddb913fb1f79625cf7982f6839c93b
70f474b0607c37b88887b54934612c29c3d2ba0d
91734 F20101130_AAABXX zhang_b_Page_036.jp2
6572e183a7a0a1fcadb5e06e99ee50e3
102b3341832f4fc67b4cf599671586176602f5a7
23844 F20101130_AAACEG zhang_b_Page_018.QC.jpg
992761cefcbd5cc2af08531106b281ad
90c5fd0ea0ff37b3232477e473ac645928c132e0
71485 F20101130_AAABZA zhang_b_Page_073.jp2
a9bbfa4606e012ac5bd075b08d1ae3c6
05bd9795105c1e51e1f52b500f5847245b5232cd
F20101130_AAACDS zhang_b_Page_115.tif
05803a7038d9c30910161de61997a353
1eeeaed54f2eb1806fa6351a97378c4794a93b64
72022 F20101130_AAABYN zhang_b_Page_056.jp2
7ac624c2607cabd1e87731e6137dd67a
cd249fa25206f922e1574c497c69f9e955f21c0e
90822 F20101130_AAABXY zhang_b_Page_037.jp2
2c277fba277caac1bb17b95ae33f439f
5aeda4aa26ab672670bff12360d8df9a6d9c6f01
4707 F20101130_AAACEH zhang_b_Page_003thm.jpg
24098d0d28a30f750ebb90e8536bee0e
749d995f42c0404eb76a8ef54caccc9e19795b83
102555 F20101130_AAABZB zhang_b_Page_074.jp2
1073692c5e40a570b3abcf5022df5bb3
dadec614cbdf45b00226dc9eeed3a20048e8ea99
F20101130_AAACDT zhang_b_Page_116.tif
741308271fb37401253a43259e16a69d
b1c5b1c80a839a79dae527fc608024001a9a31f1
74627 F20101130_AAABYO zhang_b_Page_057.jp2
53eaf4ea0f9c1b9bfe48e603b0e36149
35ab1879607fffd47485dfa88cf4965cafeb2d7c
831328 F20101130_AAABXZ zhang_b_Page_038.jp2
37ed952988c383ff0ac2e22809e8f200
9349b10f0f449a4f979430edf67a6140a23ce128
5233 F20101130_AAACEI zhang_b_Page_006thm.jpg
f26e99ea2a3a22cb641df2d4188733cc
56f3a18dedcaee041e6a1b0663a5bcd9a110ea75
41667 F20101130_AAABZC zhang_b_Page_076.jp2
d47ad4921c6c88449169cc9585600572
677bec1f78bf55d796aceeba30bdd86a1d5d8274
F20101130_AAACDU zhang_b_Page_117.tif
8567738b35c0cefe3df21879e0fcadd2
abd5da48bf1df81fe0d3a8bcbdddb516040cf2ae
90502 F20101130_AAABYP zhang_b_Page_058.jp2
99fb49a9ea7ff7a681512a8b6f3b0069
38c53d0cf476f2029920d91094b6f258bad8cae7
5876 F20101130_AAACEJ zhang_b_Page_007.QC.jpg
e0afa83764b192ea6cc131fb9b37938a
cb7c4014c4f52ec0afdc41eb3ddfe00497b37f86
97893 F20101130_AAABZD zhang_b_Page_077.jp2
299154b3de86d95b95c66baeb3f8683d
f77d2acf4a55f5ff2bb5136c06423dbd092a2483
F20101130_AAACDV zhang_b_Page_118.tif
466a8123536eb8fea8b373a96f2326b4
f4d0c71ff35829f945a96d6878f007a4df604a26
962682 F20101130_AAABYQ zhang_b_Page_061.jp2
dbcf2f5444ac378b7897d4a429b1da12
1397880576ac8d06fcdc1415d36d8119a11cb8cc
140204 F20101130_AAACEK UFE0011000_00001.xml
13a25a5107a76d02186182a3424910d3
993c049888090d4cc5c4ef0a05c3db018f728d04
106050 F20101130_AAABZE zhang_b_Page_078.jp2
d0f3a07ed79d608f11d8850929fff90a
87dff90465efaf79ccb29599df064822bc1157de
F20101130_AAACDW zhang_b_Page_122.tif
b1b744d274d25ad612aa891c9c16d74c
e8da17231589208bfd60bc8f60a11a5189b07d8e
96920 F20101130_AAABYR zhang_b_Page_063.jp2
023a68475bac421cf64c0aceb5490728
bec1e1530f5b518fe541f417c2529db6b38d8b3e
17194 F20101130_AAACFA zhang_b_Page_015.QC.jpg
d192b2f32e7994e38f15afe63124cb2e
d55b532d55135cffc0289c137b82da051112bdfd
7207 F20101130_AAACEL zhang_b_Page_001.QC.jpg
87d4922dea4ef62b349ef17e6d0cac0d
5e00d44b54ab117fb5d68a3d685ed931b0abe041
696115 F20101130_AAABZF zhang_b_Page_081.jp2
870e08dca98233282296d470ed35ff5b
32cadc8496ba1478732eacbe9a59d7bb1e65e362
F20101130_AAACDX zhang_b_Page_124.tif
6e85a485c0553658ed4008b392b19de3
8315e374f94de36a838a2e5d637a00481148d56a
71843 F20101130_AAABYS zhang_b_Page_064.jp2
ed41d7e5fde084cefb17f0027dfc3daa
08d24d7edb179b5007bfac52e51bac1a85b36ae8
20411 F20101130_AAACFB zhang_b_Page_016.QC.jpg
2391db7a61fbe7a0ecab7d29f75d44af
a344bdccca8e0a7e87781a855774c35b40046a83
15612 F20101130_AAACEM zhang_b_Page_003.QC.jpg
183b0be067083619027be5c0413b59d2
6e151dd2e684be35110c1266c96f89f795f291a2
91972 F20101130_AAABZG zhang_b_Page_083.jp2
c04089570a71369bfff9b3f7c48ced63
74b33bcf80c9c92f4ca366551848bfac06970e60
F20101130_AAACDY zhang_b_Page_125.tif
d7569ef036f9e75d505ae720b42a70ce
696055039dec3a478a28e7c9681cadc93bd08bb0
74445 F20101130_AAABYT zhang_b_Page_065.jp2
ebc1171c0cbb4015269cfb942c29db80
056974543e4dcaef64941198b58dbb545b7040aa
5964 F20101130_AAACFC zhang_b_Page_016thm.jpg
08e9f7b18d101a91eb0f2c7ab44c6f33
43779950a84b3a9873b3f464e5a60d4208db22cd
3360 F20101130_AAACEN zhang_b_Page_005thm.jpg
f481dca357a9a61f21ac9db1754bc413
ed634633c3e79b956c7bd37b1dbf8b08dea523f2
759680 F20101130_AAABZH zhang_b_Page_084.jp2
59bc3bd36bbcb139fc66ce0b04fe476b
ebe648e13992233c01bcc747cae22a2f44c47b6f
F20101130_AAACDZ zhang_b_Page_126.tif
f8c30fbaaa56765c204bea6b5107d870
61094b186b4cda76e5ab0100db72f3724c18eb95
53520 F20101130_AAABYU zhang_b_Page_066.jp2
6dda8ee4b9edbe8a8c1819cd095f2b98
6ccfc26a5040c043c439238aab8c4f394f80f885
17241 F20101130_AAACFD zhang_b_Page_017.QC.jpg
55c4fe3895e5a1f2220b1b70448268a3
8237d15600a7b225cd1e0c3a6dcfcfca779373ab
18602 F20101130_AAACEO zhang_b_Page_006.QC.jpg
bae70e4d35ee468ceb90d002709ea5d0
2e14de1e1f8423e83906064a1a38df26f9fbc377
453025 F20101130_AAABZI zhang_b_Page_086.jp2
55a79277b40a3b62ecfcfda596913022
e411913c4bed7e44042f3e980a68168a49092276
632685 F20101130_AAABYV zhang_b_Page_067.jp2
35ceb830adad078d43c634c8055ed7ac
04df809c82c108fe5a21e0bdd6bd2b975f13b4d1
2081 F20101130_AAACEP zhang_b_Page_007thm.jpg
5ece07708531ef451217f964905e02db
45fee8559bd39c758eebb5dd194e99addc3477d3
56014 F20101130_AAABZJ zhang_b_Page_087.jp2
8731f56208e684aee0a897e245c30b72
f041cb6bb3e955b5c2c801519aef391aa1b77a63
101475 F20101130_AAABYW zhang_b_Page_068.jp2
a37b58bdd64a69689318019c7b926ddc
8b5cf736d04247a2d47ef1468a49c296cc8ed66a
5568 F20101130_AAACFE zhang_b_Page_017thm.jpg
825b67548e5dcd10b325d7ad0fb7de7f
f5424b4ff1bdbace78deba82d10e1a1075797793
17685 F20101130_AAACEQ zhang_b_Page_008.QC.jpg
122a3dbcb6ecdf2bfe14249fe0e027f5
0f7951688234b88ad740c3e3a692da8dabfd753c
70311 F20101130_AAABZK zhang_b_Page_088.jp2
ccbad0c32ead986c07b1d3954b42e194
e9d9d535aa777ce131812bc8bdfd5a50be0d82a7
90416 F20101130_AAABYX zhang_b_Page_069.jp2
bb93d0ddef42b1b38a7d662fa0e9cfde
b4585b6a104b3c79af74ceac848646d747d27fb8
6549 F20101130_AAACFF zhang_b_Page_018thm.jpg
81585e9b3b5cd064b8423891acbf33bb
6f142be2d77b86bfb21195e2964d96feea7c47a6
4900 F20101130_AAACER zhang_b_Page_008thm.jpg
cfca9ff17b82f46da0f65ad713fff4f3
5c850ecad13d9a4101a3b8b68755dd0c6c4d4635
81724 F20101130_AAABZL zhang_b_Page_090.jp2
8dcc71cc10a77cc85ea56480e8a57413
580253f16207c001e5f3cce1a67a069b0a67eea0
23569 F20101130_AAACFG zhang_b_Page_019.QC.jpg
dee146354219e51d6a015b1ee8d4a65e
15fdab4cdc65712a62342d13072bad7f7985f150
19679 F20101130_AAACES zhang_b_Page_009.QC.jpg
71782127985b36e08b82893a5317f306
09f7ea594253a2e82dc7fa6ab3754f9af801fa08
473839 F20101130_AAABZM zhang_b_Page_091.jp2
cc3b3c668b4ecc85f93779e4ebf9c357
606a688e87ee394e70d31d02559c2c29f8859b3c
70663 F20101130_AAABYY zhang_b_Page_070.jp2
e2a57fb939d066f4661cf47c77170eb2
0ede6b62dee38fe9e592d90b9f46e25b0ad78ab8
4927 F20101130_AAACFH zhang_b_Page_021thm.jpg
6e9b35f88760c036e9b4e03b27e484c2
85f0400f453d6ffda9629ec55f3f69b27926ad98
5620 F20101130_AAACET zhang_b_Page_009thm.jpg
07a6314e68634bf2540f04956d6bbeb6
fe70f1085c1a89c7fb30dc8d7f1500329d802cd5
87420 F20101130_AAABZN zhang_b_Page_092.jp2
67011a019e0e35c1049d84c64d30a32f
f0b64f8a4c45d89223c1fffb51c6dc095446be30
75748 F20101130_AAABYZ zhang_b_Page_071.jp2
d4e4201a35e9e1b8ed7230a3782dfedf
8564f5eb2afb9ba79797d64527428039947131ff
2349 F20101130_AAACFI zhang_b_Page_022thm.jpg
c419227dace77f87cb926435eaeebc31
562aa9bfcd858e50d21cd780a56037d0b1b14464
1512 F20101130_AAACEU zhang_b_Page_010thm.jpg
711605245a308153405ee423187f1cd0
e0d45ae79db7812b53aebd41c8aa3966fabaca57
557951 F20101130_AAABZO zhang_b_Page_093.jp2
8f635222078bac6bcfce91484e704b9e
f40ef4a0cd9bcf28d766a63dc82efede9321c21b
20307 F20101130_AAACFJ zhang_b_Page_023.QC.jpg
53aa7d7d710fa59eb167a50176c7bcab
57b7860af852e59a7250e8b1762a56ba79f55212
17287 F20101130_AAACEV zhang_b_Page_011.QC.jpg
9cae1b325fb072ee9c1b569660af1b4a
8d92f6a46529c87f0d663348f92fd77fdd24e89d
565743 F20101130_AAABZP zhang_b_Page_094.jp2
08c7588fb01923881ec84620821f2530
4103e106a6a060d1b066676ec6672fbe05b2d030
5776 F20101130_AAACFK zhang_b_Page_023thm.jpg
74f5cfb0bd6a6d2aff9e7d81953e477b
a8552ce871f9c84666e02c3618720d82244a90d6
4899 F20101130_AAACEW zhang_b_Page_011thm.jpg
f3be5e2e7913c38849bb34d7647b62b5
ef849785aab2bb66eeb6ea57aa9fc2305c65e77d
784777 F20101130_AAABZQ zhang_b_Page_096.jp2
e6835c60f5e13dbf72533fbe67448ea0
53caa5694d0a25b33b7f77591aa8d5d96ea63429
6563 F20101130_AAACFL zhang_b_Page_024thm.jpg
f001613f6c3c228fe7b34abe47311b36
0b4dfdad6b82719aa75d8a0ee66d42f652dde11d
3125 F20101130_AAACEX zhang_b_Page_012thm.jpg
040b293553972ce9def7da5afb73360c
2a1062b5142df8e7305e2bae403fc6c6c63ebfda
F20101130_AAABZR zhang_b_Page_097.jp2
c616b76669c3b26e667bfafac9853620
f2c73d6a2c091e06eda6890f520018cac25613e9
20512 F20101130_AAACGA zhang_b_Page_036.QC.jpg
a80d0bc96999b3793eaa7f60f17eed80
fb0776f8ece7912024e3edd12872fad611c43d09
21272 F20101130_AAACFM zhang_b_Page_025.QC.jpg
f4bc614f0eca12639382607549b1be70
95405cf866c9b943d60718c5ceac68297cd9a007
19306 F20101130_AAACEY zhang_b_Page_013.QC.jpg
c98bdb5cc8ac6a10588f7c01bafa8132
a0adbfa6d37d8e20f99ed673498b23c6bfc5cc28
1043566 F20101130_AAABZS zhang_b_Page_098.jp2
741e0d1d5f00165cdb64d007a4baf2a2
08fac26209d5663065cfddef109079fe17b7b22f
20115 F20101130_AAACGB zhang_b_Page_037.QC.jpg
cc4d855fdda6d93ee90968092249d869
8149aeabbd078fc08b515a0ae44f4bb93cc779f4
6107 F20101130_AAACFN zhang_b_Page_025thm.jpg
11151c6ae048badc0ad120048e4abbf4
d8eac0459f895818a4cac8811d0b0dd94ce8f1d5
6205 F20101130_AAACEZ zhang_b_Page_014thm.jpg
3c4b9e077c3bfc0f60a2d75f78176d3b
2b124460cf0ab1a59d95780719a3da0b8ffdf1d7
946224 F20101130_AAABZT zhang_b_Page_099.jp2
da27d872d1186b757a55ff9e1ef4f368
1017127a1bc10425cb5ae10d97d465dbd27094a8
18215 F20101130_AAACGC zhang_b_Page_038.QC.jpg
30ccad3153a858a438259f2d0e79ddda
569f1c7a2e7c42fe94303ec45ee6a87f01b83fed
6117 F20101130_AAACFO zhang_b_Page_026thm.jpg
6844483134391ec3bfb58cafcde4bac2
d4b5b79831a0195f35434efcfcd6f2ed37830a1a
939608 F20101130_AAABZU zhang_b_Page_101.jp2
54407ffefacbb1bafb019d238dd08639
68c35a2e1b2aea3963a03b049c384c2ba6029684
20042 F20101130_AAACGD zhang_b_Page_039.QC.jpg
2b7d9d17608b48fe72635b6b222c10a6
03d9293efd7bb0e4f3eba9ec4d707ad432185615
21309 F20101130_AAACFP zhang_b_Page_027.QC.jpg
7427dcbfcdb04f513ec75b0fc1353df9
b08adbcce85b29a8e557bfbddba84b703e49ba99
719018 F20101130_AAABZV zhang_b_Page_102.jp2
0515ca98716b33a76e5f3038fe4a1633
5ccb213ef6d8e82f84fbe860b495b77a2282288b
19124 F20101130_AAACGE zhang_b_Page_040.QC.jpg
90a005f01b9cac7bc36798fdbe50abe3
fcd85e697450273f2a886f98fdb6f4696729a33c
6025 F20101130_AAACFQ zhang_b_Page_027thm.jpg
e41d3faa0c538b6f2f47a38cb53e12de
39e93a271011f7af9ced307aae8fa8e98fdf4de9
90712 F20101130_AAABZW zhang_b_Page_103.jp2
c93f8515059f38f8c80178136721c1fa
3b21be3ae0935ed52ff500b6b9f237db5c2f812a
20989 F20101130_AAACFR zhang_b_Page_028.QC.jpg
b51e149ba1021156c47951b3430f7c0c
c889729a939a6b9b2af0231426a55e356130230b
91839 F20101130_AAABZX zhang_b_Page_104.jp2
b999c431abe15f5eb48df3723fb380ab
af79e3a971a8041d0cb5ba82d42da4adb251ca15
9412 F20101130_AAACGF zhang_b_Page_041.QC.jpg
4ac712b7df555f8cc068708fef38f713
31f3af1f8fd2ee8a3e245219a1a25838dd796b6c
4671 F20101130_AAACFS zhang_b_Page_029thm.jpg
2832fc4d678845fab19ae686fddab941
7b67d2a70dd2ba31b77c9ad25e070501cadb8d03
77986 F20101130_AAABZY zhang_b_Page_105.jp2
55fe01930cbc123b88af6540c3a73185
8addaa52624910b7ea6c5796483193feba1f0803
14271 F20101130_AAACGG zhang_b_Page_042.QC.jpg
3a9a10e455ff7561d3c0c44effaf9a97
89cc9d3deeceb8e8478ae6b294443e3ef3a8932c
6381 F20101130_AAACFT zhang_b_Page_030thm.jpg
a9c8b6745514432b433a666fc450d8bf
e9c681a42ac8d8e290a82dc32cd9a45a4b4363ff
4198 F20101130_AAACGH zhang_b_Page_042thm.jpg
a79b1f9ec295ade132058dd31233b55f
9d097cb644f75ad65753727b0e27516c1152d932
18366 F20101130_AAACFU zhang_b_Page_032.QC.jpg
c2e6dd7a6f26a51b78c42bb224dae5f7
6ccd9650742651347f37e5d3ac2a5e526d9d1ad8
94040 F20101130_AAABZZ zhang_b_Page_106.jp2
7a243b86261de618e1164c5c58cd2d6e
52345e1f932d0267350f11e42265e0234e4fe54e
17738 F20101130_AAACGI zhang_b_Page_043.QC.jpg
8d5140da25e962d836057eeb08caef7a
58e1fe02299c4640a44f94188f738c33349adce6
5401 F20101130_AAACFV zhang_b_Page_032thm.jpg
277be1b802d5924b84fe08e2fc7108bf
e4505afe71ae3cda28220e5a493097dad8261e7e
5280 F20101130_AAACGJ zhang_b_Page_043thm.jpg
20cc0bbe7244fd602a6ccf9132bd1b07
6382baf4c962c134f39f2114823e02b5914b1235
21785 F20101130_AAACFW zhang_b_Page_033.QC.jpg
5ac97684e2d761cd83bc26e878b73ab0
5e1a3ca61d8c8913616699d281134443000757f6
4850 F20101130_AAACGK zhang_b_Page_044thm.jpg
51e4adca6c68f5a569a67af1fd495db9
f811f4ce47bbc03f427177104dba6f1ed6c21afe
18004 F20101130_AAACFX zhang_b_Page_034.QC.jpg
c6daee6012670ce3237c93ba92218a5d
58f41284be1dc8b50c820dbb1545b5f63a5ca78b
16455 F20101130_AAACHA zhang_b_Page_056.QC.jpg
a9e42de06a519e50a020c4525f81d796
b95b262acbc544dd534211bf774d4a2aab9a9f9e
9615 F20101130_AAACGL zhang_b_Page_045.QC.jpg
c51826c2d97bf241a2f89c40ca3d590d
92ccbcd940a5cbdaec51c72e041cdff5df625c96
5428 F20101130_AAACFY zhang_b_Page_034thm.jpg
8411e5973066dab9bac53031acfa91dc
c55a758a5103ae4c123695c827aa277206e52cdd
18489 F20101130_AAACHB zhang_b_Page_057.QC.jpg
325ea03257fc74094fa8e28658134eae
3e8257d7752935a60d0f9ece2adc069553dd5e05
17737 F20101130_AAACGM zhang_b_Page_046.QC.jpg
be32eee3712333f3a30fd7da3b607860
1fa2d453c6ad05598d875c791ff7b86dd5c2618a
5694 F20101130_AAACFZ zhang_b_Page_035thm.jpg
1a97fb03e55523e6b5dfb661e6e28d52
0a747626ee0bfeaa289a42c4f5399b557ffc2345
5410 F20101130_AAACHC zhang_b_Page_057thm.jpg
e5b88a6f4f495acdab2d13c441d42c46
f0cfaa40abc42995002d2aa8046e4785f77d3fe5
5069 F20101130_AAACGN zhang_b_Page_046thm.jpg
2fd2266b06f7a8c141ab5968f623d38a
629f643babbd0491485e4be2e1061c995ec5f2bd
20994 F20101130_AAACHD zhang_b_Page_058.QC.jpg
b651e4f0de7e0bda02a4b155ad8b3c31
4c64a4aa703630459e9311cf28830a66cc1e702d
5264 F20101130_AAACGO zhang_b_Page_047thm.jpg
b1bd062629708f698054f21dd53fe71e
2a7e5686aed7afc08246e1e0459c9dafe2ae33fa
6243 F20101130_AAACHE zhang_b_Page_058thm.jpg
fa8022b96939465deab8dfdfb57edd53
5dec6abe63a7ea18a4a4f6408c29504236827392
18044 F20101130_AAACGP zhang_b_Page_048.QC.jpg
54408ea28e00a05aea94b8745d506eac
20b51d3a542b0b8bc525daa6c7a95c7863f58f43
13695 F20101130_AAACHF zhang_b_Page_059.QC.jpg
8727f5be65161795275901cccf240b31
8d5e4a111928bf6b03f66329773c60d2e277f0ea
17795 F20101130_AAACGQ zhang_b_Page_049.QC.jpg
176c1676137b77132f927cc79a4fc1c2
4bdbfbe1587effb240ba96d00d996eb0dc7ff30c
18363 F20101130_AAACGR zhang_b_Page_051.QC.jpg
b2783b6c2a9d8e3e0b55a52c5c766a8d
5daa49182aa69eed8b57bfa2368f2b4e2a0302cd
4492 F20101130_AAACHG zhang_b_Page_059thm.jpg
78ab885a13c1a7f00a9ac1282e079b69
cd3343f2f499e3624737fa6f7f465a9a4dc0abce
5596 F20101130_AAACGS zhang_b_Page_051thm.jpg
4373971b6766609872862907905945dc
84f8df7d35a2c6769a645019dfc6f408f735147b
15868 F20101130_AAACHH zhang_b_Page_060.QC.jpg
8473120ba7a8eca571014eda3d6629c2
578028bcfd7bb38765f607775f4971ec1dd21da5
5369 F20101130_AAACGT zhang_b_Page_052.QC.jpg
2745c843f2fb42b3068b1090c22a49bf
30fdf614ce3a1a734dd4c1d2861d382ec493ee73
4488 F20101130_AAACHI zhang_b_Page_060thm.jpg
192fd2b0ba70e6495515dc55386a9635
93a5c5f744e0d890520b2f5bcd4e047f2c304b6f
1943 F20101130_AAACGU zhang_b_Page_052thm.jpg
ac544c2c60501793f6dbb20ac1dc58d0
b54ecd2089090ec9d40f36ff68d81f15a579b6cf
15711 F20101130_AAACHJ zhang_b_Page_061.QC.jpg
85cab4e18dbdf5487ab67f01681ba99a
088c0c295a1b6cf958edcb8b8be1ef697622fb8b
18701 F20101130_AAACGV zhang_b_Page_053.QC.jpg
421e694fc88f81d69aa27c190c4a1314
1d5572b2eaee727b39a354a9fa3ea639911eafda
4713 F20101130_AAACHK zhang_b_Page_061thm.jpg
84e6ae026ccba61e93373b959f0d0a0f
29dc4505799cfe58f8b04e2334dcb9f10bed45f3
5491 F20101130_AAACGW zhang_b_Page_053thm.jpg
ac7d5f3171782c10d806cb8138e91bae
e96ee775f848a68c1d2046c0b28546b4e73e3334
21840 F20101130_AAACIA zhang_b_Page_074.QC.jpg
81a7d13ffb3d23899945416d8ec631f2
6f441f0817797fb6cff6c0979f8e2548c3ad63d9
F20101130_AAACHL zhang_b_Page_062thm.jpg
bd4b7671c5a51241db41ecd73d75a9a3
442ca1e564fc5196bc7174bf565057546eba7a9c
13401 F20101130_AAACGX zhang_b_Page_054.QC.jpg
f021b8a0ead9776ccb4a1a695bfe99e7
e0cbe9c6ef71ba07c46729d56e675387ea5c8883
6327 F20101130_AAACIB zhang_b_Page_074thm.jpg
99020c333a1dd2562f65f5e142511ede
35a813cac3fd5102226e3c0148e889e59ebe9dd6
16105 F20101130_AAACHM zhang_b_Page_064.QC.jpg
56e1bd73ca480f81b5990f798b448642
30dcfeda07df225f760d984ca97ae41ecd600004
4537 F20101130_AAACGY zhang_b_Page_054thm.jpg
9768ff625ed4c8af97a24d39d82ba836
4202765c4897025534cf3f9fad3ef02a6b0abb17
15238 F20101130_AAACIC zhang_b_Page_075.QC.jpg
c36e9769a053855e1d0285bc0c7fe5b9
8aa4edeb39ae9828de1a4980ed64d9841bb37814
5032 F20101130_AAACHN zhang_b_Page_064thm.jpg
07f264777dc515ab29f13d730dad4fc3
701c89839457532a1ff96647943d9067d356fa13
5323 F20101130_AAACGZ zhang_b_Page_055thm.jpg
a8805b7319d066a870a5080302a1b356
d363a96060a2bf023d10c11f01f36a080daa8359
4835 F20101130_AAACID zhang_b_Page_075thm.jpg
6e8121bc592e070825537ebcc6730b9c
fc80bc6645ec64d58f17781e958c93c7e4f6c8d8
5472 F20101130_AAACHO zhang_b_Page_065thm.jpg
48b220109d2d7a4a703abd712f3e08ee
9bb0bd854a43eb3eab719141547abac17437f7c6
3130 F20101130_AAACIE zhang_b_Page_076thm.jpg
8429084c67574a06b283ee3fc6af3ba5
3b6c735a41543301450f9d057a95292147cb3fe0
12144 F20101130_AAACHP zhang_b_Page_066.QC.jpg
aa334db6b72b26d8367756af84b8f037
f202c9b4cb775856958e22b84f108607302e604d
21070 F20101130_AAACIF zhang_b_Page_077.QC.jpg
cd79fa3f71dac92861fcf2a6029323f4
bfcc143756df3e41c1c3b2a2e345b86ad6ef3b37
4060 F20101130_AAACHQ zhang_b_Page_066thm.jpg
318bbed674ebaf0822f04451d4f39844
0f9cf8783a57c1d062ebb9c2093a7dd1dda08011
6001 F20101130_AAACIG zhang_b_Page_077thm.jpg
78c14fc4dcb13e4e8d980c338a0373cc
42d2ac61092915379234ed39d39ef845bbe7ee30
21848 F20101130_AAACHR zhang_b_Page_068.QC.jpg
31b0b9e8dc9145b41a1a4d35bd76d4cb
6b32f1c14880bbd0362e06023b0958a98359c07f
6106 F20101130_AAACHS zhang_b_Page_068thm.jpg
84b195267daaf40d82fc0fb637d2c8bb
9adfdc70a180d0f6a17b9ab36739f9994a552571
23310 F20101130_AAACIH zhang_b_Page_078.QC.jpg
c0613bdf092f850775ab9ce15da443fb
ae3aca8af26468b7bc5189d142618eb543924f11
5854 F20101130_AAACHT zhang_b_Page_069thm.jpg
e7feecaeda4491772e764fd2a128a4c8
df360e42f10f286f38acda3e96e5dddf5ec0d51a
12465 F20101130_AAACII zhang_b_Page_079.QC.jpg
cee6d511406f34f73d1d4ec309b0a517
5528095b71297fcf65af04965d400cee327e618c
14945 F20101130_AAACHU zhang_b_Page_070.QC.jpg
09ccd10fa1e9d2250cdd91825bbd1c6e
123c3fad6164f3bd7018771b0a96e56ed33ede31
3935 F20101130_AAACIJ zhang_b_Page_079thm.jpg
486816d0a07d45e061a4f4a78df41797
4b3e13031293aed7ffce9f5687d4dbef957ba795
16697 F20101130_AAACHV zhang_b_Page_071.QC.jpg
a813ad6e0e7079142ff667bd02396f1e
55e19938beeaedc3c3d0a235920a921169b37fa0
15968 F20101130_AAACIK zhang_b_Page_080.QC.jpg
f092bd1172a70053cafda0e209d80710
6b7d009b70a347072ee5774a861eb20569f9f6e8
5141 F20101130_AAACHW zhang_b_Page_071thm.jpg
633bd07fdc1f4c5c1533042dc9fea4d8
ec7826c055a86c7889fe42b00a91816cc18c588d
4530 F20101130_AAACIL zhang_b_Page_080thm.jpg
689acaa80031580c8a4d4598fb83aa2d
56e3363ed6fd72056382bfe7fa56793081ad37ca
17947 F20101130_AAACHX zhang_b_Page_072.QC.jpg
a0974d54b76fcfd1f9433617fd3aca89
19e044041c8364bb345eccb099bb240f93148a88
4335 F20101130_AAACJA zhang_b_Page_089thm.jpg
ff847fe7c411a6f07151e98cb6c6ef0d
e0464a72c8257c2915addea1ce99fd6ae6c7b867
16955 F20101130_AAACIM zhang_b_Page_081.QC.jpg
a0e957bd9542a356eb89ee78e640e9ef
1ce8a49f56ab9e2d7845309b8cd2c11fabe0d857
5454 F20101130_AAACHY zhang_b_Page_072thm.jpg
67d59a8fb0182baa1b77c42e03a834d0
97774ac051ef611da8f95e9b7a2793e03008eb97
18049 F20101130_AAACJB zhang_b_Page_090.QC.jpg
972fdab189aec6ccfa9e9f2301134760
c7450065afbcffa73cb595a40898a2cabe8f89b3
5384 F20101130_AAACIN zhang_b_Page_081thm.jpg
e4a6723b8f52cd7fad87cea633f678bb
c25d5a764f938a685e6af459dac058213fa18cde
5167 F20101130_AAACHZ zhang_b_Page_073thm.jpg
01e099e0de1e0c60189b1caeb28d2a23
cbaa2272c37996bfc14e64b814e47a3202b0f74a
5196 F20101130_AAACJC zhang_b_Page_090thm.jpg
31441f9ce95c7ddf15de0ac114b70e2b
e32efb84e15f9b279ed6c3d1c2d429b777985253
21192 F20101130_AAACIO zhang_b_Page_082.QC.jpg
61cee6b373e95ec24cb3ca08e5016847
d2da8a2d5761354610e41e3f68ab947e622b5e82
20089 F20101130_AAACJD zhang_b_Page_092.QC.jpg
771108eeb0ad9cde90ed7a248bced35d
dca1af905dbc3a1ed682cc547bbdcab24a9f56d9
5835 F20101130_AAACJE zhang_b_Page_092thm.jpg
3e3c74185c1e2efae700da7fecda7182
c5539e8ace00096548b814c3c4f264c3a749d7df
20451 F20101130_AAACIP zhang_b_Page_083.QC.jpg
6c491b9685692d858b75d52c4364db79
672628a7ea3d9dbd07619a044b63dd0a5b03bfcb
12328 F20101130_AAACJF zhang_b_Page_093.QC.jpg
048c2a5ad1defdeaaf08dff5d30671f4
8945305a9d93cf86c8073f511b7bdbe9f69d9073
5879 F20101130_AAACIQ zhang_b_Page_083thm.jpg
170f3df39b9f94c045901fc15b900539
5e20d7c28170da428e11076c515bae8df034b004
3815 F20101130_AAACJG zhang_b_Page_093thm.jpg
bebe30b503619efebcaabe56f107e6f1
42c1656a91db7ba771c43bfa62f39a4f17a624c4
4637 F20101130_AAACIR zhang_b_Page_084thm.jpg
152190164baab3300da2dca350aaaeb9
b5564b64070a4368411ff1d15139fb2dd8e9a8be
10500 F20101130_AAACJH zhang_b_Page_094.QC.jpg
afd94d6d9d94653f2daef4df50b87923
30f7de00b089cba8bb61af8f83df05a59bb28467
13728 F20101130_AAACIS zhang_b_Page_085.QC.jpg
8768a2cdd9fff9e7525a94df4371cfb6
a42f2a7d3f45e03e230f252442f0075d03be5066
4309 F20101130_AAACIT zhang_b_Page_085thm.jpg
d5613c49ecc1501085a09540f217a5af
1f348d198bdf09b4789381681d807f301b9ab607
3422 F20101130_AAACJI zhang_b_Page_094thm.jpg
14b437dae3cad26b1e3896cd4d62ca0c
a63531f4e5ec824cf8dd7ebc23e251b3a65a2581
11334 F20101130_AAACIU zhang_b_Page_086.QC.jpg
e2deec40f24e10aa5cbe71fed2eb87d1
5b5445b87cbca2f96cd985270c5e61e9658d429f
5082 F20101130_AAACJJ zhang_b_Page_096thm.jpg
cc1c0f9e692d0c73cae81c0ed00956e8
95ab68b377cbf59c3f364de0f3465f7f6f9d718b
3886 F20101130_AAACIV zhang_b_Page_086thm.jpg
64569b7aa054fb3466767e9ca87fa917
ed59ba057d04e836b76501fac7077268b0be71be
24565 F20101130_AAACJK zhang_b_Page_098.QC.jpg
79a68086d6590a661f86752b74190c7b
45fed96c093e45152cd431e8dc79fcc1fc3f3e8b
16467 F20101130_AAACIW zhang_b_Page_087.QC.jpg
0d96f1eceed4c337e2d09256ac50c7ee
8816d9034071b0bef5323aa955cec830aafcac77
16210 F20101130_AAACKA zhang_b_Page_110.QC.jpg
3adaff31ec380a67baa6daf05cc8a74d
73defe75887bf2aa9549e850fecd6e336a40f491
6663 F20101130_AAACJL zhang_b_Page_098thm.jpg
488d19df641e42e54ce020e3922c4fed
91c390aec17f666524c6ea6b77abfa424127ea0b
5186 F20101130_AAACIX zhang_b_Page_087thm.jpg
7c3960cb25facf824dbc3938581a7038
1493a7ec452749254979b8756b7d9fe291a87029
4818 F20101130_AAACKB zhang_b_Page_110thm.jpg
8cd1f124fe22cef2a74a3eb85b2b8d8e
a7acbb5c0529d66525e28c90aeac7e07416e2434
6415 F20101130_AAACJM zhang_b_Page_099thm.jpg
396add22cdf649926e3f5cf85d6cd701
e775531e9ba01a8e45bb3b1ae9d45a29c011c836
16927 F20101130_AAACIY zhang_b_Page_088.QC.jpg
51dee1b84ca9e66dfd41d2d38e4c1fa7
d6f18e3818a84b0318cd850bb92f6a39f6726f86
18167 F20101130_AAACKC zhang_b_Page_111.QC.jpg
b81ccea43749a534821d9ac5977b5d36
6d36aaf6be12c7585fc32d293e17be6b52614a25
21327 F20101130_AAACJN zhang_b_Page_101.QC.jpg
bd3835b2fc5fbde705813cf673c9d77b
ba86c93325af4dd574e82e1e5ffdd2c2315a51c6
5195 F20101130_AAACIZ zhang_b_Page_088thm.jpg
7037ad21fa8049a2cc024f7119d137c7
dddc065e602a917e6c892f0357c74d174827ab5d
21780 F20101130_AAACKD zhang_b_Page_112.QC.jpg
53690ba0233fa6d66c0912a088bfd246
bc36a990c0ee9e8efb41cccb525ab4e70ec304ef
F20101130_AAACJO zhang_b_Page_101thm.jpg
c0673173f3c5eb9fead73347dfd43cbb
29cbc2fb0e21dade090c10854f424bfd03e82314
6234 F20101130_AAACKE zhang_b_Page_112thm.jpg
e7d8d8d48274ebac1d4a0f592c508c39
28ac4ab6d599bb8639a3323ca4e13eef4a4e238f
16059 F20101130_AAACJP zhang_b_Page_102.QC.jpg
cb9f35358c1b984f91bae62284d0132d
f0042749c76ee65ec99a99c3635e285a40a39240
18508 F20101130_AAACKF zhang_b_Page_113.QC.jpg
8b07dd78d2d5391f8fabafeb58669728
43fb6491b40a7063e51442d5075905d879075068
4834 F20101130_AAACJQ zhang_b_Page_102thm.jpg
9fefd428f128bc25c9489d02bd90b0b5
1326bdcb9ea7145f3d1642317cdd3e4d4e1f3384
18941 F20101130_AAACKG zhang_b_Page_114.QC.jpg
331774208cd459bcc9cac6f59ceae5c6
aadfe9591b8967fa2b783d6de461d49e37697220
19741 F20101130_AAACJR zhang_b_Page_103.QC.jpg
cddfb77905db7183448fb2de79a26b47
3c99c3fc64b7b341e707d0ffe00c80ae90a3e5d7
17724 F20101130_AAACKH zhang_b_Page_115.QC.jpg
a41e4091f0a0eead84ea0449ed7e1ef5
61121ca0e330a2b386641b2dd8555361eee881ec
20441 F20101130_AAACJS zhang_b_Page_104.QC.jpg
edd15ec26f01effa115d512e62253fd0
01220faf713b458f7590eb63714ae5c467fd7daf
5512 F20101130_AAACKI zhang_b_Page_115thm.jpg
53570f31fdfb408254862f722c54a7ea
85bf0fe0680c030abb462cd461cb21e1752b7d1c
5727 F20101130_AAACJT zhang_b_Page_105thm.jpg
7b2e29858074c21bd2ca7f17087c80ce
9f28aa0dcf0d7208210fd7cc9b19cda8efe222a5
25591 F20101130_AAACJU zhang_b_Page_106.QC.jpg
f4216ee23c429340543369074dd7bad4
54f216e0bab316f4c949027cf7ac5d47e4db928d
16560 F20101130_AAACKJ zhang_b_Page_116.QC.jpg
532b6f6dbda0e3fefa60027ec2bf4d9e
077364022c7158503edd0ac3f562f03e31808df3
6806 F20101130_AAACJV zhang_b_Page_106thm.jpg
626a45885660648679873a127e902f40
df4b46a46dfdc40b7f2ed53c984acf432f1c6ae9
4853 F20101130_AAACKK zhang_b_Page_116thm.jpg
71386509f70ca458cfefd2873005cd50
32d6b7fa979d8550a97f96d5a2fe2e6b519cd76b
18376 F20101130_AAACJW zhang_b_Page_107.QC.jpg
345b4ec551aafc2aabeb0976d20df555
eb1206adec545d6e37d5ed90e93c0101619e5a83
12596 F20101130_AAACKL zhang_b_Page_117.QC.jpg
b9835bb6cdcccc162f41ff4456c34091
79107bc73d1321fc420912f0c203da02c69c9861
5235 F20101130_AAACJX zhang_b_Page_107thm.jpg
a19cd68fd8be4d0ba88a5e4759e1aca0
a295a7f76f2b302a9b8083a331bd5935ffb32b53
6047 F20101130_AAACLA zhang_b_Page_126thm.jpg
ee1c98cd69907c1b5676058049af0ef6
5f3a82880f55c2595909361dafce0d484b4936ca
4028 F20101130_AAACKM zhang_b_Page_117thm.jpg
7def7a4647ffb80642487e5c46abf2c8
179696a44a8f9218617b71907c7e6bc8238e8ecf
4471 F20101130_AAACJY zhang_b_Page_108thm.jpg
9c67770b421b7fa27e1c5d0a0b001efc
6ec76d6f21330fcc2ed0d013df425684b0b04f24
6632 F20101130_AAACLB zhang_b_Page_127thm.jpg
f0ca8acfe089734147ac3a8a047c264b
a0ecb626da387bc468934302fd5d4bb89546dc96
22233 F20101130_AAACKN zhang_b_Page_118.QC.jpg
775df75535c7c77990598e54051c6aad
fd0571570588ca8ca1af9df28429457ff7d069d5
3095 F20101130_AAACJZ zhang_b_Page_109thm.jpg
ddb04441e17f0ea1cdd1bd8d95bde016
533787ba796f77dc2206740d0eab84a639acc67a
23400 F20101130_AAACLC zhang_b_Page_128.QC.jpg
55e04283dda5a6617360c628a6db03fb
5a0b464e339815ab32769f63f0df89ea8655b06c
3753 F20101130_AAACKO zhang_b_Page_119.QC.jpg
b9ae60e321a5ca6df4e1aec0d0fbd830
0e3b544bc0a9dfe3ea589672b1c391b23c51f176
6764 F20101130_AAACLD zhang_b_Page_128thm.jpg
d5bbb07dac0a35011e95b0ba385eddb5
6f68bb06155cdd144a734e310c3f6f7f603a80cd
1452 F20101130_AAACKP zhang_b_Page_119thm.jpg
c4dd62146274e0a1f13cb6106e3c2e49
5086cd2ce0219862727bd65db49c52e92c545b8e
5289 F20101130_AAACLE zhang_b_Page_129thm.jpg
5fa4fd9413336f1d9ac1235da555fa1b
77d0d37dea7dc9472d3dc9b98a9d67a64e9656fd
8225 F20101130_AAACKQ zhang_b_Page_120.QC.jpg
168e08c321046b197030bc415a53b4f5
a9693fac303fb762405bb8dbb021d3678548b7fa
10388 F20101130_AAACLF zhang_b_Page_130.QC.jpg
7fc8fd6611d82fe54c8e7be61ae1f912
b0b65d2162ca75bb45d39e2ae5f7162c057b79ee
2696 F20101130_AAACKR zhang_b_Page_120thm.jpg
cddba7507e291ed8e79966e45b472501
594a45cecb15310b54497633cfcb02678ad398e2
3111 F20101130_AAACLG zhang_b_Page_130thm.jpg
1fce3598b11426eef9502e2ea8dfcbcb
5f1fca39f20342a22f50203f8cc7eec59954cee9
3092 F20101130_AAACKS zhang_b_Page_121thm.jpg
af984da16ba1c7a72e42bf4a48debbb0
1fcb0778bfb54305c4d6bc8fd4c02a4223005ab6
9473 F20101130_AAACKT zhang_b_Page_122.QC.jpg
f08fb6282bde88ba19debeed9ca44e14
597d7e39771a5d2c5ed1c96faa281c164019e9db
3126 F20101130_AAACKU zhang_b_Page_122thm.jpg
c608189f792833dba8045f0de5b45376
ead041d25751b007413f8300ddcf327f4d532b1b
11499 F20101130_AAACKV zhang_b_Page_123.QC.jpg
2f648b376cdeb4258b63348bd420f89d
3e0e29dee0ccf46f68d89582cb074b181b462449
3568 F20101130_AAACKW zhang_b_Page_123thm.jpg
59e10e6025d2860ba6c9996880005198
35ace06a2b61d1b4e65e57ae7f4154458001bd23
3257 F20101130_AAACKX zhang_b_Page_124thm.jpg
f6821364a883d1e5942fc63ab9a16a03
57815e1b299b3369731175d2e5176dc48c742a89
3243 F20101130_AAACKY zhang_b_Page_125thm.jpg
3f5decf61daef2544fae620da73eef14
2d1114b975dcbb3a61f5c0535c18daa4bdb4e502
21955 F20101130_AAACKZ zhang_b_Page_126.QC.jpg
cc7fd6acc55741485bd7a3cfbb81be52
c9c400d6e2aecaf8ab25d5194027a92ee50dbdae
F20101130_AAABMC zhang_b_Page_107.tif
555674d699775b0c88a0e47b89697d2e
1c892a8f531b342e1149a17ca6f1689d7d19deb6
63331 F20101130_AAABMD zhang_b_Page_006.jpg
561d919f84aa4be13d7e799f87590b81
a72dce327dfbba0260e93eb40c61670420f743cf
4094 F20101130_AAABME zhang_b_Page_031thm.jpg
8f6708111c47072644ee244083a754da
aaadff4c8a038dfae6d7b64d73e4c59552d6e048



PAGE 1

DESIGN AND IMPLEMENTATION OF A 6 DOF PARALLEL MANIPULATOR WITH PASSIVE FORCE CONTROL By BO ZHANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Bo Zhang

PAGE 3

iii ACKNOWLEDGMENTS The author wishes to express his great gr atitude to Dr. Carl D. Crane, III, his dissertation advisor, for sugge sting the topic of this wor k, providing continuous guidance, support and encouragement throughout this re search. Also the au thor would like to acknowledge the efforts of his supervisory co mmittee members: Dr. John C. Ziegert, Dr. Gloria J. Wiens, Dr. A. Antonio Arroyo and Dr. Rodney G. Roberts, whose time and expertise were greatly appreciated. The author wishes to acknowledge the gui dance from Dr. Joseph Duffy; he will be greatly missed. The author would like to thank Dr. Carol Chesney, Shannon Ridgeway, Dr. Michael Griffis, Dr. Bob Bicker and Jean-Francois Kama th for their great help. He also would like to thank his fellow students in the Center for Intelligent Machines and Robotics for sharing their knowledge. This research was performed with funding from the Department of Energy through the University Research Program in Robotics (URPR), grant number DE-FG0486NE37967. Finally, the author gives speci al appreciation to his parents, for their love, patience and support.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... xi CHAPTER 1 MOTIVATION AND INTRODUCTION....................................................................1 Motivation..................................................................................................................... 1 Introduction................................................................................................................... 4 2 BACKGROUND AND LI TERATURE REVIEW....................................................11 Kinematics Analysis...................................................................................................11 Singularity Analysis....................................................................................................13 Statics and Compliance Analysis................................................................................14 Compliance and Stiffness Matrix...............................................................................17 3 PARALLEL PLATFORM DESIGN..........................................................................20 Design Specification...................................................................................................20 Conceptual Design......................................................................................................22 Prototype Design........................................................................................................28 4 SINGULARITY ANALYSIS.....................................................................................34 5 STIFFNESS ANALYSIS OF COMPLIANT DEVICE.............................................41 Simple Planar Case Stiffness Analysis.......................................................................41 Planar Displacement and Force Representation.........................................................46 Stiffness Mapping for a Planar System......................................................................48 Stiffness Matrix for Spatial Compliant Structures......................................................51

PAGE 5

v 6 FORWARD ANALYSIS FOR SPEC IAL 6-6 PARALLEL PLATFORM...............65 Forward Kinematic Analysis for a 3-3 Platform........................................................65 Forward Kinematic Analysis for Special 6-6 In-Parallel Platform............................70 7 RESULTS AND CONCLUSION...............................................................................78 Calibration Experiment for the Force Sensor.............................................................78 Individual Leg Calibration Experiment......................................................................80 Parallel Platform Force/Wrench Testing Experiment.................................................85 Determination of Stiffness Matrix at a Loaded Position............................................95 Future Research........................................................................................................106 APPENDIX MECHANICAL DRAW INGS OF THE PARTS FOR PCCFC..............107 LIST OF REFERENCES.................................................................................................114 BIOGRAPHICAL SKETCH...........................................................................................118

PAGE 6

vi LIST OF TABLES Table page 3-1 Design objective specification.................................................................................21 6-1 Mappings of angles of sp herical four bar mechanism..............................................69 6-2 Parameter substitutions............................................................................................75 7-1 Load cell regression statistics...................................................................................80 7-2 Force-torque Sensor Resolution...............................................................................85 7-3 Repeatability experiment, method 1 (Encoder Counts)...........................................86 7-4 Repeatability experiment, method 2 (Encoder Counts)...........................................87 7-5 Forward analysis......................................................................................................88 7-6 Spring constants.......................................................................................................92 7-7 Numerical experiment 1encoder counts and leg lengths.......................................93 7-8 Numerical experiment 2encoder counts and leg lengths.......................................93 7-9 Numerical experiment 3encoder counts and leg lengths.......................................93 7-10 Numerical experiment 4encoder counts and leg lengths.......................................93 7-11 Numerical experiment 5encoder counts and leg lengths.......................................93 7-12 Numerical experiment 6encoder counts and leg lengths.......................................93 7-13 Numerical experiment 1 wrench............................................................................94 7-14 Numerical experiment 2 wrench............................................................................94 7-15 Numerical experiment 3 wrench............................................................................94 7-16 Numerical experiment 4 wrench............................................................................94 7-17 Numerical experiment 5 wrench............................................................................94

PAGE 7

vii 7-18 Numerical experiment 6 wrench............................................................................95 7-19 Measured leg connector lengths at positions A through F (all units mm).............100 7-20 Displacement of leg connectors from unloaded home position at positions A through F (all units are encoder counts).................................................................101

PAGE 8

viii LIST OF FIGURES Figure page 1-1 Passive compliance device for contact force control.................................................3 1-2 Serial structure........................................................................................................... .4 1-3 In-parallel platform....................................................................................................5 1-4 General parallel platform...........................................................................................9 1-5 3-3 parallel platform...................................................................................................9 2-1 Remote Center Compliance Coupling......................................................................17 3-1 Plan view of octahedr al 3-3 parallel platform..........................................................23 3-2 Special 6-6 para llel platform....................................................................................26 3-3 3-D model for special 6-6 parallel mechanism........................................................27 3-4 Plan view of sp ecial 6-6 platform............................................................................29 3-5 Flow chart of the design process..............................................................................30 3-6 Assembled model for speci al 6-6 parallel platform.................................................32 3-7 Photo of the assembled parallel platform.................................................................33 4-1 Planar singularity......................................................................................................35 4-2 Plan view of a special 6-6 parallel platform.............................................................36 4-3 Plan view of the specia l singularity configuration...................................................37 4-4 Photo of the parallel platform in singularity............................................................39 5-1 Planar in-parallel springs..........................................................................................42 5-2 Planar serially c onnected springs system.................................................................43 5-3 Planar 2 DOF spring.................................................................................................44

PAGE 9

ix 5-4 Planar compliant system...........................................................................................49 5-5 Stiffness mapping for a planar compliant system....................................................50 5-6 3-D Model for special 6-6 parallel mechanism........................................................55 5-7 Top view of the special 6-6 configuration...............................................................55 6-1 3-D drawing of the 3-3 in-parallel platform.............................................................67 6-2 A general spherical four-bar mechanism.................................................................69 6-3 3-3 in-parallel mechanism with three spherical four-bar linkages...........................70 6-4 Special 6-6 platform a nd equivalent 3-3 platform...................................................72 6-5 Definition of angles and .....................................................................................73 6-6 Planar triangle..........................................................................................................74 7-1 Load cell calibration experiment..............................................................................79 7-2 Load cell calibration experi ment opposite direction..............................................79 7-3 Calibration plot for leg-1..........................................................................................81 7-4 Calibration plot for leg-2..........................................................................................82 7-5 Calibration plot for leg-3..........................................................................................82 7-6 Calibration plot for leg-4..........................................................................................83 7-7 Calibration plot for leg-5..........................................................................................83 7-8 Calibration plot for leg-6..........................................................................................84 7-9 Photo of the testing experi ment for the parallel platform........................................85 7-10 Coordinate systems of the parallel platform............................................................90 7-11 Compliant Platform Model.......................................................................................96 A-1 Strip holder.............................................................................................................10 8 A-2 Lower leg part........................................................................................................109 A-3 Upper leg part.........................................................................................................110 A-4 Leg........................................................................................................................ ..111

PAGE 10

x A-5 Top plate.................................................................................................................1 12 A-6 Base plate...............................................................................................................11 3

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DESIGN AND IMPLEMENTATION OF A 6 DOF PARALLEL MANIPULATOR WITH PASSIVE FORCE CONTROL By Bo Zhang August 2005 Chair: Carl D. Crane III Major Department: Mechanic al and Aerospace Engineering Parallel mechanism has been studied for se veral decades. It has various advantages such as high stiffness, high accuracy a nd high payload capacity compared to the commonly used serial mechanism. This work presents the design, analysis, and control strategy for a parallel compliance coupler for force control (PCCFC) based on a parallel platform design. The device is installed on the distal end of an indus trial manipulator to regulate the contact wrench experienced when the manipulator comes into contact with objects in its environment. The parallel mechanism is comprised of a top platform and a base platform that are connect ed by six instrumented complia nt leg connectors. The pose of the top platform relative to the base as well as the external wrench applied to the top platform is determined by measuring the displ acements of the individual leg connectors. The serial robot then moves the PCCFC in orde r to achieve the desired contact wrench at the distal end.

PAGE 12

xii A mechanical system has been fabricated with an emphasis placed on minimizing the system size and minimizing friction at th e joints. Each leg has been calibrated individually off-line to determine connector properties, i.e., spring constant and free length. The spatial compliance matrix of the PCCF C has been studied to better understand the compliant property of the passive manipulat or. The forward analysis for the special 6-6 parallel platform as well as the kinematic control is also studied. The outcome of this research will advance the contact force /torque and pose regul ation for in-contact operations.

PAGE 13

1 CHAPTER 1 MOTIVATION AND INTRODUCTION Motivation With the development of robotics and c ontrol technologies, many tasks can now be done with higher efficiency than ever befo re. Robotic systems can perform extremely high precision operations in automatic assembly lines as well as perform operations in environments that are too dange rous for humans to work in directly. Simple operations can be completed by position control alone and manipulators are capable of high repeatability which makes these operations eas y to control. More complex operations require the robot to be controlled as it mani pulates a work piece which comes into contact with other objects in its environment. The control of forces during in-contact ope rations is a more difficult problem than that which occurs during non-contact operati ons. For example, during an in-contact operation it is necessary to monitor the loads applied externally to the end effector to ensure that the forces and torques applied to the object being manipulated and the objects in the environment do not exceed allowable sp ecified values. The manipulator should provide the appropriate loads to the object to function pr operly without exceeding the external load limits of the object. As an ex ample, the manipulator may need to implant a prism-shaped part into a corresponding hol e of a precision instrument. During the assembly operation, the part should be carefully fitted with minimal collision so as not to damage the component or the instrument.

PAGE 14

2 Industrial manipulators have been in use since the mid 1970s. Traditional serial industrial robots can be positioned and orie nted very accurately and moved along a desired trajectory. However, without some form of force control, any positional misalignment of the manipulator could cause unexpected yielding interaction with other parts if the manipulated end-effector is rigidly positioned. To satisfy these kinds of requirements, one solution is to integrate a force control scheme into the manipulator controller where measured forces are fed back in a closedloop approach. Load cells that are commerci ally available are capable of measuring a spatial wrench. These devices, however, are quite stiff and the response of the manipulator may not be fast enough to prevent high contact forces. The focus of this dissertation is on the creation a nd analysis of a more compliant spatial device that can be incorporated into the control loop of a ma nipulator in order to successfully accomplish the in-contact force control task. Compliance control must be integrated in the robotic control when the manipulator is handling some fragile or da ngerous objects with force regul ation. A properly designed compliance control system has following advantages [Hua98]: 1. Avoiding collision and damage of the object and the environment. 2. Regulating the loads and wrench applied on the manipulator to meet any special external load requirements. 3. Compensating for the inevitable positional inaccuracies that results form rigid position/trajectory control. A simple example is shown in Figure 1-1. A serial pair of actua ted prismatic joints supports a wheel via a two-spring system. The actuated prismatic join ts are controlled so that the wheel could maintain a desired cont act force with a rigid wall [Gri91A, Duf96].

PAGE 15

3 Figure 1-1. Passive compliance device for contact force control The objective of this simple application is to control the contact force between the wheel and the rigid wall when the wheel is ro lling along the surface. The serial pair of actuated prismatic joints suppor ts the body containing points B1 and B2. This body is connected to the wheel by two compliant connectors (1 B C and2 B C ). The actuator drives the end effector body with pure motion in the i and j directions to sustain the desired contact force between the wh eel and the environment and to move the wheel along the surface as specified. This can be accomplishe d given the compliant propriety of the two connectors (spring constants and free lengt hs) and the geometri cal values of the

PAGE 16

4 mechanism ( 1 and 2). The relationship between a ch ange in the actuator positions (1d and2d ) and the change in contact force can be analytically determined. Introduction Most industrial manipulators are serial in st ructure. The serial structure is an openended structure consisting of several links c onnected one after anothe r as shown in Figure 1-2. The human arm is a good example of a se rial manipulator. The kinematic diagrams of most industrial manipulators look simila r to that of the G.E. P-60 and PUMA 560 industrial robots in that most consist of seven links (i ncluding ground) interconnected by six rotational joints using a sp ecial geometry (such as having three consecutive joint axes being parallel or intersecting) These structures are well constructed, highly developed, and are widely used in the i ndustrial applications. Serial manipulators do not have closed kinematical loops and are actuated at each joint along the serial linkage. Figure 1-2. Serial structure Accordingly, each actuator can significantl y contribute to the net force/torque that is applied to the end effector link, either to accelerate this link to cau se it to move or to influence the contact force and torque if th e motion is restricted by the environment. Since motions are provided serially, the e ffects of control and actuation errors are

PAGE 17

5 compounded. Further, with serially connected links, the stiffness of the whole structure may be low and it may be very difficult to re alize very fast and highly accurate motions with high stiffness. Compared to the serial ma nipulator, the parallel ma nipulator is a closed-loop mechanism in which the end-effecter (mobile platform) is connected to the base by at least two independent kinematic chains (see for example Figure 1-3). With the multiple closed loops, it can improve the stiffness of the manipulator because all the leg connectors sustain the payload in a distributive manner as long as the device is far from a singular configuration. The probl em of end-point positioning e rror is also reduced due to no accumulation of errors. Figure 1-3. In-parallel platform Hence this type of manipulator enjoys the advantages of compact, high speed, high accuracy, high loading capacity, and high stiffn ess, compared to serial manipulators. Disadvantages, however, include a limited work space due to connector actuation limits and interference.

PAGE 18

6 Therefore, the in-parallel mechanism is perceived to be a good counterpart and necessary complement to that of the serial mani pulator. It is very useful to combine these two types of structures to utilize both adva ntages and decrease the weaknesses. One of these applications is force control of serial manipulators with an in-parallel manipulator, whose leg connectors are compliant and instrume nted, is mounted at the end of the serial arm. This is the approach that will be taken here in this dissertati on; i.e., a commercially available serial industrial ma nipulator will be augmented by attaching a compliant inparallel platform to the end effector so that contact forces can be effectively controlled. Compared to parallel manipul ators, the direct forward pos ition analysis of serial manipulators is quite simple and straightforwar d and the reverse positio n analysis is very complex and often requires the solution of multiple non-linear equations to obtain multiple solution sets. For the parallel manipul ator, the opposite is the case in that the reverse position analysis is straightforwar d while the forward position analysis is complicated. This kind of phenomenon is usually referred to as seri al-parallel duality. There are similar duality propert ies between serial robots an d fully parallel manipulator with regards to instantaneous kinematics and statics. The first parallel spatial indus trial robot is credited to Pollards five degree-offreedom parallel spray painting manipulator [Pol40] that had three branches. This manipulator was never actually built. In la te 1950s, Dr. Eric Gough invented the first well-known octahedral hexapod with six stru ts symmetrically forming an octahedron called the universal tire-testing machine to respond to the proble m of aero-landing loads [Gou62]. Then in 1965, Stewart published hi s paper of designing a parallel-actuated mechanisms as a six-DOF flight simulator, which is different from the octahedral

PAGE 19

7 hexapod and widely referred to as the Ste wart platform [Ste65]. Stewarts paper gained much attention and has had a great impact on the subsequent development of parallel mechanisms. Since then, much work has been done in the field of parallel geometry and kinematics such as geometri c analysis [Lee00, Hun98, Tsa99], kinematics and statics [Das98A], and parallel dynamics and controls [Gri91B]. Relative works will be discussed in a more detailed manner in the literature review section. In addition to theoretical studies, experi mental works on prototypes of the Stewart platform have also been conducted for studyi ng its property and performance. The fine positioning and orienting capability of the Stew art platform make it very suitable for use in control applications as dexterous wrists and various constructions of Stewart platform based wrists. Because there are different nomenclatur es and definitions used by different researchers in the field of parallel mechan isms, it is quite necessary to define the nomenclature [Mer00] used in current re search. The following terms are defined: 1. Parallel mechanism: A closed-loop mechanism in which the end-effecter (mobile platform) is connected to the base by at le ast two independent kinematic chains. It is also called Parallel Platform or a Parallel Kinematic Mechanism (PKM). 2. In-parallel mechanism: A 6-DOF parall el mechanism with two rigid bodies connected through six identical leg connectors, such as for example six extensible legs each with spherical joints at both e nds or with a spherical joint at one end and a universal joint at the other. 3. Base Platform: The immovable plate of the pa rallel platform. It is also called the fixed platform. 4. Top Platform: The moving body connected to th e base platform via extensible legs. It is also called moving plat form or mobile platform. 5. Legs: The independent kinematic chains in parallel connecting the top platform and the base platform. Also called a connector. 6. Joint: Kinematic connection between two ri gid bodies providing relative motion.

PAGE 20

8 7. PCCFC: Parallel Compliance Coupler for Fo rce Control. A parallel mechanism whose leg connectors are compliant and instrumented. 8. Platform Configuration: The combined positions and orientations of all leg connectors and the top platform. 9. Direct Analysis: Given the kinematic prope rties of the legs, determine the position and orientation of the top platform. Also called the forward analysis. 10. Reverse Analysis: Given the position and or ientation of the top platform, determine the kinematic properties of the leg actuators. 11. M-N PKM: A PKM with M joint connecti on points on the base platform and N joint connection points on the top platform. 12. DOF: An abbreviating for degree of freedom. A generalized parallel platform is shown in Figure 1-4 [Rid04]. In this example the base platform and top platform are connect ed by six extensible legs. The joints connecting the legs to the top platform and base platform are spherical joints, which would introduce an additional tr ivial rotation of legs about their axes. These additional freedoms will not affect the overall system performance and could be eliminated by replacing the spherical joint on one end of each connector with a universal joint. The above parallel platform has six legs connecting the base pl atform with the top platform and there are six separate joint poi nts on the base platform and top platform respectively. This geometric arrangement it is called a 6-6 PKM or -6 platform according to the above nomenclature. The si x joint points on the base platform are not necessarily located on one plane, but it is often the case that all joint points lie on a single plane and are arranged in some symmetric pattern.

PAGE 21

9 Figure 1-4. General para llel platform [Rid02]. Besides the general 6-6 parall el platform, there are some other configurations. One common configuration of a para llel mechanism is the 3-3 parallel manipulator as shown in Figure 1-5. Figure 1-5. 3-3 parallel platform. Base Platfor m To p Pla t for m Le g connecto r Joint

PAGE 22

10 The 3-3 parallel platform also has six conn ector legs, but each leg shares one joint point with another leg on the base platform and similarly on the top platform. The three shared joint points form a tria ngle on both the planar top platform and the planar base platform. A literature review of these types of devices will be presented in the following chapter.

PAGE 23

11 CHAPTER 2 BACKGROUND AND LITERATURE REVIEW Although it was Dr. Eric Gough [Gou62] who invented the first variable-lengthstrut octahedral hexapod in England in the 1950 s, his parallel mechanism, also called the universal tire-testing machine, did not dr aw much public attention. Then in 1965, Stewart presented his paper on the design of a flight simulator based upon a 6 DOF parallel platform [Ste65]. This work had a great impact on subsequent developments in the field of parallel mechanisms. Since th en, many researchers have done much work on parallel mechanisms and both theoretical anal yses and practical applications have been studied. Kinematics Analysis The forward kinematic analysis of parall el mechanisms was one of the central research interests in this field in the 1980s and 1990s. While the reverse kinematics analysis, which is to calculate connector pr operties based on the top platform position and orientation information, is quite straightfo rward, the forward kinematic analysis is comparatively difficult to solve. The problem is to determine the position and orientation information of the top platform (mobile plat form) based on the connect or properties, i.e., typically the connector lengths Usually the legs are composed of actuated prismatic joints that are connected to th e top platform and base platfo rm by a spherical joint at one end and either a spherical or universal joint at the other. The connector properties are generally referred to as the connector length or leg length as this parameter is usually measured as part of a closed-loop feedback scheme to control the prismatic actuator.

PAGE 24

12 It has been observed [Das00]that th e closed-form solution for the forward kinematics problem could be simplified by regul ating the connector joint locations. The simplest case is the 3-3 platform, which is degenerated by grouping the connector joints into three pairs for both the top platform a nd base platform; each pair of connectors shares one joint point. Alt hough it is difficult to design a ph ysical octahedral structure with the concentric double-sphe rical joints that would have sufficient loading capacity and adequate workspace due to collisions of the leg connectors themselves, the forward analysis is the simplest for this class [Hun98]. The forwar d analysis of this device was first solved by Griffis and Duffy [Gri89] w ho showed how the position and orientation of the top platform can be determined with respec t to the base when given the lengths of the six connectors as well as the geometry of th e connection points on the top and base. This solution is a closed-form solution and is based on the analysis of the input/output relationship of a series of three spherical four bar mechanisms. The resulting solution yielded a single eighth degree po lynomial in the square of one variable which resulted in a total of sixteen distinct solution poses fo r a given set of connector lengths. Eight solutions were reflected about the plane fo rmed by the base connector points. The solution technique was extended to solve other special configurations such as the 4-4 and 4-5 platforms with extended methods [L in90, Lin94]. Other methods [Inn93, Mer92] used constraints to solve the problem. First, part of the structure is ignored so that the locus curves of the connector joints could be determined according to the connector and joint properties. Then the removed part toge ther with the angular and distance constraints is added back for further determinations of the solution for th e forward kinematics analysis.

PAGE 25

13 Some theoretical analyses [Rag93, Wen94] found that there are as many as 40 solutions for the forward kinematic equations Hunt and Primrose used a geometrical method to determine the maximum number of the assembly [Hun93]. In 1996, Wampler [Wam96], found that the maximum number of possible configurations for a general Stewart platform is forty by using a continuation method. Singularity Analysis A singular configuration is some special configuration in which the parallel mechanism gains some uncontrollable freedom. For parallel manipulators, there are different types of singularity c onditions based on the analysis of the Jacobian matrix that is formed from the lines of acti on of the six leg connectors [Zha04]. Both Tsai [Tsa99] and Merlet [Mer00] poi nt out in their books that there are 3 types of singularities based on the Jacobian matrix analysis: the inverse kinematic singularities where the manipulat or loses one or more degree s of freedom and can resist external loads in some dire ctions; the direct kinematic singularities where the top platform gains additional degree(s) of freedom and the parallel platform cannot sustain external loads in certa in direction (or just uncontrollable); and th e combined singularities which could exhibit the features of both th e direct kinematic singularities and reverse kinematic singularities. Other researchers [Dan95, Ma91] fo und an interesting phenomenon called architecture singularities, which means that some parallel manipulators with particular configurations exhibit conti nuous motion capabilities in a re latively large portion of the whole workspace with all actuators fixed. Th is kind of singularity should be avoided in the early stage of design.

PAGE 26

14 One of the reasons to study singularities is to avoid or restri ct any singularities from occurring within the effective worksp ace. Dasgupta [Das98B] developed a method to plan paths with good performance in th e workspace of the manipulator for some special cases although there is not solid evidence to prove that the method is applicable to all begin-pose-to-endpose path planning. Statics and Compliance Analysis The forward static analysis is very st raightforward as the end effector (top platform) force can be directly mapped fr om the connector forces, as described by [Fic86]. Duffy [Duf96] perf ormed stiffness mapping analyses when considering the relationship between the twist (instantaneous motion of the top platform) in response to a change in the externally applied wrench. The mapping matrix (compliance matrix) is related to the stiffness properties of the c onnectors as well as the geometric dimension values. The static equilibrium equation for a para llel mechanism with six leg connectors can be written as = { f ; m } = { f1 ; m01} + { f2 ; m02} + + { f6 ; m06} (2-1) where is the applied external wrench applied to the top platform which is comprised of a force f applied through the origin point of the reference coordinate system and a moment m (dyname interpretation). The terms fi and mi, i = 1..6, represen t the pure force that is applied along each of the leg connectors, i.e., fi represents the force applied along connector i and m0i represents the moment of the fo rce along connector i measured with respect to the origin of the reference frame. This equation may also be written as = f{ S ; S0} = f1{ S1 ; S0L1} + f2{ S2 ; S0L2} + + f6{ S6 ; S0L6} (2-2)

PAGE 27

15 where f is the magnitude of the external wrench, { S ; S0} are the coordinates of the screw associated with the wrench, fi represents the magnitude of the force along connector i, and { Si ; S0Li} are the Plcker line coordinates of the line along connector i [Cra01]. This equation can also be written in matrix format as 1 2 1234563 00102030405064 5 6 f f f f f f f SSSSSSS SSSSSSS (2-3) where the 6 matrix formed by the Plcker coordinates of the lin es along the six leg connectors comprises the Jacobian matrix that relates the force ma gnitudes along each leg to the externally applied wrench. A derivative of the above equation will yi eld a relationship between the changes in individual leg forces to the change in the externally applied wrench. Griffis [Gri91A] extended this analysis to show how the change in the externally applied wrench could be mapped to the instantaneous motion of the top platform. According to the static force analysis presented above, it is straightforward to design a compliant parallel plat form to detect forces and torques. Knowledge of the spring constant and free length of each conn ector and measurement of the current leg lengths allows for the computation of the external wrench, once a forward position analysis is completed to determine the curren t pose. The first sensor of this type was developed by Gaillet [Gai83] based on the 3-3 pa rallel octahedral structure. Griffis and Duffy [Gri91B] then introduced the kinestatic control theory to simultaneously control force and displacement for a certain constrai ned manipulator based on the general spatial

PAGE 28

16 stiffness of a compliant coupling. In their th eoretical analysis of the compliant control strategy, a model of a passive Stewart platform with compliant legs was utilized to describe the spatial stiffness of the pa rallel platform based wrench sensor. Compared to the open loop Remote Center Compliance Modules (RCC) that are commercially available (see Figure 2-1), the St ewart-platform-based force-torque sensor can provide additional information about external wrench to assist force/ position control. Dwarakanath and Crane [Dwa 00], and Nguyen and Antrazi [Ngu91] studied a Stewart platform based force sensor, using a LVDT to measure the compliant deflection for wrench detection. Some researchers [Das94, Sv i95] also considered the optimality of the condition number of the force transformation matrix for isotropy and stability problems. It is important that any wrench sens ing device be far from a singularity configuration. Lee [Lee94] defined the problem of closeness to a singularity measure by defining what is known as quality index (QI) for planer in-paralle l devices. Lee et al. [Lee98] extended the definition of quality inde x to spatial 3-3 in-parallel devices. The quality index is the ratio of the determinan t of the matrix formed by the Plcker line coordinates of the six connector legs of the platform in some arbitrary position to the maximum value of the determinant that is possible for the in-parallel mechanism. The index ranges from 0 to 1 and the maximum valu e of 1 refers to the optimal configuration. Lee [Lee00] provides a detailed quality index presentation on several typical in-parallel structures.

PAGE 29

17 Figure 2-1. Remote Center Compliance Coupling Compliance and Stiffness Matrix The theory of screws is a very powerfu l tool to investigate the compliance or stiffness characteristic of a compliant device. Ball [Bal00] first introduced the theory and used it to describe the general motion of a rigid body. He presented the idea of the principal screw of a rigid body, where a wrench applied along a principal screw of inertia generates a twist on the same screw. The principal screw can be found by solving the spatial eigenvector problem. Dimentberg [Dim65] analyzed the static and dynamic properties of a spring suspended system using screw theory. Patterson and Lipkin [Pat90] studied the n ecessary and sufficient conditions for the existence of compliant axes and showed a new classification of general compliance matrix based on the numbers of the compliant axes produced.

PAGE 30

18 Loncaric [Lon87] also did research on unl oaded compliant systems for the spatial stiffness matrix and normal form analysis by using Lie groups. Then Loncaric [Lon91A] used a geometric approach to define and an alyze the elastic system s of rigid bodies and described how to choose coordinates to simp lify the interpretation of the compliance and stiffness matrix. Lipkin and Duffy [Lip88] presented a hybrid control theory for twist and wrench control of a rigid body, and showed that it is based on the metric of elliptic geometry and is noninvariant with the cha nge of Euclidean unit length and change of basis. Duffy [Duf96] analyzed plan er parallel spring systems theoretically. Griffis and Duffy [Gri91B] studied the non-symmetric sti ffness behavior for a special octahedron parallel platform with 6 springs as the c onnectors, and the stiffness mapping could be represented by a 6 matrix [K]. Patte rson and Lipkin [Pat92] also found that a coordinate transformation changes the stiffness matrix into an asymmetric matrix but the eigenvectors and eigenvalues are invariant under the coordinate tr ansformation, and the eigenvalue problems for compliance matrix and stiffness matrix are shown to be equivalent. Lipkin [Lip92A] introduce a geometric decomposition to diagonalize the 6 compliance matrix and the diagonal elements are the linear and rotational compliance and stiffness values. It is also proved that th e decomposition always ex ists for all cases. Another paper [Lip92B] presents several geom etric results of the 6 compliance matrix via screw theory. Huang and Schimmels [Hua00] also studi ed the decomposition of the stiffness matrix by evaluating the rank-1 matrices that compose a spatial stiffness matrix by using

PAGE 31

19 the stiffness-coupling index. The decom position, which is also called eigenscrew decomposition, is shown to be invariant in co ordinate transformation with the concept of screw springs. Roberts [Rob02] points out that the normal s tiffness matrix with a preload wrench is asymmetric. However when only a small twist from the equilibrium pose is considered, the spatial stiffness matrix is a symmetric positive semi-defined matrix and the normal form could be derived from it. Ciblak and Lipkin [Cib94] have shown that for a preload system with relative large deflection from its original pos ition, the stiffness matrix is no longer symmetric. In a more recent study, Ciblak and Lipkin [Cib99] present a systematic approach to the synthesis of Cartesian stiffness by springs us ing screw theory, and shown that a stiffness matrix with rank n can be synthesized by at least n springs. Researchers [Cho02, Hua98, Lon91B, Rob00] also work on synthesis for a predefined spatial stiffness matrix.

PAGE 32

20 CHAPTER 3 PARALLEL PLATFORM DESIGN The parallel mechanism offers a straight forward method to determine the external wrench applied to the top platform based on measured leg lengths and knowledge of the system geometry (location of joint points in top and base platforms), connector spring constants, and connector spring free lengths. Because of this, the Passive Compliant Coupler for Force Control (PCCFC) can be ut ilized in force feedback applications for serial robots by employing it as a wrist elemen t that is between the distal end of the manipulator and the environment. In this ch apter, a prototype for the compliant parallel platform has been designed and tested in order to study the position/wrench control with the PKM device. Design Specification The design process could roughly be composed by 3 stages [Tsa99]: 1. Product specification and planning stage, 2. Conceptual design stage, 3. Physical design stage. During the first stage, the functionalit y, dimension, and other requirements are specified for the development of the produc t. In the second stage, several rough conceptual design alternatives are develope d based on the product specification and the one with best overall performance is chosen to be developed. In the last stage the dimensional and functional design, analysis and optimization, assembly simulation, material selection, and engineering docume ntation are completed. Although the third

PAGE 33

21 stage usually is the most time-consuming stage, the first two stages have great impact on design results and cost control. In the PCCFC specification and planning stage, the products geometrical and functional requirements are specified as follows: SP1: The mechanism should be compliant in nature to measure contact loads; SP2: The mechanism should have 6 degrees of freedom; SP3: The mechanism should be a spatial parallel structure; SP4: The mechanism should be ability to measure loads without large errors; SP5: the structure should have a relatively compact size. As mentioned in the previous chapter, the Stewart platform is a promising structure for such spatial compliant devices. The Stew art platform structure satisfies the SP2 and SP3 criteria. With each connector being co mposed of a prismatic joint and compliant component, the SP1 and SP4 criteria could also be satisfied. One of the main advantages of the Stewart platform is that it can with stand a large payload (compared to serial geometries) with relative compact size. As a result, the dimensi onal requirement SP5 could be met during the conceptual design and ph ysical design stage. The detailed design specifications are presented in Table 3-1. Note that in this table, the direction of the z axis that is referred to is perpendicu lar to the plane of the base platform. Table 3-1. Design objective specification Specification Range Individual Connector Deflection : Magnitude 5mm Maximum perpendicular load: Magnitude 50N Motion range in Z direction: Magnitude 4mm Motion range in X,Y direction: Magnitude 3.5mm

PAGE 34

22 Table 3-1. Continued Specification Range Rotation range about Z axis: Magnitude 6 Rotation range about X,Y axis: Magnitude 4 Position Measuring Resolution: 0.01mm Conceptual Design In the conceptual design stage, a best conceptual design is selected from among alternatives. The 3-3 octahedral structure was the first to be considered since it is the simplest structure of its kind. The mobility of general spatial mechanisms can be calculated from the Grbler Criterion [Tsa99]: j 1 i i) f ( ) 1 n ( M, (3-1) where: M: Mobility, or number of degr ee of freedom of the system. : Degrees of freedom of the space in which a mechanism is intended to function, for the spatial case, = 6 n: Number of links in the mech anism, including the fixed link j : Numbers of joints in a mechanism, a ssuming that all the joints are binary. i f : Degrees of relative motion permitted by joint i. The octahedral 3-3 parallel platform is shown in Figure 3-1. It is the simplest geometry for a spatial parallel platform. Fo r the 3-3 parallel platform, the octahedron has a top platform and a base platform, 6 connector s, and 6 concentric s pherical joints, three on the base platform and three on the top platform. Each conn ector is treated as 2 bodies

PAGE 35

23 connected by a prismatic joint. Thus there are a total of 14 bodies; 2 for each of the 6 legs, the top platform, and the base platform. Each spherical joint has 3 relative degrees of freedom and thus there are a total of 12 joints that each has 3 degrees of relative freedom. Figure 3-1. Plan view of octa hedral 3-3 parallel platform For this case, the mobility equation becomes 12 30 36 78 ) 1 6 ( ) 3 6 ( ) 1 14 ( 6 M6 1 i 12 1 i (3-2) Six out of the 12 degrees of fr eedom are trivial in that e ach leg connector can rotate about its own axis. Eliminating these from c onsideration results in a 6 degree of freedom device. The forward kinematic analysis was first performed by Griffis and Duffy [Gri89]. They discovered that a maximum of sixteen pose configurations can exist for this device (eight above the base platform and eight more reflections below the base platform) when given a set of leg connector lengths. Hunt does a comprehe nsive study of the octahedral

PAGE 36

24 3-3 parallel platform [Hun98]. L ee [Lee94] analyzed the singul ar configurations of this device and defined the problem of closeness to a singularity by de fining what is known as the quality index (QI) for planer in-para llel devices. Lee et al [Lee98] extended the definition of quality index to the spatial 3-3 in-parallel device. The quality index is the ratio of the determinant of the matrix formed by the Plcker line coordinates of the six connector legs of the platform in some arbitrary position to the maximum value of determinant that is possible for the in-paralle l mechanism. This index is a dimensionless value: max| | | | J J (3-3) where J is the matrix of the line coordinates. The quality index has a maximum value of 1 at an optimal configuration that is show n to correspond to the maximum value of the determinant. As the top platform departs from this configuration relative to the base, the determinant would decrease, and at some positions and orientations it can reach zero which indicates an uncontrollable singularity state. As previously stated, the mechanism has been designed such that the determinant of the matrix J will be near the maximum at the unloaded home position. Lee [Lee00] showed that for a 3-3 platform with a top tr iangle of side length a and a bottom triangle of side length b, that the maximum value of th e determinant of J would occur when the top platform was parallel to the base and was rotated as depict ed by the plan view shown in Figure 3-1. The determinant of the matrix J was shown to be given by 3 2 2 2 3 3 3h 3 b ab a 4 h b a 3 3 J (3-4)

PAGE 37

25 where h is the separation distance or height of the top platform with respect to the base. Lee proved that the maximum value for the determinant of J would occur when 221 () 3 haabb (3-5) and 2 ba (3-6) From (3-5) and (3-6), it is easy to obtain 22 bah (3-7) Although the geometric configur ation of the 3-3 parallel platform has the simplest geometry with regards to the forward analysis it is difficult to design and implement in practice due to the pair of concentric spherical joints or other type of universal joints that connect each leg to the base and top platform These kind of concentric joint pairs are not only very difficult to manuf acture but also very hard to have sufficient payload capacity. This configuration also induces unwanted interference be tween moving legs. One other alternative is the 6-3 platform which has six points of connection on the base platform and three points of connec tion on the top platform. While this 6-3 configuration has less concentric joints pair in the base platform, it still has the same difficulty and complicated arrangement of joint connections on the top platform. In order to completely eliminate the need for coincide nt connections, the 6-6 parallel platform was studied. The forward analysis for a general 6-6 platform, however, is very complicated requiring lengthy computation and resulting in a maximum of forty configurations. Griffis and Duffy [Gri93] developed a sp ecial 6-6 platform geometry, however, where the forward analysis was comparable in difficulty to that of the 3-3 platform. As

PAGE 38

26 opposed to a general 6-6 parallel platform, th e special 6-6 platform has six leg connector points in a plane on the top a nd base platforms. The arrangement of these connecting points is such that three conn ecting points of the platform de fine a triangle as the apices and the other three connecting points lie on the sides of this triangle. There are a few configurations of the leg connection: one configuration, known as the apex to apex, has three of the legs attaching to the corners or apices of the triangle of the top platform and the base platform, and the remaining legs are attached to the sides or edges of the platform. The other pref erred configuration is known as apex to midline as shown in Figure 3-2. Figure 3-2. Special 66 parallel platform.

PAGE 39

27 In order to eliminate the six additional ro tational freedoms, the six spherical joints connecting the top platform and leg connectors are replaced by universal joints. Equation (3-1) is utilized here to ca lculate the mobility of the special 6-6 parallel platform 6 30 24 18 78 ) 1 6 ( ) 2 6 ( ) 3 6 ( ) 1 14 ( 6 M6 1 i 6 1 i 6 1 i (3-8) The legs 1-6 are arranged in such a way that each of them is connecting to a corner connecting point on either the t op or base platform while the other connecting point is in the midline of the triangle of the opposing platform. For example, leg 6 is connecting the top platform 19 on the corner connecting point 18 with the base platform 20 on the midline connecting point 12. Griffis noted that this particular configuration is singular Figure 3-3. 3-D model for speci al 6-6 parallel mechanism If the midline connecting points are not exac tly on the middle of the triangle side and for each side of the triangle, and the poi nts are distributed symmetrically, there are two different configurations which are us ually called clockwise configuration and

PAGE 40

28 anti-clockwise (the rotation direction of the top platform when pressed) configuration based on the plan view of the structure. Th ere is no significant analytical difference between the two configurations for the kine matic analysis and c ontrol, so only one configuration is used to bu ild the prototype. One computer generated model of these configurations is s hown in Figure 3-3. Prototype Design After choosing the desired configurati on of the prototype, more detailed and specific information about the parallel platform should be studied in order to build the physical mechanism. One plan view of the configuration of the parallel platform is shown in Figure 3-4. The dimensional relationship between the top/base platform and the free length of the legs was studied by Lee [Lee00] based on the optimal dimensional restriction for the special 6-6 parallel platform stated in the previous section. Th e free lengths of the connectors are calculated by the following equations: (3-9) As shown in Figure 3-4: l: free length of the legs, divi ded into two groups based on th e fact that there are two different values for free length of the legs. a: refers to the functional triangl e edge length of the top form b: refers to the functional triangl e edge length of the base form. : refers to the rotation angle of th e top platform about the z-axis. 2222 22221 3(331)(cos3(21)sin) 3 1 3(cos3(21)sin)(331) 3long shortlhaabb lhaabb

PAGE 41

29 a,b: refer to the distance between adjacent jo int points on the base platform and top platform. Figure 3-4. Plan view of special 6-6 platform The design process is show n in the Figure 3-5 below

PAGE 42

30 Figure 3-5. Flow char t of the design process Consider the requirements of feasibility with regards to dimensional limits for manufacturing parts, encoders available in the market, a nd the need for a continuous working space with no interf erence of moving parts, a 3-D modeling and assembly simulation resulted in the following choices fo r the size of the base and top platforms and the lengths of the two sets of conn ectors at the unloaded home position:

PAGE 43

31 a = 60 mm, b =120 mm, = 28/120 = 0.2333, h = a= 60mm, Lshort = 68.021 mm, Llong = 80.969 mm. The elastic component is a crucial elem ent for the passive compliant parallel manipulator prototype. Preci sion linear springs were selected for use as the elastic component to provide elongation and compression of the leg connectors. After carefully studying and comparing different springs, a grou p of cylindrical precision springs with different spring constants and the same high linear coefficient ( 5%) were selected to use as the elastic component in the legs. The spring rates were respectively 8.755 N/mm, 2.627 N/mm and 0.8755 N/mm (or 50 lb/in, 15lb/in and 5 lb/in). By using different groups of springs, the stiffness matrix of the parallel platform could be changed by replacing the spring and thus the device can be applied to different applications where the expected force ranges are different. The free length variation of each connector is measured by a linear optical encoder. Each leg connector is comprised of two parts that translate with respect to each other and are interconnected by the spring. The encoder read head is attached to one of the leg parts and the encoder linear scale is attach ed to the other. The two leg parts are constrained by a ball spline to ensure that they translate re lative to one another and to help maintain alignment of the encoder read head and linear scale.

PAGE 44

32 The initial free length of each connector is measured by micro-photography with a specially designed clamp appara tus. Since the six connectin g points on the base platform are in the same plane, the centers of the ps eudo-spherical joints ar e also in the same virtual plane. The free length of the connect ors is defined as the distance between the two centers of the joints, which connect the c onnector with the top platform and the base platform. Figure 3-6 shows renderings of the protot ype device. Figure 3-7 presents a photograph of the prototype. A B Figure 3-6. Assembled model for special 6-6 parallel platform. A) Solid model, B) Frame model.

PAGE 45

33 Figure 3-7. Photo of the a ssembled parallel platform

PAGE 46

34 CHAPTER 4 SINGULARITY ANALYSIS In this chapter, a singula rity analysis is presente d and a special singularity configuration for a special 6-6 parallel platform is analyzed. It is important singular ity configurations, which are identified as the case when the Plcker line coordinates of the six le g connectors become linearly dependent, be avoided throughout the effective workspace of th e device. Singularities can occur in both planar and spatial mechanisms. As an example, a planar structure is shown in Figure 4-1. The structure is in a singular configuration when point S is at location (0, 0). At this instant, considering the displacement of point S as the input and the displacement of point D as the output, it is not possible to obtain a user desired velocity for point D, no matter what the velocity of the input is. This c ondition where point S is at (0, 0) is called a singular point as distinct from other simple points, which are also called regular points [Hun78]. For this group of mechanism, a singularity only occurs when the mechanism comes to some discrete special configuration while for all other point s or configurations the mechanism is normal or controllable. Partic ularly for parallel st ructures, it could be said that singularity configurations exist in the workspace at those configurations where the parallel structure gains some uncontrollable freedom.

PAGE 47

35 Figure 4-1. Planar singularity For spatial parallel manipulators, there ar e different types of singularity conditions based on the analysis of the Jacobian matrix th at is formed from the lines of action of the six leg connectors. For example, a parallel manipulator may be able to resist external loads in some directions with zero actuator forces at an inverse kinematical singular configuration. Some manipulat ors can also have a different kind of singularity condition, in which with all the actuators are lock ed, the top platform could still move infinitesimally in some direction. Further, there are combined singul arities that occur for some special kinematic architecture wh en both conditions mentioned above occur [Tsa99]. All the singularity conditions described a bove are temporary and conditional. This means that only when the special geometrica l configuration occurs, the mechanism will function differently from the norm. The special 6-6 parallel platform is considered here and the joint arrangement is shown in Figure 4-2. Lee [Lee00] showed that when the dimensional ratio p=q=1/2, where p and q represent the offset percentage of the joints in the base and top plat form respectively, the determinant of the Jacobian matrix is zero and the platform is in a singularity. He also D S (0, 0)

PAGE 48

36 identified the singularity case that occu rs when the top platform rotates 90 around the Zaxis. Figure 4-2. Plan view of a special 6-6 parallel platform Certain geometries exist, however, that ar e always singular. One planar example is the parallelogram. A parallelogram with four revolute joints is normally unstable i.e., it could not sustain external load s with internal angles unchange d. Compared to the very stable triangle, such special geometries ar e often referred to as unstable as opposed to being described as being in a singular state [Tsa 99]. Examples of unstable spatial mechanisms can also be found. Figure 4-3 shows the particular case that was identified in this research. This special geometry is always unstable, or in a singularity condition. It will be shown that the top platform will have continuous mobility when the lengths of the six leg connectors are fixed. For this pb qa

PAGE 49

37 mechanism, both the top and the base platform s are equilateral triangles. The following notation is used for this speci al singularity configuration: a: the length of the side of the top equilateral triangle, b: the length of the side of the base equilateral triangle, q,p: a dimensionless number in the range of 0 to 0.5 that defines the offset distance of the connection points (for example, the distan ce between points Ec1 and Ec2 is pb and the distance between points B1 and B2 is qa), h: the vertical distance from the geometric cen ter of the base plate to the center of top plate. Figure 4-3. Plan view of the special singularity configuration In order to simplify the process, here the structure will be analyzed only in an initial configuration where the two plates ar e parallel and the relative rotation angle is zero. The analysis process is the same for other positions and orientations. The matrix of the line coordinates along the connectors for this configurat ion is written as qa

PAGE 50

38 (1)(21)(12)(1) 222222 3[(31)]3()3()3[(31)]3[(23)2]3[2(23)] 666666 3333(31)233(23) 666666 (21)(1) 0 22222 33 6 qabapbqabapbqapb aqbbabaapbaqbapb hhhhhh bhbhbhpbhbhpbh bhpbhbhpbhpbh qabp 3333 66666 abqabpabqabpab (4-1) The determinant of the Jacobian matrix is ,(4-2) It was shown by Lee [Lee00] when analyzing the mechanism depicted in Figure 42 that when p=q=0.5, the determinant of the J acobian matrix equals zero and the structure is unstable. For all other values for p and q in the range of 0 to 0.5, the structure is stable. For the structure shown in Figure 4-3, th e result is quite different. Clearly the determinant of J is zero when p=q. Under the constraints 0 q, p 0.5, the term 2()3(1)(1) p qpqpq is clearly nonnegativ e and is zero precisely when p=q=0. Consequently, we have that under the natura l physical constraints 0
PAGE 51

39 A photo of the platform model is shown in Figure 4-4. From th e picture it can be seen that the structure has collapsed due to th e gravity effect until leg interference occurs. The structure could sustain the gravity of the top plate in the shown state only because of collisions of adjacent leg connectors which st ops it from moving downwards further. The top plate could easily be moved upwards with little upwards force applied. Figure 4-4. Photo of the para llel platform in singularity Although singular conditions are typically so mething to be avoided, there is one application where this particular geometry w ould be of use. Platforms that incorporate tensegrity must be in a state that when no exte rnal load is applied to the top platform, the lines of action of the six leg connectors mu st be linearly dependent. Tensegrity based platforms incorporate compliance such as spri ngs to pre-stress certa in leg connectors in tension which will place the other connectors in compression. The internal forces in the

PAGE 52

40 leg connectors can only sum to zero when the lines of acti on are linearly dependent. Thus, the platform geometry discussed here could have particul ar application when tensegrity is incorporated in the platform.

PAGE 53

41 CHAPTER 5 STIFFNESS ANALYSIS OF COMPLIANT DEVICE For a compliant component, such as a spring, the basic property of it is its stiffness. Theoretically, no object is infinitely stiff; the only difference between a so called stiff component and compliant device is that comparing with the compliant component, the stiff component has a relati vely much higher rigidity or stiffness. The spring is the most commonly used compliant component. The stiffness property of a spring is usually quantified by a ratio called the spring consta nt, also known as the el astic coefficient or stiffness coefficient and is used to describe compliant devices. In this chapter, simple analyses for planar compliant devices are introduced together with an example of a compliant 2-D contact force control system. Then the stiffness analysis of a 3-D system is studied using Screw theory. The stiffness matrix for the special 6-6 parallel manipul ator is developed and a numer ical example is provided for the specific design. Simple Planar Case Stiffness Analysis Hookes Law describes the fundamental rela tionship between an external force and the compliant displacement in static equilibrium 0()s F klkll (5-1) where s F is the force exerted on th e spring (or more generally, the compliant component), k is the spring constant, l is the current length of the spring, l0 is the free length, and l is the displacement from the equilibrium position.

PAGE 54

42 Consider two springs that are connected in-parallel, with spring constants1k and2k respectively. An external force is applie d on the right end of th e two springs and the force direction is along the axes of the springs. Figure 5-1. Planar in-parallel springs The force in each spring can be calculated as fs1 = -k1 l1 fs2 = -k2 l2 where in this case l1 = l2 = l The total force applied to the across the two springs can be written as fs = f1 + f2 (5-2) The equivalent mapping of stiffness of the model is derived by: 112212()()s e f klklkkl kl (5-3) where ek is the equivalent stiffness/sp ring constant of the in-par allel connected springs. It is the sum of the individual spring consta nts. This shows that when compliant components are connected in-parallel, the overa ll equivalent stiffness is much higher than the stiffness of any individual compliant component. s f 1 s f 2 s f 1k 2k

PAGE 55

43 Figure 5-2. Planar serially connected springs system Now consider two springs connected in se ries, as shown in Figure 5-2 above. The axes of the two springs are collinear and they are connected end to e nd. The direction of the external force applied on them is also al ong the axes of the springs. The derivation is quite simple, so only the result is provided here as 12()sss f klkll (5-4) where 12111skkk The term s k is the equivalent stiffness/spring c onstant of the serially connected springs. Its reciprocal value is the sum of the reciprocal values of the individual spring constants. This means that when compliant components are connected serially, the overall equivalent stiffness is less than the stiffness of individual compliant component. If 12kkk then 2ek k and if12kk, then1ekk This shows that serially connected components with wi dely different spring consta nts, have an equivalent stiffness, which is more dependent on the component with the smaller spring constant. According to this result, if one compliant component is connected to a relatively stiff bar, the equivalent stiffness of this two-component system is very close to the stiffness of the compliant component. The systems discussed above have one degree-offreedom (DOF). A more general planar case would have two or more DOF. s f 1 s f 2 s f 1k 2k

PAGE 56

44 X Figure 5-3. Planar 2 DOF spring Figure 5-3 shows a planar 2 DOF spring case. In this spring system, two translational springs are connect ed at one end P and grounded se parately at pivot points A and B respectively. Here translational springs behave like prismatic joints in revoluteprismatic-revolute (RPR) serial chains. In the X-Y plane, two such springs form the simple compliant coupling as shown in Figure 5-3. The two-spring compliant coupling system is equivalent to a planar twodimensional spring. The spring is two-dime nsional because two independent forces act in the translational spring, and it is plan ar since the forces remain in a plane. The external force applied at point P is in static equilibrium with the forces acting in the springs. The two-dimensional spring remains in quasi static equilibrium as the point P move gradually. In order to analyze the two-dimensiona l force/displacement relationship, or mapping of stiffness, it is necessary to decompose both the external force and displacements into standard Cartes ian coordinate vectors such that xy f fxfy A P B fx fy K2 K1 2 1 Y

PAGE 57

45 The locations of points A, B, and P, the initial and current lengths of AP, BP, and the angles 12, are all known. The free length of AP is 01l and the free length of BP is02l The spring constants are1k and2k The current length of AP and BP are 1l and 2l respectively. To simplify the e quations, dimensionless parameters 1011ll and 2022ll are introduced. By definition, these two scalar values are always positive, and no negative spring lengths are allowed. When i >1, the corresponding spring is elongated, and if i is less than 1, then it is compressed. The external force applied on the spring system is given by: 12111 12222(1) (1)x yf cckl f sskl (5-5) where cos()iic andsin()iis Differentiating the above equation will result in the following equation x yf x k f y (5-6) where [k] is the mapping of the stiffness of the system which can be written as [Gri91A] 111212111 212212222 121111 1222220 [] 0 (1)0 0(1) kkcckcs k kksskcs ssksc ccksc (5-7) The spring matrix [k] can be written in the form (1)TT iiikjkjjkj (5-8) where [j] is the static Jacobian matrix of the system, [ j ] is the differential matrix with respect to 1 and2 and ik, 1iik are 2 2 diagonal matrices. In general12

PAGE 58

46 as if the two angles are equal, then th e spring matrix expression is singular and meaningless. For that case the two springs are also parall el and equation (5-3) can be applied instead. Planar Displacement and Force Representation Screw theory is utilized in the analysis of more complex compliant systems, such as the structure shown in Figure 1-1. In screw theory [Bal00, Duf96], a line in the XY plane is written using Plcker coordinates as a triple of real numbers {L,M; R}. L and M represent the direction of the line in the XY plane and as such are dimensionless. R represent the moment of the line about the Z axis and has units of length. For this analysis the direction of th e line will be represented by a unit vector and thus 1 22 2()LM =1. The Plcker coordinates of a line are often written as } ; { 0S S s (5-9) where S= (Li + Mj) and S0= (Rk). Although Plcker coordinates are homogeneous, i.e., { S ; S0} describe the same line for all man zero values for it will be assumed throughout this analysis that S is a unit vector. The subscript 0 is introduced to indicate that the moment of the line about the Z ax is will change with a translation of the reference coordinate system. A force is represented by a line multiplied by a force magnitude. A twist is represented by a line multiplied by an angular ve locity. It is interesting to note that a pure moment of magnitude m is written as {0, 0; m} and a pure translation of magnitude v is write as {0, 0; v} which represent lines at infinity that are multiplied by the moment magnitude and the velocity magnitude respectively.

PAGE 59

47 The resultant of an arbitrary set of planar forces fi {Si; S0i}, i=1..n, can always be represented by a wrench which in the planar case reduces to a particular line-bound force which may be calculated as } ; { f } ; { fi 0 i n 1 i i 0S S S S w (5-10) This resultant is often written as } ; { 0m f w (5-11) where f = f S and m0 = f S0. Now, consider that three forces in the XY plane of magnitude 123,, f ff are applied to a rigid body. The equivalent resu ltant force can be determined by the sum of the force vectors as 3 12 112233 3 12 1 000123 23 123wfsfsfs SSS f+f+f SSS off ff mm mm (5-12) Equation (5-12) can be further transf ormed with the following expression 123 1123 12323ccc wfs+fs+fs p pp (5-13) where si and ci represent the sine and cosine of an orientation angle of line i measured with respect to the X axis and pi represent the moment of line i Equation (5-13) may be writt en in matrix form as wj (5-14) where

PAGE 60

48 123 123 123ccc jsss p pp 1 2 3 f f f Given the magnitude of the external forces and their coordinates of each line of action, the coordinates of the resultant force (both magnitude and the line of action) can be determined by the above equation. For stat ic equilibrium [Duf96], there must be one corresponding external force also applied on the rigid body with equal magnitude and opposite direction. Stiffness Mapping for a Planar System The general mapping of stiffness is a one -to-one correspondence that associates a twist describing the relative displacement between the bodies with the corresponding resultant wrench which interacts between them Considering a simple 2-D planar system shown in Figure 5-4, the rigid moving body is connected to ground by three RPR compliant serial chains. A linear spring is located inside the prismatic joint and axial spring forces are applied both on the top movi ng stiff body (to simplify, it is also called top platform) and the ground. According to equation (5-13), the re sultant force can be calculated. In order to maintain equilibri um, an external force with the exact same magnitude f and opposite coordinatesw much be applied on the top platform.

PAGE 61

49 Figure 5-4. Planar compliant system 112233 110112202222022 ()()()wfsfsfs kllskllsklls (5-15) where twist wcontains the magnitude and geometrical information of the resultant force, 1 f ,2 f and 3 f are the magnitudes of the sp ring forces correspondingly, 12 ,ss and 3sare the line coordinate of the axis of the connectors, 1k,2kand 3kare the spring constants and 0(),(1,2,3)iilli is the change in length between the current connector length and the initial length correspondingly. The detailed derivation process can be f ound in [Duf96], and only the result of the stiffness mapping is provided here as wKD (5-16)

PAGE 62

50 where the stiffness matrix [K] is (1)T TKjkjBkC (5-17) where 1 2 300 00 00 k kk k 11 22 33(1)00 [(1)]0(1)0 00(1) k kk k and 0/iiill 123 123 123 ccc jsss p pp 123sss, 123 111123 123BBB BBBsss B sssccc qqq and 123 123123 123CCC CCCsss Csssccc qqq /iBiisdsd are the coordinates of the line that is perpendicular to is that passes though a fixed pivot i B icsare the coordinates of the line iCS that is parallel toiBSand through pivot iC, as shown in Figure 5-5, Figure 5-5. Stiffness mapping fo r a planar compliant system

PAGE 63

51 Stiffness Matrix for Spatial Compliant Structures The analysis of stiffness matrix for the planar structure was presented in the previous sections. Now the spatial stiffness property will be explored In this section, necessary concepts of projective geometry fo r point and line-vector are provided briefly, followed by the analysis of a spatial 3-3 oc tahedron and the desired special 6-6 spatial parallel manipulator. In order to perform the analysis, it is esse ntial to expand some of the concepts from the 2-D plane to 3-D space. In the planar case, the Cartesian coor dinate for a point is defined by two dimensionless scalar values: X and Y, which are also dependent on the reference point of the selected coordina te system. In Screw theory [Cra01], homogeneous coordinates (w; x, y, z) are us ed to describe the location of a point and iii i i x yz r w i j k refers to the position vector from th e reference point O to the point A with Cartesian coordinate(,,)iii x yz. Usually it is assumed that w=1 so that the ratios of x/w, y/w, z/w are equal to x, y and z. This homogeneous coordinate system also allows for a point at infinity. When |w|=0, the point A is at infinity in the direction parallel to () x yz i j k For points not at infini ty, w is always a non-zero value. To simplify the expressions in this work, the coordinate of a point will also be expressed as (x, y, z). It is clear that in the planar case, the equation of a line can be expressed by three numbers L, M, and R, which are first introdu ced by Plcker, and ar e called Plcker line coordinates. In 3-D space, the ray coordina te of a line is defined by any two distinct points on the line. The coordina te of point 1 and point 2 are 1111(,,) x yzr and

PAGE 64

52 2222(,,) x yzr respectively. The vector S whose direction is para llel to the line could be written as: () 21Sr-r (5-18) or LMN Si j k (5-19) So vector S provides directional information fo r the line segment composed by the two points. Now cons ider another vector r from the reference origin to any general point 111(,,) x yzon the line, the cross product of the two vectors defines a new vector which is perpendicular to both of the vectors. This is written as OL SrS (5-20) The vector OLS is the moment of the line about th e origin O and is clearly origin dependent, while the vector S provides the directional information for the line. Thus the coordinate of a line is written as {}OLS;S and this is also called the Plcker coordinates of the line. The semi-colon in the coordi nates indicates that the dimensions of S andOLS are different, i.e., S is dimensionless while S0L has units of length. From equation (5-20), the moment vector OLS is: 111OL x yzPQR LMN ijk Si j k, (5-21) where 11 11 11PyNzL QzLxN R xMyL (5-22) The Plcker coordinates for the line jo ining two points with coordinates 1111;,, x yzand 2221;,, x yzwas expressed by Grassmann by the six 2 determinants of the array:

PAGE 65

53 111 2221 1 x yz x yz (5-23) as 1 21 1 x L x 1 21 1 y M y 1 21 1 z N z and 11 22yz P yz 11 22zx Q zx 11 22 x y P x y The Plcker line coordinates {}{,,;,,}OLLMNPQR S;S are homogeneous since non-zero scalar multiples of all the coordinates still determine the sa me line [Cra01]. The semicolon inside the coordinate separates the (L, M, N) and (P, Q, R) because their dimensions are different. From equation (5-19) the direction ratios (L, M, N) are related to the distance | S | by 2222||LMN S (5-24) It is useful to unitize these directional va lues of these homogeneous coordinates to simplify the application; 212121,, |||||| x xyyzz LMN SSS (5-25) which are also called direction co sines of the lines or unit dir ection ratios. Therefore, the new set of L, M, and N has the new restriction 2221 LMN (5-26) Using the Plcker line coordinates, a force fis expressed as a scalar multiple f S. The reference point is selected in such a way that the moment of the force fabout this reference point, 0m, can be expressed as a scalar multiple OL f Swhere OLSis the moment vector of the line. So the action of the fo rce applied on the body can be expressed as a scalar multiple of the standard Plucker line coordinates and the coordinates of the force are:

PAGE 66

54 {}{,}OLwff 0SS;Sfm (5-27) where {}OLS;Sis the Plucker line coordinates and 1 S. It is clear that fis a line bound vector and coordinate inde pendent, while the moment 0mis origin dependent. For spatial parallel mechanisms, the forward static analysis consists of computing the resultant wrench {}wf;m due to linearly independent fo rces generated in the legs acting upon the moving platform. The resultan t wrench could be simply expressed: 101202303 110122023303 {}{}{}{}... {;}{;}{;}...w fSSfSSfSS f;mf;mf;mf;m (5-28) Equation (5-28) may be written as w j (5-29) where j is the Jacobian matrix of the structure. The columns of the matrix are composed of the Plcker coordinates of the lin es of the leg connector forces as 123 010203... ... SSS SSS j. (5-30) is the column vector with the scal ar of the forces as its elements. 1 2 3... f f f (5-31)

PAGE 67

55 Figure 5-6. 3-D Model for speci al 6-6 parallel mechanism Figure 5-7. Top view of th e special 6-6 configuration

PAGE 68

56 Now consider the proposed special 6-6 para llel passive structure, its 3-D model and plan view are shown in Figures 5-6 and 5-7. This special 6-6 spatial parallel manipulat or has a moveable top platform connected to the fixed platform (ground) by six translati onal springs acting in-par allel. Each leg has a conventional spring as the compliant component and as acting in the prismatic joint of spherical-prismatic-spherical se rial chain. The parallel manipulator with six such compliant legs as well as one top platform a nd one base platform thus has six degrees of freedom. The top platform is connected to the base platform by the initially unloaded compliant coupling which restricts any rela tive spatial motion between the bodies. The corresponding stiffness mapping is thus a one -to-one correspondence that associates a twist describing the relative displacement between the top and base platform with the corresponding resultant wrench prov ided by the springs between them. The basic shapes of the top platform a nd the base platform of this special configuration are equilateral triangles with th e top and base triangle side lengths being a and b respectively while pa and pb are the join ts separation distance between the pivots. The distance h between the geometrical center of the top platform and the geometrical center of the base platform is selected in such a way that it is qualify the following expression based on the analysis in the previous chapters: 2b ha (5-32) The top platform has 6 spherical pivot points1,1,1,1,1,1ABCDEF located along its triangle sides and the base platform has 6 spherical pivot points ,,,,,ABCDEF along its triangle sides. The six springs are connected respectively and in order to simplify the

PAGE 69

57 process of the analysis, it is necessary to establish the geometry of the system. The following are position vectors that define th e directions of the six legs and their magnitudes define the lengths of the six legs: 111211 311411 511611, , lAAOAOAlBBOBOB lCCOCOClDDODOD lEEOEOElFFOFOF (5-33) One coordinate system is attached to th e base platform, which is fixed to ground, with the z axis pointing upwards to the top plat form. The origin of the coordinate system is located right on the geometrical cent er of the base triangle platform. Now considering some external wrench is applied on the top platform and the whole system is in static equilibrium w ith the six spring forces as the moving top platform twists relative to the fixed base platform. Here a wrench w 0f;mcan be thought of as a force f;0acting though the origin togeth er with a general couple 00;m, adding these sets of coordinates together re produces the original wr ench coordinates. Similar to the wrench, a twist (w ritten in axial coordinates) can be thought of as a rotation 0, about a line through the origin toge ther with a general translation X,0, adding these sets of coordinates together reprodu ces the original six twist coordinates: DX, .[Gri91A] In order to maintain the system equilibrium the top platform moves as the external wrench changes. The incremental change of wrench may be written as ; w fm, and the form of the desired stiffness mapping of the compliant structure may be written as

PAGE 70

58 wKD (5-34) where Kis the 6 stiffness matrix, which relates the incremental twist ( ,DX ) of the top platform relative to th e fixed base platform/ground to the incremental change of the wrench. As it is specified in the previous chapters, the geometrical properties of the platform are known: the leg lengths are meas ured, while the dimensional values of the top and base triangles are inva riant and pre-defined. The i ndividual spring constants of the springs are also known. In order to determine the mapping of (5-34), it is necessary to first differentiate (529) [Pig98], which could also be written as 1101 2202 3303 4404 5505 6606() () () () () () kll kll kll w kll kll kll 0f j j m, (5-35) where ikis the individual sp ring constant of the thispring, and 0,iill are the current length and free length of the thispring respectively, so that0()iiikll is the force in thethileg due to the extension of the thispring. The Jacobian matrix j is: 123456 123456 SSSSSS j OASOBSOCSODSOESOFS (5-36) where,iSis the direction cosine of the axial line of the thileg. These direction cosines could be calculated a nd unitized via (5-25):

PAGE 71

59 123456 123456111111 ,,,,,llllllAABBCCDDEEFF SSSSSS The lower three rows of the 6 matrix in (5-36) are the moments of the six lines relative to the referenc e point O. Because OA, OB, OC, OD, OE and OF are all fixed line vectors on the base platfo rm triangle, and are constant, this matrix is solely a function of the direction cosines of the six legs. The differential of equation (5-35) is expressed as 11 22 12345633 12345644 55 66 1101 2202 1234563303 123456() () ( kl kl SSSSSSkl OASOBSOCSODSOESOFSkl kl kl kll kll SSSSSSkll OASOBSOCSODSOESOFS o f m4404 5505 6606) () () () kll kll kll (5-37) Hereil is related toD by the following expression: TlD j (5-38) where j is the Jacobian matrix of the parallel st ructure. The column of this 6 matrix is the line coordinates of the thileg [Gri91A]. Griffis demonstrated that in order to relate iS with D each leg needs to have two derivatives, which are perpendicula r to the axis of that leg at its base point, to describe the iS Because each leg is individually connected to the base platform and top platform via a ball joint and a hook joint, it only possess two degrees of freedom: two

PAGE 72

60 rotations whose rotation axis are mutual perpendicular. Griffis also provides the stiffness mapping analysis for a 3-3 octahedral spatia l platform: the analysis is based on Screw theory and utilizes some geometrical restric tions and configurations of the octahedral structure. Three examples include that two le gs sharing one concentr ic ball joint at one fixed pivot, four legs sharing one triangle side and the pivots are located on the vertexes of the triangles. Although the analysis resu lt is exclusively fo r the 3-3 octahedral mechanism, the conceptual approach is applicab le to the general spatial stiffness analysis. The stiffness mapping of the special 6-6 para llel manipulator is determined by following the same procedure. The general expression of the stiffness matrix for the special 6-6 parallel platform is presented as (1)(1) (1)(1)TTT iiiii TT iiiiKkkk kVkV jjjjjj jj, (5-39) where 123456 123456 SSSSSS j OASOBSOCSODSOESOFS (5-40) 1 2 3 4 5 600000 00000 00000 00000 00000 00000ik k k k k k k 11 22 33 44 55 66(1)00000 0(1)0000 00(1)000 (1) 000(1)00 0000(1)0 00000(1)iik k k k k k k (5-41)

PAGE 73

61 and where0/iiill are dimensionless ratios incorporat ed to simplify the expressions. Further, 123456 123456 SSSSSS j OASOBSOCSODSOESOFS(5-42) and 123456 123456 SSSSSS j OASOBSOCSODSOESOFS, (5-43) wherei S, i Sare all unitized di rection cosine vectors for the derivatives of the iS vector; the three vectors are mutually perpendicular at the pivot point. Next, six unit vectors are defi ned as intermediate variable s describing the directional information of the base platform triangle sides as ,,,,, 123456EAACACECECEA uuuuuu EAACACECECEA. (5-44) iV is given by the follow expression: i i i i iuS V uS (5-45) ii SV and iii SVS (5-46) 112233445566llllll 000000 V VVVVVV (5-47) 111222333444555666llllll 000000 V SVSVSVSVSVSV (5-48) where 0 is a zero vector. Substituting all the necessary components in equation (5-45) yields the global stiffness matrix of the special 6-6 parallel passive mechanism via the stiffness mapping analysis.

PAGE 74

62 It is clear that the stiffness matrix can be written as the sum of five different matrices, the first three matrices are sy mmetric, while the last two matrices are asymmetric, so the overall stiffness matrix is asymmetric. It is obvious that the stiffness matrix is dependent on the selection of th e coordinate system, so by using different coordinate systems, it might be possible to have a shorter or longe r expression, but the change of coordinate system does not change the stiffness properties of the system, such as the rank of the matrix, or the eigenvector s and eigenvalues of the stiffness matrix (which are also called corre sponding eigen-screw of sti ffness and eigen-stiffness) [Gri91B, Sel02]. It is also intuitive that the stiffness property of a given compliant manipulator should not change just by using a different c oordinate system, and neither could it be changed to be a symmetric matr ix by choosing a different coordinate system In the five matrices constituting the global stiffness matrix, only the first matrix does not have the diagonal matrix of (1)iik When the parallel compliant manipulator sustains a relatively small de flection (and a correspondi ngly small twist), and each leg also endures a relatively small amount of axial force, therefore each leg length is close to its initial length. Then because 00 01ii i iill ll all the other four matrices components who contain the diagonal matrix of (1)iik are so close to the zero matr ix that their effects can be neglected and only the first part of the global stiffness matrix is left. With these simplifications, the stiffness matrix is a positive definite symmetric matrix. Consider an example parallel passive plat form with the special 6-6 configuration shown in Figure 5-7) that ha s the following specifications:

PAGE 75

63 a = 60 mm, b =120 mm, p=0.233, 20ik lb/ in = 3.5003 N/mm. The stiffness matrix of the system will be calculated for the case when it is sustaining a small external wrench in static equilibrium and the top platform is slightly deflected from its original position. The corresponding Jacobian matrix of the structure is calculated as -0.65457-0.235290.19761-0.235290.456960.47059 0.149740.40754-0.64174-0.407540.4920 0.741020.882350.741020.882350.741020.88235 -25.67-9.169751.3439.735-25.67-30.566 -44.461-40.588012.35344.46128.235 -13.6j 9116.302-13.69116.302-13.69116.302 (5-49) where each column is the Plcker line coordina tes of the axis for each leg, the upper three numbers are unit less, the lower th ree numbers have the unit of mm. The corresponding stiffness ma trix is calculated as 3.53010022.256242.750 03.53010242.7522.2560 0013.9420044.511 22.256242.7502293000 242.7522.25600229300 0044.511004758.7 K (5-50) In this matrix, the units are also differe nt. The four 3 sub-matrices have the following units / NmmN NNmm Introduce the equation (5-38) into the stiffness mapping (5-34):

PAGE 76

64 TwKDKj l (5-51) For a given set of minute deflections of th e legs, the above equation can be used to calculate the applied external wrench. -2.2912-0.82360.69168-0.82361.59951.6472 0.524141.4265-2.2463-1.42651.72220 2.59383.08852.59383.08852.59383.0885 -89.852-32.097179.7139.0989.852-106.99 -155.63-142.07043.239155.6398.832 -47.92157.061-47w1 2 3 4 5 6.92157.061-47.92157.061l l l l l l (5-52) If the top platform of the special 6-6 pa rallel passive platform is deflected much from its unladed home position, then the compliance matrix can be calculated from equation (5-39).

PAGE 77

65 CHAPTER 6 FORWARD ANALYSIS FOR SPECIAL 6-6 PARALLEL PLATFORM The kinematic forward analysis is very impor tant to control applications. In this chapter, the forward kinematic analysis for pa rallel 3-3 manipulator is introduced. Then a geometrical method, patented by Duffy and Griffis [Gri93], of determining the equivalent 3-3 parallel structure for a special 6-6 configuration mechanism is described. Forward Kinematic Analys is for a 3-3 Platform It is necessary to determine the position a nd orientation of the end effector of the manipulator. Given the geomet rical properties of the para llel platform, including leg length, connector pivots locations and etc., th e forward kinematic analysis determines the relation between the end effector pose and it geometrical properties. As stated in the previous chapters, the forward kinematic analysis is complicat ed compared to the reverse kinematic analysis. This problem is also geometrically equivalent to the problem of finding out ways to place a rigid body such th at six of its given points lie on six given spheres. During the late 80s and early 90s of la st century, many researchers had worked on the forward kinematic analysis of parallel ma nipulators. Many differe nt approaches such as closed-form solutions of special cases, num erical schemes, and analytical approaches were considered for both special cases a nd general 6-6 configur ations. Hunt [Hun98] studied the geometry and mobility of the 33 in-parallel manipulator. Wen and Liang [Wen94] used an analytical approach and solv ed the problem for the 6-6 Stewart platform with a planar base and platform by reducing of the kinematic equations into a uni-variate

PAGE 78

66 polynomial and concluded that the upper bound of the solutions for forward kinematic problems is 40 for this class. Merlet [Mer 92, Mer00] had studied direct kinematic solutions for general 6-6 parallel platform and special cases with additional sensors. Various researchers have stated that for a general 6-6 in-parallel platform and with a given set of fixed leg lengths, it is possible to assemble it in 40 di fferent configurations [Das00, Hun98]. It is not likely that all of the 40 configurations are real and applicable. Due to the difficulty of the forward kinematic analysis for the general 6-6 in-parallel platform, it is worthwhile to consider some special cases with reduced complexity. The simplest form of this class is first analyzed. Although there are different configurations a nd designs for a simple spatial parallel platform, the symmetric 3-3 parallel platfo rm shown in Figure 6-1 is among the most important and widely analyzed structures. This 3-3 parallel platform is also called an octahedron. It has eight triangular faces, six vertices and twelve edges. Four edges are concurrent at each vertex. Every vertex is contained in four faces. The devi ce has all six degrees of freedom, ignoring the trivial freedom of each leg rotating about its own axis. The forward kinematic analysis for the 3-3 in-parallel manipulator was first solved by Griffis and Duffy [Gri89] who showed th at the position and orientation of the top platform can be determined with respect to the base when given the lengths of the six connectors as well as the geometry of the connection points on the top and base. The solution is a closed-form solution and is based on the analysis of the input/output relationship of three spherical four bar m echanisms. The three spherical four bar mechanisms (see Figure 6-3) result in three e quations in the tan-half angle of the three

PAGE 79

67 unknowns x, y, and z. These tan-half angles are simply named x, y, and z. Elimination of the variables y and z resu lts in the following polynomial wh ich contains the variable x as its only unknown [Gri89]. Figure 6-1. 3-D drawing of the 3-3 in-parallel platform 2 124 (6-1) where

PAGE 80

68 222222 332123312331123321 22 3312123311233221 22 3312331112331212 2222 33112332213312 222 331213321332424 244 822 448 242ABaccABcbACaccACcb AEaaccAEacbAEacb AEbbADccbbBCaacc BCacbBCacbBCbb BEaacBEabBD 2112 222 331223312331212 2222 33121231212312 222222222 312312312 331233123312 331233123312 2 1111 2 2222242 2 4 4acbb CEaacCEabCDacbb DEaabbDaaccAcc BcaCacEaa AEbbADccDEaa BCbbBDcaCDac bac bac 22 11111111 22 22222222,0.5, ,0.5,aAxCbDxcBxE aAxBbDxcCxE The coefficients ,...,,1,2,3iiAEi are expressed in terms of known quantities. Their values for the generic spherical four bar mechanism shown in Figure 6-2 may be written as 12413412413441341223124134 12413412413441341223124134 12413412413441341223124134 1234 124134124134413412231241344Ascssscssccccc B scssscssccccc Cscssscssccccc Dss Escssscssccccc where sin()ijijs and cos()iic For this problem, the coefficients are obtained by replacing the vertex as the or igin of the three spherical f our-bar linkages (shown in Figure 6-3) with points o, p, and q respectively. The corresponding mapping of the angles of the spherical four-bar linkages is shown in table 6-1.

PAGE 81

69 Figure 6-2. A general spherical f our-bar mechanism, Griffis [Gri89] Table 6-1. Mappings of angles of spherical four bar mechanism. Origin o p Q Output: 12a qor ops p qr Coupler: 23a ros spt tqr Input: 34a sop tpq rqo Ground: 41a p oq p qo oqp The resulting solution shown above is an ei ghth degree polynomial in the square of one variable which yields a maximum of si xteen solution poses for any given set of connector lengths and base-top tr iangles lengths. Eight soluti ons were reflected about the

PAGE 82

70 plane formed by the base connector points. There can be up to 16 real solutions and therefore, there can be 0,2,4,6 or 8 pairs of real, reflected solutions. Figure 6-3. 3-3 in-parallel mechanism with three spherical four-bar linkages, Griffis [Gri89] Forward Kinematic Analysis for Special 6-6 In-Parallel Platform The configuration of the special 6-6 in-par allel platform is carefully chosen such that the forward kinematic anal ysis can use a method similar to the method for the 3-3 inparallel platform. This section will show the geometric relationship between the special 6-6 in-parallel platform and an equivalent 3-3 platform and how to solve the direct kinematic analysis problem by using this conv ersion. The relationship between a Special 6-6 platform and its equivalent 3-3 platform was discovered by Griffis and Duffy [Gri93] and is presented here for completeness as the fo rward analysis is an important part of the control algorithm for the fo rce control device develope d in this dissertation.

PAGE 83

71 A Special 6-6 platform is defined as one which is geometrically reducible to an equivalent 3-3 platform Figure 6-4 depicts a perspective and plan view of a 6-6 platform where the leg connector points R0, S0, and T0 lie along the edges of the triangle defined by points O0, P0, and Q0 and the leg connector points O1, P1, and Q1 lie along the edges of the triangle defined by the points R1, S1, and T1. The objective here is to determine the distance between the pairs of points O0 R1, O0 S1, P0 S1, P0 -T1 Q0 T1, and Q0 R1. These distances represent the leg connector lengths for an equivalent 3-3 platform. Throughout this analys is, the notation minj will be used to represent the distance between the two points Mi and Nj and the notation minj will represent the vector from point Mi to Nj. Using this notation, the probl em statement can be written as: given: o0o1, p0p1, q0q1, r0r1, s0s1, t0t1 connector lengths fo r Special 6-6 platform o0p0, p0q0, q0o0, o0s0, p0t0, q0r0 base triangle parameters r1s1, s1t1, t1r1, r1o1, s1p1, t1q1 top triangle parameters find: o0r1, o0s1, p0s1, p0t1, q0t1, q0r1 connector lengths for equivalent 3-3 platform Obviously it is the case that s0p0 = o0p0 o0s0 ; t0q0 = p0q0 p0t0 ; r0o0 = q0o0 q0r0 o1s1 = r1s1 r1o1 ; p1t1 = s1t1 s1p1 ; q1r1 = t1r1 t1q1 (6-2 ) The six leg connectors of the equivalent 3-3 platform de fine an octahedron, i.e., there are eight triangular f aces; the top and bottom platform triangles and six faces defined by two intersecting connectors and an e dge of either the base or top platform. The solution begins by first defining angles and which, for each of the six faces

PAGE 84

72 defined by intersecting connectors, defines the angle in the plane between a connector of the Special 6-6 platform and an edge of either the top or base platform as appropriate. A B Figure 6-4. Special 66 platform and equivalent 3-3 pl atform, A) Perspective view, B) Plan view Figure 6-5 shows the angles 1 and 1. The angle 2 is defined as the angle between r0r1 and r0o0 and the angle 3 is defined as the angle between s0s1 and s0o0. Similarly, the O0 O1 P0 P1 Q0 Q1 R1 R0 S1 S0 T1 T0 O0 O1 P0 P1 Q0 Q1 R1 R0 S1 S0 T1 T0

PAGE 85

73 angle 2 is defined as the angle between q1t1 and q1q0 and the angle 3 is defined as the angle between o1s1 and o1o0. Figure 6-5. Defi nition of angles and Figure 6-6 shows a planar triangle that is defined by the vertex points G1, G2, and G3. A point G0 is defined as a point on the line defined by G1 and G2 and the angle is shown as the angle between g0g2 and g0g3. The cosine law for the triangle defined by points G0, G2, and G3 yields c b 2 c b cos2 b 2 2 (6-3) Next consider that the triangle lies in a plane defined by u and v coordinate axes. The coordinates of any point Gi may then be written as (ui, vi). The distance a may then be expressed as P0 Q0 O0 S1 R1 T1 P1 T0 1 1

PAGE 86

74 cos c a 2 c a cos c a 2 a ) cos (sin c ) sin c ( ) cos c a ( ) v v ( ) u u (2 2 2 2 2 2 2 2 2 1 3 2 1 3 2 a (6-4) Figure 6-6. Planar triangle Substituting (6-3) into (6-4) and simplifying gives ) a c a b a c b a b ( b 1 ) c b ( b a c a2 b 2 2 2 2 2 b 2 2 2 2 2 a (6-5) Regrouping this equation gives 222222()()()abbaabcabababcab (6-6) The result of equation (6-6) can be applie d to each of the six side faces of the octahedron defined by the leg connectors of th e equivalent 3-3 platform. The parameter substitutions are shown in Table 6-2. G1 G2 G0 G3 a b c ab u v

PAGE 87

75 Table 6-2. Parameter substitutions G0 G1 G2 G3 a b c a b 1 P1 T1 S1 P0 p1t1 p1s1 p0p1 p0t1 p0s1 2 Q1 R1 T1 Q0 q1r1 t1q1 q0q1 q0r1 q0t1 3 O1 S1 R1 O0 s1o1 o1r1 o0o1 o0s1 o0r1 1 T0 P0 Q0 T1 p0t0 t0q0 t0t1 p0t1 q0t1 2 R0 Q0 O0 R1 q0r0 r0o0 r0r1 q0r1 o0r1 3 S0 O0 P0 S1 o0s0 s0p0 s0s1 o0s1 p0s1 Using Table 6-2 to perform appropriate parameter substitutions in (A-5) will result in six equations that can be written in matrix form as A q = M (6-7) where 6 5 4 3 2 1 2 1 0 2 1 0 2 1 0 2 1 0 2 1 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1M M M M M M r o s o t q r q s p t p 0 p s 0 0 s o 0 r q 0 0 o r 0 0 0 0 t p 0 0 q t o s r o 0 0 0 0 0 0 r q q t 0 0 0 0 0 0 t p s p M q A (6-8) where M1 = (p1t1+p1s1) (p0p12 + p1t1 p1s1), M2 = (q1r1+t1q1) (q0q12 + q1r1 t1q1), M3 = (s1o1+o1r1) (o0o12 + s1o1 o1r1), M4 = (p0t0+t0q0) (t0t12 + p0t0 t0q0), M5 = (q0r0+r0o0) (r0r12 + q0r0 r0o0), M6 = (o0s0+s0p0) (s0s12 + o0s0 s0p0). (6-9)

PAGE 88

76 Then these equations are manipulated su ch that dividing the first rows of A and M by p1t1+p1s1= s1t1, the second rows by q1r1+t1q1= r1t1, and so on to obtain AqM (6-10) where 0 0 0 0 2 1 0 0 0 0 0 2 1 0 0 0 0 0 2 1 0 1 1 1 1 2 1 0 1 1 1 1 2 1 0 1 1 1 1 2 1 00 1 0 0 0 0 0 1 0 0 0 0 0 0 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 p s s o s s o r r q r r q t t p t t r o o s o o q t r q q q s p t p p p s s r r t t w w q q p p M A with 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1p o o s p s o s o s s s r r o s o r o r o w o q q r o r q r q r r t r t q r q t q t q q q p p t q t p t p t t t s s p t p s p s p p Note that 1 , , 0 s r t w q p The matrix A is dimensionless and depends only on the ratios of where the connections occur along the sides of the upper and lower bases. Matrix A is also independent of the shapes of the upper and lo wer triangles. Now all the terms in matrix A are known as they are dimensionless ratios expressed in terms of given distances of the top and bottom platform. All the terms in vector M are also known as they are expressed in terms of given dimensions as well as the square of the lengths of the six leg connectors of the Special 6-6 platform. Thus the square of the lengths of the connectors for the equivalent 3-3 platform may be determined as:

PAGE 89

77 1qAM (6-11) Further more, the determinant of A is: pqwtrs s r t w q p ) 1 )( 1 )( 1 )( 1 )( 1 )( 1 ( ) det( A (6-12) It is interesting to note that the matrix A is singular if and only if the following condition is satisfied: 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0t p t p s o s o r q r q r o r o q t q t p s p s (6-13) One case where this could happen would be if all the middle connector pointes S0, T0, R0, O1, P1, and Q1 are located at the midpoints of the sides of the upper and lower triangles. We can also conclude that for A to be singular, if some of the parameters s r t w q p , , are less than then some of the other parameters must be greater than Thus we immediately conclude th at if all of the parameters ar e less than (as is the case in Figure A-1), then A is non-singular. The same statemen t of course is true if all the parameters are greater than Once the leg length dimensions of the equi valent 3-3 platform are determined, the forward analysis of the 3-3 device as descri bed in Section 6.1 is used to determine the position and orientation of the top platform relative to the base

PAGE 90

78 CHAPTER 7 RESULTS AND CONCLUSION In this chapter, the individual components of the designed manipulator are tested to validate the machine design and theoretical anal ysis presented in the previous chapters. In order to calibrate each legs stiffness, firs t the force sensor is calibrated. An optical encoder measures the compliant displacement of the leg and the lo ad cell records the applied force. The force-displacement relationship is provided for each leg and the results are analyzed. The parallel platform is then assembled for the 6 DOF wrench measuring testing. Several experiments are presented and the results are analyzed. At the end of this chapter, a summary of the work is followed by conclusions and future work suggestions. Calibration Experiment for the Force Sensor In order to get the force/ displacement re lationship of the compliant connectors, the detected force signal and the corresponding displacement signal should be recorded simultaneously during the individua l leg calibration experiment. A load cell is used to measure the axial force applied on the individual leg. The rated capacity of the lo ad cell is 5 lb, (22.246 N) and th e output of the load cell is an analog voltage signal, ranging from -5~+ 5 Vdc. The rated output is 2mV/V20%. It is necessary to calibrate the load cell first to ensure the validity and accuracy of the following experiments results.

PAGE 91

79 0 5 10 15 20 25 -1 0 1 2 3 4 5 6 In p ut Force ( Ns ) Output Voltage (Volts)Load cell linearity calibration experiment plot Figure 7-1. Load cell calibration experiment 0 2 4 6 8 10 12 14 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 Input Force (Ns)Output Voltage (Volts)Load cell linearity calibration experiment plot: Fine plot Figure 7-2. Load cell calibration experiment opposite direction

PAGE 92

80 The load cell output analog signal is conn ected to an A/D port of a multi-functional I/O board and the value of the voltage is recorded with the co rresponding known force, which is provided by a set of standard weights for accurate force calibration. The mapping of the force/voltage relationship is shown in Figure 7-1 and Figure 7-2. Five additional sets of expe rimental data were analyzed and the plots are very similar, so only one plot is shown here. Th e linear equation that relates the applied force to the measured voltage can be written as 1ykx (7-1) where the coefficient k1 is determined to equal 0.00225 Volt/gram. The linear regression statistics are shown in Table 7-1. Table 7-1. Load cell regression statistics Regression Statistics Multiple R 0.999115584 R Square 0.998231949 Adjusted R Square 0.966981949 Standard Error 0.059572297(V) From the table and the figure above, the lo ad cell shows very high linearity and is adequate to be used in the following experiments. Individual Leg Calibration Experiment After calibrating the load cell, it is necessa ry to calibrate the compliant connectors and identify the stiffness property for each of them. During the experiment, the compliant connector is fixed vertically and is attached to the load cell in such a way that the applied external force causing either elongation or compression can be detected properly. The two physical quantities, force and displacement, are detected by the load

PAGE 93

81 cell and the optical encoder. The optical enc oder is attached on the compliant leg both for the calibration experiment and for the designed operation. Axial force is applied on the top of the co mpliant leg with variable magnitude and direction in order to get the stiffness mapping of the leg for both compression and elongation. The resolution of the optical encoder is 1000 counts/inch. The individual plot of the stiffness ma pping for each of the six leg connectors is shown in Figures 7-3 through 7-8. -2 -1 0 1 2 3 4 5 -5 0 5 10 15 Leg 1 Calibration: Force Vs. Displacement X Dis p lacement ( mm ) Y Force (N) Figure 7-3. Calibration plot for leg-1

PAGE 94

82 -2 -1 0 1 2 3 4 5 -6 -4 -2 0 2 4 6 8 10 12 14 Leg 2 Calibration: Force Vs. Displacement X Displacement (mm)Y Force (N) Figure 7-4. Calibration plot for leg-2 -2 -1 0 1 2 3 4 5 -5 0 5 10 15 Leg 3 Calibration: Force Vs. Displacement X Dis p lacement ( mm ) Y Force (N) Figure 7-5. Calibration plot for leg-3

PAGE 95

83 -2 -1 0 1 2 3 4 5 -6 -4 -2 0 2 4 6 8 10 12 14 Leg 4 Calibration: Force Vs. Displacement X Displacement (mm)Y Force (N) Figure 7-6. Calibration plot for leg-4 -2 -1 0 1 2 3 4 5 6 -6 -4 -2 0 2 4 6 8 10 12 14 Leg 5 C alibration: Force Vs. Displacement X Dis p lacement ( mm ) Y Force (N) Figure 7-7 Calibration plot for leg-5

PAGE 96

84 -2 -1 0 1 2 3 4 5 -6 -4 -2 0 2 4 6 8 10 12 14 Leg 6 Calibration: Force Vs. Displacement X Displacement (mm)Y Force (N) Figure 7-8 Calibration plot for leg-6 In the calibration plots, the displacement is measured in mm and force measured in Newtons. In the original data file, the disp lacement at each sample time is recorded as encoder counts and the force is recorded as digitized voltage value. Then the units are converted by the load cell ca libration experiment results together with the following conversion equations: 125.4 453.64.448 inmm lbgram (7-2) Then, the data in the plots is used to build six individual look-up tables of displacement vs. force for each leg. Duri ng operation, the relative displacement is measured and the corresponding force value is obtained from the look-up table. These calibration tables can also be replaced by linea r regression of the data for less-accurate operations, and thus a linear spring constant is assigned to each leg.

PAGE 97

85 Parallel Platform Force/Wrench Testing Experiment The force/torque sensor has been factory ca librated, so only the resolution is listed here. The resolutions for force and torque are shown in the table below Table 7-2. Force-torque Sensor Resolution The parallel platform is attached on the t op of the force/torque sensor as shown in Figure 7-9 and its mass center is carefully balanced such that its weight is directly over the origin of the sensors refe rence coordinate system. This coordinate system is based on the force/torque sensor, so there is no mome nt introduced. This z-direction bias force is easily eliminated in the sampling program, so it will not be discussed further. Figure 7-9. Photo of the testing e xperiment for the parallel platform Fx Fy Fz Tx Ty Tz 0.1 (N) 0.1 (N) 0.2 (N) 5 (N -mm) 5 (N-mm) 5 (N-mm)

PAGE 98

86 The home position refers to the position of th e platform when it is attached with the force/torque sensor and placed horizontally with no other force applied on the top platform or any leg. The home position is also called the initial position/ orientation, and it can be specified by its relative displacem ent from the encoders indexed position. Before use, it is necessary to home the device. This is done by applying some small force/ wrench to the top platform, so each le g pass through its index position. Thus after homing, the current position and orientation can be accurately measured even when it is not in its home position. Repeatability test It is important to know the repeatabi lity of the manipulator. The previous prototype, built by Tyler [Dwa00], did not have good repeatability. Significant friction in the spherical joints was noted when an external force was applied and then removed. This friction impacted the ability of the top platform to return repeatedly to the same home position when external loading was removed. In order to test its repeatability, tw o sets of experiment are performed. 1. Apply random force/wrench to the top pl ate and then remove it, measure the unloaded position/ orientation via the leg lengths measurements 2. Apply a weight of 500 grams on the top plate several times, and compare the measured leg lengths. Table 7-3. Repeatability expe riment, method 1 (Encoder Counts) Test Number Leg 1 Leg 2 Leg 3 Leg 4 Leg 5 Leg 6 #1 -1 0 0 0 0 -1 #2 0 0 -1 1 2 -1 #3 -6 0 -4 -6 1 -3 #4 -5 2 0 -4 3 -1 #5 -3 -3 -1 -1 -2 -1 #6 -2 0 -1 0 2 2 Average -2.83333 -0.16667 -1.16667 -1.66667 1 -0.83333 STD 2.316607 1.602082 1.47196 2.73252 1.788854 1.602082

PAGE 99

87 Table 7-4. Repeatabil ity experiment, method 2 (Encoder Counts) Test Number Leg 1 Leg 2 Leg 3 Leg 4 Leg 5 Leg 6 #1 -200 -64 -5 -103 101 98 #2 -202 -65 -5 -104 101 100 #3 -207 -66 -5 -104 101 102 #4 -197 -63 -4 -100 100 97 #5 -203 -66 -5 -102 101 98 #6 -202 -65 -4 -103 100 101 Average -201.833 -64.8333 -4.66667 -102.667 100.6667 99.33333 STD 3.311596 1.169045 0.516398 1.505545 0.516398 1.966384 The experiment results are shown in Tabl es 7-3 and 7-4. The numbers in the column shows the encoder counts when there is no external load a pplied (for method 1) and when the external load is applied on the platform (for method 2). The overall repeatability for the parallel platform is reasonable. Forward analysis verification The forward kinematic analysis was presen ted in Chapter 5. An experiment is shown here to validate the analysis result and al so to provide for the static analysis process. The parallel platform has some external wrench applied on the top plate. The counts of the relative displacement for each le g is then recorded. Table 7-5 shows the lengths of the six leg connectors, both in units of encoder counts from the unloaded position and in units of mm, with the external wrench applied. The table also shows the calcu lated lengths of the six c onnectors of the equivalent 3-3 platform. After performing the forward an alysis of the equivalent 3-3 platform, a reverse analysis of the special 6-6 platform is conducted to calculate the leg lengths and to verify the forward analysis result.

PAGE 100

88 Table 7-5. Forward analysis Leg 1 Leg 2 Leg 3 Leg 4 Leg 5 Leg 6 Encoder Counts -15 -10 -2 -13 -26 -19 Leg length for special 6-6 (measured) (mm) 80.588 67.746 80.918 67.669 80.308 67.517 Leg length for the equivalent 3-3 platform (mm) 84.595 84.829 84.722 84.145 84.469 80.588 Leg length for the 6-6 platform (reverse analysis) (mm) 80.587 67.746 80.918 67.669 80.309 67.517 From the table, it is clear that the forwar d analysis is very accurate, the leg lengths calculated from the reverse analysis, which ar e also based on the re sults of the forward kinematic analysis, are very close to the leg length values me asured by the optical encoder. The coordinate systems of the parallel plat form are shown in Figure 7-10. The base points are 00,, OP and0Q the origin of the base coor dinate system is at point0O 0P is on the X axis, and 0Q is on the XY plane. This coordinate system is fixed in the base. The top plate corner points are111,, R ST the origin of the top platfo rm coordinate system is at point1 R 1S is on the x axis, and 1T is in the xy plane. After finding the leg lengths of the equiva lent 3-3 platform by the geometrical conversion, the position and orie ntation of the platform can be calculated by the forward kinematic analysis. The detail of this anal ysis is presented in Chapter 6, so only the results are presented here.

PAGE 101

89 For the measured leg lengths listed in Ta ble 7-5, the forward analysis has only 8 solutions.1 Four are above the base and four ar e reflected through the base. The four configurations above the base are discussed he re. Each solution can be represented by a 44 transformation matrix. The first three elemen ts in the fourth column have units of length (mm) and all other elements are dimens ionless. From now on, the units for the element in the transformation matrices are omitted. The four calculated transformation matrices that relate the top a nd base coordinate systems are base top10.51050.10950.852929.2029 0.85980.07180.505552.4117 0.00590.99140.130859.4408 0001 T base top20.50210.86480.000729.7107 0.86480.50200.000752.1185 0.00580.00420.999959.4469 0001 T base top30.49870.86680.006129.9149 0.12110.06270.990752.0006 0.85830.49470.136359.4478 0001 T (7-3) base top40.35260.37570.857080.9928 0.36700.78690.496022.5108 0.86080.48940.13968.1510 0001 T Although all four solutions ar e real and satisfy the geometrical conditions, there is only one solution that corresponds to the actual current c onfiguration. With these transformation matrices, the coordinates of all points of the 3-3 structure can be determined. Then drawings of these confi gurations (or the numeri cal determination of 1 In general, the forward analysis of the special 6-6 platform will yield 16 possible solutions, 8 above the base and 8 reflections through the base. However for the measured set of leg lengths, only 8 real solutions exist, 4 above the base and 4 reflections through the base.

PAGE 102

90 the configuration closest to the most previ ous pose) can be used to choose the correct solution set. Figure 7-10. Coordinate syst ems of the parallel platform Here the geometrical meanings of the 44 transformation matrices are used to make the selection. By using homogeneous coordinates, the 44 transformation matrix can be represented as 00001AA BB A B RP T, (7-4) where 0A BPis the location of the origin of the B c oordinate system measured with respect to the A coordinate system, AAAA BBBB RXYZ, its columns vectors are the orientation of the B coordinate system measured with respect to the A coordinate system. O0 O1 P0 P1 Q0 Q1 R1 R0 S1 S0 T1 T0 X Z Y z x y

PAGE 103

91 So by comparing the columns of the transfor mation matrices, the right solution can be properly selected. The transformation matrix for the home position is given below as base tophome0.50560.8628029.6651 0.86280.5056052.1547 00159.9987 0001 T (7-5) Clearly, the second solution is the correct one since physical ly the Z axis of the top coordinate system remained relatively parallel to the Z axis of the base coordinate system after the loading was applied. For all othe r solutions, the coordinate systems have too much rotation relative to the base coordinates system. Wrench and force testing. After choosing the correct transformation ma trix, the coordinates of all points are known and the Jacobian matrices of the structur e at any instant be ca lculated. Using the equilibrium condition, the follo wing equation can be used to calculate the applied force/ wrench as stated in the previous chapters. w j (7-6) Numerical example: In order to verify the accuracy of the PCCFC, several numerical experiments have been done. For each set of experiment data the external wrench applied on the top platform of the PCCFC is measured by the force/torque sensor on which the device is mounted for this experiment and compared to the calculated wrench based on the sensed leg lengths. The force/torque sensor date is reported in te rms of the sensors coordinate system and the calculated wrench is calculated in terms of the base coordinate system. The wrenches are compared by converting the calculated wrench to the sensor coordinate

PAGE 104

92 system. It is important to note that the for ce/torque sensor has a left-handed coordinate system. That is its Z-axis and X-axis are parallel to the corres ponding axes of the base coordinate system of the PCCFC, but the Y-axes are anti-parallel. The sensor coordinate system is used as the common reference system and the coordinates of the external wrench det ected by the PCCFC are transformed by the following equation: 221111 00{;}{;} SSSSpS (7-7) where superscript 1 and 2 refer to the Screw co ordinates in the former coordinate system and new coordinate system respectively, 1p refers the coordinates of origin of the second coordinate system in terms of the first coordinate system. This transformation will determine the calculated wrench in terms of a right-handed coordi nate system whose origin is at the origin of the sensor coordinate system. The x and z axes will be parallel to the x and z axes of the sensor coordinate system but the y axis will be anti-parallel to that of the sensor coordinate system. For each set of experiment data, 0, SS will be different, but 1p will remain the same as 1[60,203,19]p mm. Several numerical expe riments are conducted here. The spring constants are shown in Table 7-6. Table 7-6. Spring constants Leg_1 Leg_2 Leg_3 Leg_4 Leg_5 Leg_6 Spring Constant(N/mm) 3.0914 3.0469 3.1003 3.1047 2.9757 3.038 Numerical experiments: Six different loads were applied to the top platform. The encoder counts and the corresponding leg lengths for the six cases are shown in Tables 7-7 through 7-12.

PAGE 105

93 Table 7-7. Numerical experiment 1encoder counts and leg lengths Table 7-8. Numerical experiment 2encoder counts and leg lengths Table 7-9. Numerical experiment 3encoder counts and leg lengths Table 7-10. Numerical experiment 4encoder counts and leg lengths Table 7-11. Numerical experiment 5encoder counts and leg lengths Table 7-12. Numerical experiment 6encoder counts and leg lengths Leg_1 Leg_2 Leg_3 Leg_4 Leg_5 Leg_6 Counts -15 -15 -12 -73 -61 8 Leg length(mm) 80.59 67.62 80.66 66.15 79.42 68.20 Leg_1 Leg_2 Leg_3 Leg_4 Leg_5 Leg_6 Counts -76 13 -13 -4 -5 -87 Leg length(mm) 79.04 68.33 80.64 67.90 80.84 65.79 Leg_1 Leg_2 Leg_3 Leg_4 Leg_5 Leg_6 Counts -6 -80 -63 -6 -20 2 Leg length(mm) 80.82 65.97 79.37 67.85 80.46 68.05 Leg_1 Leg_2 Leg_3 Leg_4 Leg_5 Leg_6 Counts 74 -30 -37 -54 -21 -83 Leg length(mm) 82.85 67.24 80.03 66.63 80.44 65.89 Leg_1 Leg_2 Leg_3 Leg_4 Leg_5 Leg_6 Counts 46 -31 -37 -47 -22 -67 Leg length(mm) 82.14 67.21 80.03 66.81 80.41 66.30 Leg_1 Leg_2 Leg_3 Leg_4 Leg_5 Leg_6 Counts 43 -101 -27 -66 59 -35 Leg length(mm) 82.06 65.43 80.28 66.32 82.47 67.11

PAGE 106

94 The corresponding wrench information for each of the six cases is shown in Tables 7-13 through 7-18. The sensor data are shown in the first row, the calculated data from PCCFC in terms of its base coordinate system are shown in the second row. Table 7-13. Numerical experiment 1 wrench Fx(N) Fy(N) Fz(N) Tx(Nmm)Ty(Nmm) Tz(Nmm) F/T sensor -0.1 0.3 -9.9 -56.5 -192.1 -8.5 PCCFC (base coord system) -0.1 -0.4 -10.4 -429.2 821.1 -36.2 PCCFC (Transformed) -0.1 -0.4 -10.4 -60.0 193.7 -14.0 Table 7-14. Numerical experiment 2 wrench Fx(N) Fy(N) Fz(N) Tx(Nmm) Ty(Nmm) Tz(Nmm) F/T sensor 0.1 0 -9.8 -107.3 240.1 0 PCCFC (base coord system) -0.1 -0.1 -10.6 499.7 380.9 -4.8 PCCFC (Transformed) -0.1 -0.1 -10.6 -130.6 -257.4 -3.1 Table 7-15. Numerical experiment 3 wrench Fx(N) Fy(N) Fz(N) Tx(Nmm) Ty(Nmm) Tz(Nmm) F/T sensor -0.2 0.1 -10.5 237.3 -31.1 -5.6 PCCFC (base coord system) -0.2 -0.3 -10.7 -115.8 691.0 -29.8 PCCFC (Transformed) -0.2 -0.3 -10.7 263.9 41.5 -14.9 Table 7-16. Numerical experiment 4 wrench Fx(N) Fy(N) Fz(N) Tx(Nmm) Ty(Nmm) Tz(Nmm) F/T sensor 4.3 -3.6 -11.0 -420.4 -519.7 -189.2 PCCFC (base coord system) 5.3 4.0 -10.8 -674.3 1051.6 163.0 PCCFC (Transformed) 5.3 4.0 -10.8 -379.3 507.7 -223.2 Table 7-17. Numerical experiment 5 wrench Fx(N) Fy(N) Fz(N) Tx(Nmm) Ty(Nmm) Tz(Nmm) F/T sensor 3.1 -2.5 -10.8 -310.7 -401.1 -144.0 PCCFC (base coord system) 3.9 2.8 -10.8 -576.3 951.2 -133.7 PCCFC (Transformed) 3.9 2.8 -10.8 -256.7 379.5 -166.7

PAGE 107

95 Table 7-18. Numerical experiment 6 wrench Fx(N) Fy(N) Fz(N) Tx(Nmm) Ty(Nmm) Tz(Nmm) F/T sensor 1.8 1.9 -9.4 420.8 -254.2 -276.8 PCCFC (base coord system) 2.0 -2.3 -9.9 29.6 794.8 -540.9 PCCFC (Transformed) 2.0 -2.3 -9.9 417.9 239.0 -327.8 The different signs between Fy and Ty of the two set of data are caused by the different coordinate systems. Considering the resolution of the sensor and environmental noise, the results are very good. Determination of Stiffness Ma trix at a Loaded Position In this section the stiffness matrix for th e prototype device when it is at a loaded position will be determined in two ways i.e., analytically and experi mentally, and the two results will be compared. Figure 7-11 show s a model of the compliant platform with coordinate system 0 attached to the base a nd coordinate system 1 attached to the top platform. The constant platform dimensions are given as mm 14 L L L mm 60 L L L mm 28 L L L mm 120 L L L1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0Q T P S O R R T T S S R R Q T P S O O Q Q P P O (7-8) where the notation j iL is used to represent the distance between points i and j. A load was applied to the top platform and the leg lengths were measured as mm 9492 67 L mm 2032 68 L mm 8570 66 L mm 8928 80 L mm 8928 80 L mm 1054 80 L1 0 1 0 1 0 1 o 1 0 1 0T T S S R R Q Q P P O O (7-9)

PAGE 108

96 Figure 7-11. Compliant Platform Model At this position, a forward displacement analysis was conducted and four possible real positions and orientations of the t op platform were determined. During the experiment, the z axis of the top platform rema ined close to parallel to the z axis of the bottom platform and the solution which most closely satisfied this constraint was selected. The 4 transformation matrix that relates the top and base coordinate systems was determined to be 1 0 0 0 5839 58 9996 0 0133 0 0256 0 6518 51 0154 0 5021 0 8647 0 1091 30 0243 0 8647 0 5017 00 1T (7-10) the units of the terms in the first three colu mns are dimensionless and the units of the top three terms of the fourth column are millimeters. At this position, the force in each of the six leg conn ectors was determined. The displacement of each of the legs from the unloaded home position measured in terms of encoder counts was as follows:

PAGE 109

97 2 L 8 L 45 L 3 L 3 L 34 L1 0 1 0 1 0 1 o 1 0 1 0T T S S R R Q Q P P O O (7-11) The resolution of the encoder is 1000 c ounts/inch (39 counts/mm). The spring constant of each leg was previously determined experimentally as m m N 1047 3 k mm N 0469 3 k mm N 0380 3 k mm N 9757 2 k mm N 1003 3 k mm N 0914 3 k1 0 1 0 1 0 1 o 1 0 1 0T T S S R R Q Q P P O O (7-12) The force in each of the legs is thus calculated as 1577 0 6191 0 4724 3 2267 0 2362 0 6697 21 0 1 0 1 0 1 1 0 1 0N f N f N f N f N f N fT T S S R R Q Q P P O Oo (7-13) The external wrench applied at position 1, 1w is calculated as the sum of the individual forces as

PAGE 110

98 7457 2 5602 179 1979 263 9436 4 1829 0 0263 01w (7-14) where the first three terms have units of Newt ons and the last three terms have units of Newton-millimeters. The stiffness matrix for the device at this position will now be computed analytically and experimentally. Analytical Determinatio n of Stiffness Matrix In chapter 5, the spatial stiffness matrix for PCCFC was studied and presented. The spatial stiffness matrix can be calcul ated analytically ba sed on the geometrical dimension values, spring constants of the compliant devices and the current transformation matrix. Equation (5-39) is used to determine the spatial stiffness matrix. For the specified configur ation discussed above, its stiffness matrix can be calculated as 3.05050.04260.023921.406211.79101.96 0.04263.03650.0265205.7821.241180.95 0.0240.026512.101412.15732.3842.647 21.406210.72411.97337772489510396 216.7321.241732.36248986451911169 101.7818aK 0.9342.647102261090618475 (7-15) where stands for analytically determined. As stated in Chapter 5, the four 3 submatrices have the following units /NmmN NNmm

PAGE 111

99 Experimental Determination of Stiffness Matrix An additional external load was used to displace the platform slightly from the position defined by (7-10) (to be referred to as position 1) and the leg lengths were recorded. This procedure was repeated for a to tal of six times and these six cases will be identified by the letters A through F. Fo r each of these cases it was necessary to determine the change in the applied wrench (to be written in ray coordinates) and the twist that describes the motion of the platfo rm from position 1 to the new position (to be written in axis coordinates). The wrench and twist will both be evaluated in terms of the coordinate system attached to the base platform. As previously stated, the stiffness matrix K relates the change in the applied wrench to the instantaneous mo tion of the top platform as D K w ] [ (7-16) By evaluating the change in the wrench and the twist six times, (7-16) can be written in matrix format as ] [ ] [ ] [6 6 6 6 D K w (7-17) where ] [6 6 w and ] [6 6D are 6 matrices defined as ] [ ] [F 1 E 1 D 1 C 1 B 1 A 1 6 6 w w w w w w w (7-18) ] [ ] [F 1 E 1 D 1 C 1 B 1 A 1 6 6 D D D D D D D (7-19) The columns of ] [6 6w are the changes in the extern al wrench as the platform moved from position 1 to position A, position 1 to position B, etc. The columns of ] [6 6D are the twists that represent the motion of the platform from position 1 to position

PAGE 112

100 A, from position 1 to position B, etc. Once the matrices ] [6 6 w and ] [6 6D are determined, the stiffness matrix can be calculated as 1 6 6 6 6] ][ [ ] [ D w K (7-20) Prior to calculating ] [6 6w and ] [6 6D it is necessary to document the measured leg lengths at the six new poses of the top plat form. Table 7-19 presents this information. Table7-19. Measured leg connector lengt hs at positions A through F (all units mm) 1 0O OL 1 0P PL 1 0Q QL 1 0R RL 1 0S SL 1 0T TL A 80.5118 79.3688 79.9530 66.9586 69.5748 68.3048 B 79.9530 80.2578 80.9182 66.9840 67.3142 68.0508 C 79.9784 80.7912 80.5372 66.9840 68.2032 67.1364 D 80.4102 81.3246 81.3500 66.4760 67.5428 67.7206 E 81.2738 79.6228 80.6134 66.5014 69.2192 67.7206 F 79.3434 80.4356 82.2390 67.4158 68.0762 67.4412 The position and orientation of the top platform was determined via a forward displacement analysis. The transforma tion matrices were determined as 1 0 0 0 5063 58 9995 0 0222 0 0225 0 0301 50 0084 0 4998 0 8661 0 4078 30 0305 0 8659 0 4993 0 1 0 0 0 7887 58 9997 0 0064 0 0217 0 0331 52 0154 0 5099 0 8601 0 7774 31 0165 0 8602 0 5097 0 1 0 0 0 6251 58 9996 0 0181 0 0222 0 4777 .52 0109 0 4782 0 8782 0 4274 30 0265 0 8781 0 4778 0 1 0 0 0 6024 58 9997 0 0036 0 0259 0 4415 51 0242 0 5020 0 8645 0 0147 30 0099 0 8648 0 5019 0 1 0 0 0 6874 58 9997 0 0239 0 0061 0 5888 51 0068 0 5040 0 8637 0 0674 30 0237 0 8633 05041 0 1 0 0 0 5788 58 9997 0 0072 0 0241 0 0055 51 0164 0 5394 0 8419 0 8315 30 0191 0 8420 0 5391 00 0 0 0 0 0 T T T T T TF E D C B A (7-21) The matrix ] [6 6w is relatively simple to calcula te. For example, the wrench at position 1, 1 w, is calculated as the sum of the fo rces along the leg connector lines at

PAGE 113

101 position 1. Similarly, th e wrench at position A, A w, is also calculated as the sum of the forces along the leg connector lines at position A. The change in the wrench, A 1w is simply calculated as the difference between these two wrenches as 1 A A 1 w w w (7-22) Similar calculations are performed for the other five legs. The displacements of each leg measured in encoder counts from the unloaded home position are listed in the table below. Table 7-20. Displacement of leg connectors from unloaded home position at positions A through F (all units are encoder counts) 1 0O OL 1 0P PL 1 0Q QL 1 0R RL 1 0S SL 1 0T TL A -18 -63 -40 -41 62 12 B -40 -28 -2 -40 -27 2 C -39 -34 -17 -40 8 -17 D -22 14 15 -60 -18 -11 E 12 -53 -14 -59 48 -11 F -64 -21 50 -23 3 -22 The calculated wrenches at each of the six positions are presented as 0032 329 4282 212 7469 125 0937 5 9762 4 3890 0 7374 47 3909 600 4532 420 8078 4 0262 2 8796 5 9454 88 5356 156 3332 257 7095 5 1190 0 6523 0 1008 63 1178 465 3918. 370 1272 8 5456 0 3539 0 4402 21 1519 363 6900 223 4476 8 4166 0 3449 0 8671 138 2140 508 9854 433 8295 4 2219 2 7730 4 F E D C B Aw w w w w w (7-23)

PAGE 114

102 where the first three components of these vector s have units of Newtons and the last three components have units of Newton-millimeters. From (7-21) and (7-14) the matrix ] [6 6w is calculated as 75 331 483 50 691 91 847 65 186 24 12 136 868 41 82 429 025 14 56 294 59 192 65 337 94 388 26 157 8646 5 19 107 508 39 79 170 1501 0 1358 0 7659 0 1836 3 5041 3 1141 0 1591 5 8434 1 0639 0 7285 0 5994 0 0391 2 3628 0 8533 5 6785 0 3277 .0 3712 0 7467 4 ] [6 6w (7-24) The determination of the twists as the top platform moves from position 1 to each of the six other positions is slightly more complicated. The determination of the twist A 1D will be presented. The twist A 1D is comprised of six components. When written in axis coordinates, the last three repres ent the instantaneous angular velo city of the moving body and the first three represent the instan taneous linear velocity of a poi nt in the moving body that is coincident with the origin of the refere nce frame. In this case, both of these instantaneous quantities must be approximated from the finite motion of the platform as it moves between pose 1 and pose A. The angular velocity was approximated by cal culating the axis and angle of rotation that would rotate the top coordi nate system from its orientatio n at pose 1 to its orientation at pose A. This was accomplished by first evaluating the net rotation matrix that relates the A pose to the 1 pose as R R R0 A T 0 1 1 A (7-25)

PAGE 115

103 where the 3 rotation matrices R0 1 and R0 A are obtained as the upper left 3 elements of the transformation matricesT0 1 and T0 A respectively. For this case the rotation matrix R1 A was evaluated as 0000 1 0049 0 0020 0 0050 0 999 0 0437 0 0018 0 0438 0 999 01RA (7-26) From [Cra98] the angle of ro tation and axis of rotation th at would rotate coordinate system 1 to be aligned with coordi nate system A are calculated as =2.526 degrees and 1m = [-0.1124, 0.0431, 0.9927]T. The superscript associated with the axis vector m is used to indicate that the rotation axis that was determined from (7-26) was expressed in terms of the coordinate system attached to th e top platform at position 1. The direction of the rotation axis can be expressed in the ground, 0, coordinate system via m R m1 0 1 0 (7-27) Evaluating the axis of rotation in the 0 coordinate system yielded 0m = [-0.0433, 0.1342, 0.9900]T. Finally, the last three components of A 1 D are approximated by ( 0m) and will be dimensionless (radians). As previously stated, the first three components of A 1D represent the instantaneous linear velo city of a point in the moving body that is coincident with the origin of the reference frame. This quantity is estimated by first determining the coordinates of the origin point of the ground, 0, coordinate system in the 1 coordinate system, i.e., 1P0orig. This quantity is determined as orig 1 0 T 0 1 orig 0 1P R P (7-28)

PAGE 116

104 where orig 1 0P is obtained as the top three terms of the fourth column of T0 1 which was numerically determined in (7-10). For infini tesimal motion, the coordinates of the origin of the ground coordinate system should be th e same when measured in the 1 or the A coordinate systems. Thus it is assumed that orig 0 1 orig 0 AP P where the notation 0'orig is introduced. This point is then transf ormed to the 1 coordinate system as orig 0 A 1 A orig 0 1P T P (7-29) where T T T0 A 1 0 1 1 A (7-30) The translation of the origin point of coor dinate system 0 as seen from coordinate system 1 can be written as the net displacement of the point as orig 0 1 orig 0 1 orig 0 1P P v (7-31) Lastly, this translation vector can be ev aluated in the 0 coordinate system as orig 0 1 0 1 o 0v R v (7-32) For this particular case, the translation of the point in the top platform that is coincident with the origin of the reference coordinate system, for the case where the top platform moves from pose 1 to pose A, was evaluated as mm 0052 0 6462 0 7223 0o 0 v (7-33) Thus for the case where the top platform moves from position 1 to position A, the instantaneous twist written in axis coordinates is estimated as

PAGE 117

105 04364 0 00592 0 00191 0 0052 0 6462 0 7223 0 0 o 0 A 1m v D (7-34) where the first three components have units of mm and the last three components have units of rad. The process can be repeated to determine the twists associated with the motion from pose 1 to pose B, pose 1 to pose C, pose 1 to pose F. The resulting twists were determined as 0273 0 0025 0 0053 0 0411 0 8259 0 3183 0 0001 0 0144 0 0088 0 0185 0 2102 0 0944 0 0027 0 0006 0 0222 0 1035 0 6291 0 04179 0 1 1 1 D C BD D D (7-35) 0025 0 0062 0 0071 0 0776 0 6217 1 2987 0 0091 0 0079 0 0001 0 2048 0 3813 0 6683 1 1 1F ED D Lastly, substituting these values into (7-19) and then substituting (7-19) and (7-24) into (7-17) yields the stiffness matrix for the device at pose 1. This matrix was determined as

PAGE 118

106 2.90450.04991.781267.167136.5397.152 0.02392.96370.8046167.0556.028181.38 0.20850.030811.182364.17693.5332.81 8.5556205.44326.45293902112210796 189.5329.633524.65144595573310174 98.136179.K 7118.616113959997.318204 (7-36) Comparison of analytical res ult and experimental result Comparing the stiffness matrix determined analytically with the stiffness matrix calculated experimentally, it is clear that bo th matrices are non-symmetric and they are very close to each other. This result also vali dates the previous stiffness analysis as well as the force/ twist approximation. Future Research The passive parallel platform has been de signed, fabricated, and tested. Future experiments should be applied to test its dynamic properties such as the effects of vibration, velocity, and acceler ation. The performance of kinematic control depends not only on the stiffness matrix of the complia nt component but also on the whole robot system. The system stiffness is dependent on the configuration of the robot system, including the serial i ndustrial robot, passive parallel platform, a gri pper or tool, and any permanently attached compliant devices. Th e ability of accurately modeling the system stiffness ensures the control perfor mance for displacement and force. There are ways to improve the design of the PCCFC. The theoretical analysis presented in this work can be applied dire ctly to the future project. A non-contact displacement laser sensor can be used to re place the optical encoder to greatly reduce the size of the structure. Also a plastic or rubber cover can be designed to protect the connectors for use of the device in a far ranging set of applications.

PAGE 119

APPENDIX MECHANICAL DRAWINGS OF THE PARTS FOR PCCFC

PAGE 120

108 Figure A-1. Strip holder

PAGE 121

109 Figure A-2. Lower leg part.

PAGE 122

110 Figure A-3. Upper leg part

PAGE 123

111 Figure A-4. Leg

PAGE 124

112 Figure A-5. Top plate.

PAGE 125

113 Figure A-6. Base plate.

PAGE 126

114 LIST OF REFERENCES [Bal00] Ball, Robert Stawell, Treatise on the Theory of Screws, Cambridge University Press, Cambridge, UK, 1900. [Cho02] Choi, K., and Jiang, S., Spatial Sti ffness Realization with Parallel Springs Using Geometric Parameters, IEEE Tran s. on Robotics and Automation, Vol. 18 (3), pp. 274-284, 2002. [Cib99] Ciblak, N., and Lipki n, H., Synthesis of Cartesian Stiffness for Robotic Applications, in Proc. IEEE Int. Conf Robotics and Automation, Detroit, MI, May 1999 pp. 2147, 1999. [Cra98] Crane, C.D., and Duffy, J., Kinematic Analysis of Robot manipulators, Cambridge University press, Cambridge, UK, 1998. [Cra01] Crane, C.D., and Duffy, J.,Class Notes: Screw Theory and its Application to Spatial Robot Manipulators, MAE, University of Florida, Gainesville, FL, 2001. [Dan95] Daniali H.R.M., and Zsombor-Murray P.J., Singularity Analysis of a General Class of Planar Parallel Manipulators, In IEEE In t. Conf. on Robotics and Automation, Nagoya Japan, pp. 1547-1552, 1995. [Das94] Dasgupta B., Reddy S., and Mruthyunjay a, T.S., Synthesis of a Force-Torque Sensor Based on the Stewart Platform M echanism, Proc. National Convention of Industrial problems in Machines and M echanisms, Bangalore, India, pp.14~23, 1994. [Das98A] Dasgupta, B., and Mruthyunjaya, T. S., A Newton-Euler Formulation for the Inverse Dynamics of the Stewart Platform Manipulator, Mech. Mach. Theory, Vol. 33(8), pp. 1135-1152, 1998. [Das98B] Dasgupta, B. and Mrut hyunjaya, T.S. Singularit y-Free Path Planning for the Stewart Platform Manipulator, Mechanism and Machine Theory, Vol. 33(6), pp.711-725, 1998. [Das00]Dasgupta, B., and Mruthyunjaya, T.S ., The Stewart Platform Manipulator: A Review, Mechanism and Machine Theory, Vol. 35, pp. 15~40, 2000. [Dim65] Dimentberg, F.M., The Screw Calc ulus and its Application in Mechanics, Foreign Technology Division, Wright-patte rson Air Force Base, Ohio. Document No. FTD-HT-23-1632-67, 1965.

PAGE 127

115 [Duf96] Duffy, J, Statics and Kinematics with Applications to Robotics, Cambridge University press, Cambridge, UK, 1996. [Dwa00]Dwarakanath, T., and Crane, C., I n-Parallel Passive Co mpliant Coupler for Robot Force Control, Proceedings of the ASME Mechanisms Conference, Baltimore MD, pp. 214-221, 2000. [Fic86] Fichter, E.F., A Stewart Platfo rm-Based Manipulator: General Theory and Practical Construction, Int. Journal of Robotics Research, Vol.5 Issue 2, pp. 157~181, 1986. [Gai83] Gaillet, A and Reboulet, C., An Is ostatic Six Component Force and Torque Sensor, Proc. 13th Int. Symposium on Industrial Robotics, Chicago, IL 1983. [Gou62] Gough, V.E. and Whitehall, S.G., "Uni versal Tyre Test Machine," Proceedings of the FISITA Ninth Int. Technical Congress, London, UK, pp. 117-137, 1962. [Gri89] Griffis, M., and Duffy, J., A Fo rward Displacement Analysis of a Class of Stewart Platforms, Journal of Robotic Systems, Vol. 6(6), pp. 703-720, 1989. [Gri91A] Griffis, M., Kinestatic control: A Novel Theo ry for Simultaneously Regulating Force and Displacement, Ph.D. Dissertation, University of Florida, Gainesville, FL, 1991. [Gri91B] Griffis, M., and Duffy, J., K inestatic Control: A Novel Theory for Simultaneously Regulating Force and Disp lacement, Trans. ASME Journal of Mechanical Design, Vol. 113, No. 4, pp. 508-515, 1991. [Gri93] Griffis, M., and Duffy, J., Method and Apparatus for Controlling Geometrically Simple Parallel Mechanisms with Distinctive Connections, United States Patent, Patent Number 5,179,525, Jan.12, 1993. [Hua98] Huang, S., The Analysis and Synthesis of Spatial Compliance, PhD dissertation, Marquette Univ., Milwaukee, WI, 1998. [Hun93] Hunt K.H. and Primrose E.J.F., Asse mbly Configurations of Some In-Parallel Actuated Manipulators, Mechanism and Machine Theory, Vol. 28(1), pp. 31-42, 1993. [Hun98] Hunt, K.H., and McAree, P. R., The Octahedral Manipulator: Geometry and Mobility, The Int. Journal of Robotics Research, Vol. 17, No. 8, pp. 868-885, 1998. [Inn93] Innocenti C. and Parenti C. V., Close d-Form Direct Position Analysis of a 5-5 Parallel Mechanism, ASME Journal of Mechanical Design, Vol. 115(3), pp. 515521, 1993.

PAGE 128

116 [Lee94] Lee, J., An Investigation of A Qu ality Index for The Stability of In-Parallel Planar Platform Devices, Master's Thesis, University of Florida, Gainesville, FL, 1996. [Lee98] Lee, J., Duffy, J., and Hunt K., A Practical Quality Index Based On the Octahedral Manipulator, Int. Journal of Robotic Research, Vol.17 (10), pp. 10811090, 1998. [Lee00] Lee, J., Investigation of Quality Indices of In-Parallel Platform Manipulators and Development of Web Based Analysis Tools, PhD Dissertation, University of Florida, Gainesville, FL, 2000. [Lin 90] Lin, W., Duffy, J., a nd Griffis, M., "Forward Displacement Analysis of the 4-4 Stewart Platforms," in Proc. 21st Biennial Mech. Conf. ASME, Chicago, IL, Vol. DE-25, pp. 263-269, 1990. [Lin94] Lin, W., Crane, C.D., and Duffy J. Closed Form Forward Displacement Analysis of the 4-5 In-Parallel Platform s, ASME Journal of Mechanical Design, Vol. 116, pp. 47-53, 1994. [Lon87] Loncaric, J., Normal Forms of S tiffness and Compliance Matrices,, IEEE Journal of Robotics and Automati on, Vol. 3(6), pp. 567-572, 1987. [Ma91] Ma O. and Angeles J, Architecture Si ngularities of Platform Manipulator, In. IEEE Int. Conf. on Robotics and Automation, Sacramento, CA, USA, pp. 15421547, 1991. [Mer92] Merlet, J,, Direct Kinematics and A ssembly Models of Pa rallel Manipulators, Int. Journal of Robotics Research, Vol. 11(2), pp.150-162, 1992. [Mer00] Merlet, J., Parallel Robots, Kluwer Academic Publishers, Boston, MA, 2000. [Ngu91] Nguyen C.C, and Antr azi, S.S. Analysis and Implementation of a Stewart Platform-Based Force Sensor for Passiv e Compliant Robotic Assembly, In IEEE Proc. of the Southeast Conf. 91, W illiamsburg, VA, pp. 880-884, Williamsburg, 1991 [Pig98] Pigoski, T., Griffis, M., and Duffy, J., Stiffness Mappings Employing Different Frames of Reference, Mech. Mach. Theory Vol.33, No.6. pp. 825-838, 1998. [Pol40] Pollard, W.L.G., "Spray Painting Machine," US Patent No. 2,213,108, August 26, 1940. [Rag93] Raghavan, M., The Stewart Pl atform of General Geometry has 40 Configurations, Trans ASME J Mech. Trans. Automat Des Vol. 115, pp. 277~282, 1993.

PAGE 129

117 [Rid02] Ridgeway, S., Research Proposal for Parallel Manipulator, MAE, University of Florida, Gainesville, FL, 2002. [Rob00] Roberts, R.G., Minimal Realization of an Arbitrary Spatial Stiffness Matrix with a Parallel Connectio n of Simple and Complex Springs, IEEE Trans. on Robotics and Automation, Vol. 16(5), pp. 603-608, 2000. [Rob02] Roberts, R.G., A Note on the Normal Form of a Spatial Stiffness Matrix, IEEE Trans. on Robotics and Automation, Vol. 17(6), pp.968-972, 2001. [Sel02] Selig, J.M. and Ding, X., Structure of the Spatial Stiffness Matrix, Int. Journal of Robotics and Automation, Vol. 17(1), pp.1-16, 2002. [Ste65] Stewart, D., "A Platform with Six Degrees of Freedom," Proc of the IMechE, Vol. 180, Pt. 1, No. 15, pp. 371-385, 1965-66. [Svi95] Svinin, M.M., and Uchiyama, M, Optimal Geometric Structures of Force/Torque Sensors, Int. journal of Robotics Research, Vol. 14(6), pp.560~573, 1995. [Tsa99] Tsai, L.W., Robot Analysis, the Mechanics of Serial and Para llel Manipulators, John Wiley& Sons, New York, USA, 1999. [Wam96] Wampler, C.W., Forward Displacem ent Analysis of General Six-In-Parallel SPS (Stewart) Platform Manipulators Usi ng soma coordinates, Mechanism and Machine Theory, Vol. 31(3), pp.331-337,1996. [Wen94] Wen F. and Liang C., Displacem ent Analysis of the 6-6 Platform Mechanisms, Mechanism and Machine Theory, Vol. 29(4), pp. 547-557, 1994. [Zha04] Zhang, B, and Crane, C., Special Singularity Analysis for a 6-DOF Parallel Platform, 10th Int. Conference on Robotics & Remote Systems for Hazardous Environments, Gainesville, FL, USA, 2004.

PAGE 130

118 BIOGRAPHICAL SKETCH Bo Zhang was born in Xian, China, 1975. He received his Bachelor of Science degree in mechanical engin eering from XiDian University in his home town in 1997. Then he went to Beijing to continue his st udy and received his Master of Science degree in mechanical engineering from Tsinghua University in 2000. He found his strong interest in robotics and joined the Department of Mechanical and Aerospace Engineering, University of Florida, for his PhD degree in mechanical engineering. Since then, he has continued his studies and worked as a research assistant with Dr. Carl D. Crane III at the Center for Intelligent Machines and Robotics.