<%BANNER%>

Design and Analysis of Fixturing Methods for Mesoscale Manufacturing

Permanent Link: http://ufdc.ufl.edu/UFE0025106/00001

Material Information

Title: Design and Analysis of Fixturing Methods for Mesoscale Manufacturing
Physical Description: 1 online resource (77 p.)
Language: english
Creator: Rao, Koustubh
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: compliant, fixturing, manufacturing, mechanism, mesoscale, runout, stiffness, variable
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The workpiece-fixture and the tool-workpiece interactions are important factors for a manufacturing process and play a major role in the efficacy of the process. They lead to non-beneficial machine performance in terms of inaccurate dimensioning and improper finishing of the workpiece. These interactions become even more important for mesoscale manufacturing because of the higher accuracy requirements. In addition, the type of fixturing used on mesoscale machine tool system sets a limit on the ability to achieve high accuracy of dimensions. Fixturing also remains a critical issue impeding the integration and autonomous operation of micro/mesoscale manufacturing systems. To push beyond these accuracy limits, innovative work-holding approaches are needed. This study presents an investigation of four fixture element (fixel) designs for fixturing to be incorporated into mesoscale manufacturing systems. These fixels help in improving the process accuracy by taking into account the effects of tool runout. Using compliant mechanisms and components (e.g., monolithic four-bar mechanisms and/or cantilever beams), fixels exhibiting mechanically adjustable stiffness characteristics are achievable. Manually or automating the stiffness adjustments, these fixels provide a functionality for enabling greater control of the response of the workpiece due to runout and variation in contact forces at the tool-workpiece-fixture interface. The method of implementing these fixel designs (using four fixels) for common manufacturing operations is suggested in addition to the designs. Each design has specific mechanical parameters for its fixels and adjustable stiffness characteristics are achieved by varying these parameters. The monolithic four bar design has two mechanical variables whereas the three cantilever/compliant beam fixel designs require actuation of only one mechanical parameter. To quantify the fixel functionality and its dynamic range, the theoretical models of the stiffness characteristics expressed as a function of these mechanical variables are presented for each of the designs. Upon establishing a common stiffness range for the three compliant beam fixel designs, a metric is formed for better comparison between the designs. This metric is based on the sensitivity of stiffness expressed as a function of slenderness ratio and an operation range, bounded by a minimum possible stiffness value shared by the cantilever beam fixture models. The slenderness ratio is defined as the ratio of the minimum length of the cantilever beam to its maximum width while the operation range is determined by the ratio of the minimum and maximum possible values of the fixel s mechanical variable. Using this metric, results are generated for each of the designs and then compared with one another to highlight the advantages and disadvantages of each design. Models for one of the fixel designs are then developed to perform dynamic analysis and understand the behavior of the fixturing design under manufacturing operation conditions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Koustubh Rao.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Wiens, Gloria J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025106:00001

Permanent Link: http://ufdc.ufl.edu/UFE0025106/00001

Material Information

Title: Design and Analysis of Fixturing Methods for Mesoscale Manufacturing
Physical Description: 1 online resource (77 p.)
Language: english
Creator: Rao, Koustubh
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: compliant, fixturing, manufacturing, mechanism, mesoscale, runout, stiffness, variable
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The workpiece-fixture and the tool-workpiece interactions are important factors for a manufacturing process and play a major role in the efficacy of the process. They lead to non-beneficial machine performance in terms of inaccurate dimensioning and improper finishing of the workpiece. These interactions become even more important for mesoscale manufacturing because of the higher accuracy requirements. In addition, the type of fixturing used on mesoscale machine tool system sets a limit on the ability to achieve high accuracy of dimensions. Fixturing also remains a critical issue impeding the integration and autonomous operation of micro/mesoscale manufacturing systems. To push beyond these accuracy limits, innovative work-holding approaches are needed. This study presents an investigation of four fixture element (fixel) designs for fixturing to be incorporated into mesoscale manufacturing systems. These fixels help in improving the process accuracy by taking into account the effects of tool runout. Using compliant mechanisms and components (e.g., monolithic four-bar mechanisms and/or cantilever beams), fixels exhibiting mechanically adjustable stiffness characteristics are achievable. Manually or automating the stiffness adjustments, these fixels provide a functionality for enabling greater control of the response of the workpiece due to runout and variation in contact forces at the tool-workpiece-fixture interface. The method of implementing these fixel designs (using four fixels) for common manufacturing operations is suggested in addition to the designs. Each design has specific mechanical parameters for its fixels and adjustable stiffness characteristics are achieved by varying these parameters. The monolithic four bar design has two mechanical variables whereas the three cantilever/compliant beam fixel designs require actuation of only one mechanical parameter. To quantify the fixel functionality and its dynamic range, the theoretical models of the stiffness characteristics expressed as a function of these mechanical variables are presented for each of the designs. Upon establishing a common stiffness range for the three compliant beam fixel designs, a metric is formed for better comparison between the designs. This metric is based on the sensitivity of stiffness expressed as a function of slenderness ratio and an operation range, bounded by a minimum possible stiffness value shared by the cantilever beam fixture models. The slenderness ratio is defined as the ratio of the minimum length of the cantilever beam to its maximum width while the operation range is determined by the ratio of the minimum and maximum possible values of the fixel s mechanical variable. Using this metric, results are generated for each of the designs and then compared with one another to highlight the advantages and disadvantages of each design. Models for one of the fixel designs are then developed to perform dynamic analysis and understand the behavior of the fixturing design under manufacturing operation conditions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Koustubh Rao.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Wiens, Gloria J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025106:00001


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 E20110216_AAAAHS INGEST_TIME 2011-02-16T20:03:38Z PACKAGE UFE0025106_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 24176 DFID F20110216_AABISV ORIGIN DEPOSITOR PATH rao_k_Page_75.pro GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
479adef431e89f57cd00a56652868ce8
SHA-1
80a4771f5b4eeafc4b55069c1d0b60a654074973
51638 F20110216_AABIBF rao_k_Page_64.pro
3c6df975e7158280d9f86808b07ea2e8
0020833b0bc2552d48a246494151e8c0617e9f08
112635 F20110216_AABINX rao_k_Page_05.jpg
d4604cb43572b71aae0bdc1e9f7eb466
e189534fd8f1c324afd4a35e598a1abdb7d2ed17
32572 F20110216_AABIGC rao_k_Page_49.jpg
8b803907111d0c99c287d5ebf73d858f
e65541fd6ab95e2bc37963d465362b87f2086979
6001 F20110216_AABISW rao_k_Page_03.jpg
786f4245c9449fe9736349e3f28c3826
1205341a3503331a7e28cee86fc2a5192fedb18d
51645 F20110216_AABILA rao_k_Page_62.pro
07423f8ff141fb1fbdd5caa3cd88247d
1473154fa3433a1f119a897652ab11502dd4ee00
3113 F20110216_AABINY rao_k_Page_51.pro
fa95e18119c1b4789f1f2f5bc1479f80
3cf68a11fb64b4d15a94bd1b0c933129e584bb97
30350 F20110216_AABIGD rao_k_Page_33.QC.jpg
3aff2c0cb42fa34ad102c7975d04488d
d090e9fd05e43e8871e0830c8cdc6d7d1799aea2
43734 F20110216_AABIBG rao_k_Page_53.jpg
8775a3682f629a1bdc3f7965d87649d4
4d45abe42a0f40565ee448aba874f5530f76c71b
1053954 F20110216_AABILB rao_k_Page_35.tif
7509904dcb8d40744e80191642d6d498
4b492f0a84cfd8e10dd6c22bc48f77528c7a2d90
F20110216_AABINZ rao_k_Page_57.tif
1db4c87522ed0b9786ae330939572e51
ce69e5d3d7891d3953ac18fbf8d22f752e2f0740
35889 F20110216_AABIGE rao_k_Page_16.QC.jpg
b4620d8c07ea107d15fe0c549d54f06a
0b389c7f047e2e418c4a9da0520b0e38e5e647ff
46547 F20110216_AABISX rao_k_Page_06.jpg
eb375be2411497b120a043e7be492051
2dd47a0b1a76eefd5b3b2075547f8b379062fd7e
5688 F20110216_AABIBH rao_k_Page_18thm.jpg
baadd25aa356b5ec80db1d6f222b1ef6
60b6d15de8acb15ec82eabbfa851f89097d91b3c
1593 F20110216_AABILC rao_k_Page_35.txt
78f4a87053b6caa6613b98cd9837b8ea
7f45ae502646db30871f1235f979256aabd9e433
22447 F20110216_AABIGF rao_k_Page_09.QC.jpg
a46b21b72bc9934c3d0889019ce291f8
d654ff7ad0e9dbbb91531623921a687deacb6c10
47839 F20110216_AABISY rao_k_Page_07.jpg
3828a3bd30b249967983d3a85ff8076d
3ccdcb63cea7c18e0e5982dabfcbef8da873f0d5
F20110216_AABIQA rao_k_Page_02.tif
d97eb372dd65af1e4e93513857603f56
529dd47ef718beb0f7e2a06b8bbad82d514b889d
47861 F20110216_AABILD rao_k_Page_13.pro
487476d8c76022f8a7ee9d2e8a16b537
f10c711ce97320a060ee597239cc1615891de931
23646 F20110216_AABIGG rao_k_Page_67.pro
423392f1e714f6d8a57b6a368eec6632
2c3ab165c75ee4a0eeb205d48d1b0e3a9598458c
14792 F20110216_AABIBI rao_k_Page_48.QC.jpg
83111e96d2eb9d7a88b9c0da39511e84
bfef61f56c9c21263dda0eed50412874060f4995
35370 F20110216_AABISZ rao_k_Page_08.QC.jpg
0154f4bb3b2cea8ea563ada07ad51ab8
f402dd6e7bb602e8343cb78d0d02347a6e154f42
73132 F20110216_AABIQB rao_k_Page_18.jp2
22f231ee818e7f23ca412066b0fe0068
759842859b33f9f06be6459948b683d3ccb00440
688889 F20110216_AABILE rao_k_Page_55.jp2
9b98db412eb36a4f5da0f49be997c5d1
af9fca5a8503d7b8d45ae84c0ee22f723f7ca73b
2129 F20110216_AABIGH rao_k_Page_74.txt
e77836cd8201e81e502d7596a7886aee
c660a35e2e0d37cf86b3aa4705ffb3cb3117453b
347675 F20110216_AABIBJ rao_k_Page_68.jp2
4b9c56d723587c1e252699bb3d5ec58e
3c49ffc79349777a47ae1214b618155106f9ed9b
2067 F20110216_AABIQC rao_k_Page_21.txt
9dcb2ba2c986efd19c65ca4fbcf4f659
482dc23c5c374213e0735270cd478b5dcc05b537
168 F20110216_AABILF rao_k_Page_48.txt
c2f3d703c8d25f24dcab8ea203a6390a
6ef3b7503f9be83c8f12d1aaae87c9b893cb5c59
F20110216_AABIGI rao_k_Page_45.tif
8399b195ffd18397622836b1401faf44
f53a01eeba4a76a4ffeaf27ca8003b4c716814e3
7631 F20110216_AABIBK rao_k_Page_01.QC.jpg
712b913f1109ebd9fa525af32baf3a5d
ef661d4f48630a0ac07dbfa63ae1aa14b5b5f8e3
29370 F20110216_AABIQD rao_k_Page_47.pro
d061eecee7770e1145e73a3d97b973bf
208220e7fb00ce489beef41e63badfa89d3c3ec4
2648 F20110216_AABILG rao_k_Page_41thm.jpg
225730e89c510a5f850e8619bff34844
b69c956f9ae757bda8abb8f7c366c14bc6057a3d
32397 F20110216_AABIGJ rao_k_Page_18.pro
e471bbfa7d8204a4e3dd146ab825ee52
28ce79fac2d894f8265b8e137795a86ba5c26975
86140 F20110216_AABIBL rao_k_Page_46.jpg
cc79884b165575f751cd9c88dd53c038
110cb5ed3c6d21b89f050ec02248ed7121a2e254
515636 F20110216_AABIQE rao_k_Page_69.jp2
a246f00f38d8933a29ca188e322b20df
513a73d89719106efd91b76c6dac15af5811e2ca
25271604 F20110216_AABILH rao_k_Page_19.tif
0b32ed984b3319000b639f8f08c3f647
c62e10600fe6c1fea9e4f079c57b14b2e1a7670d
16649 F20110216_AABIGK rao_k_Page_04.pro
c414924f6791ee93141c641ef36fe96a
11df6872b5265d92d5ba5d60df653c788ad2a385
477 F20110216_AABIBM rao_k_Page_71.txt
7b8fe20138f876e583544765e26e7495
10babd85538b80131188f13215c6e8e027b26156
F20110216_AABIQF rao_k_Page_56.tif
39b2e9b3efc233cc41b2d85447d8b995
b45fcaaeda15a2664b897e5166108c20665b4268
896804 F20110216_AABILI rao_k_Page_32.jp2
c50986fd63d54f90dbfc1d40fc191f09
2754ac49c141e1289696f4b119290f0fd4273163
21438 F20110216_AABIBN rao_k_Page_07.pro
b07564115f23b9f5b21433bd5c7f27ca
320005fd47cd2d442fc246603265ad6b5c45d6a2
56229 F20110216_AABIQG rao_k_Page_71.jpg
e6ac7c7cf89a7d453eb426fab560cf22
8ea2ca8adf19b21d6e953306eac307d0c962c47e
3985 F20110216_AABILJ rao_k_Page_48.pro
b1eb0b2792107830a0182181a8d30975
7abf27f05790ea8ab0254e1b85a52455508a2a31
F20110216_AABIGL rao_k_Page_51.tif
71e7075302d338485b1f3e2300c7a924
7abf6a839becd37b2646a829202f69c5ab1fad27
1108 F20110216_AABIBO rao_k_Page_67.txt
c5f8e44a3c7025dfbb5522c230b7fecd
35fd001ef5bafffa09000e8fcdcb684147c4bbcc
7778 F20110216_AABIQH rao_k_Page_01.pro
888283bce6269538f2c09aad27189f9b
7f8fb18162bbb7e7a405b3b7371d436e946ea7d0
11245 F20110216_AABILK rao_k_Page_37.QC.jpg
9fcf65292b9cc3efc03d74f7afb00f77
b0d2c0edd124b85ee8bbd74583aeae69add29b81
7862 F20110216_AABIGM rao_k_Page_13thm.jpg
1afd366db21a11263fdddf9d0f3851df
e64b980662211f1fd5bade6fc7ab532e9784d67c
1999 F20110216_AABIBP rao_k_Page_73.txt
4cec8649ea318bb4ffcac2ce13a118fd
a8d074029513d496bd545e96c630ba8b0b9c81c6
49145 F20110216_AABIQI rao_k_Page_21.pro
8ae5c0d04a9e89f62d3d6c278a387c35
7829f23f05ecd2c984049111882ebd41c019a79f
10500 F20110216_AABILL rao_k_Page_49.QC.jpg
1d517ac76fee872eafa062be7ec5eb72
f50f471ba8c8bcd6a8030ee4a8f8a73052e1d289
28245 F20110216_AABIGN rao_k_Page_12.QC.jpg
6c1f6320fffb7879e3d455436b307706
66f731d6f2d387f657dcc76c2baeab77592deb1b
F20110216_AABIBQ rao_k_Page_14.tif
5e530904375a72403f1514552d964a10
71303691f0f9e75010da2da50883181eb7a7aefd
903050 F20110216_AABIQJ rao_k_Page_07.jp2
1a3328f1bc111c8cd6bd56f457b812a9
dd31fdf91adda7253e1890ac90d17622d14ce3c9
F20110216_AABILM rao_k_Page_27.tif
e1076cd5fe7e8c12f6fed9609f4a1cbd
c1962b647632f683cb35b987b5fed6d20aa79fbf
172 F20110216_AABIGO rao_k_Page_53.txt
01787a00a8d0fd2b6744e507a2ca4323
be2c42715a0bc93f5e8cec8389d26f4e8f51eaef
2041 F20110216_AABIBR rao_k_Page_62.txt
5bb3c7ff629d2aeb3101223ec2991683
c0c69dc60befc27f62884034f59b6ef62101193e
16813 F20110216_AABIQK rao_k_Page_75.QC.jpg
27f64492a2f9ee5750d9da4bca838fd9
8031c62145d7ecd8f841fce985ae087ee80e5b14
F20110216_AABILN rao_k_Page_54.tif
70c9debc23371f25d8c4500388aed82c
05d103f191ed13536dd2cbcdc164fea4dd03b9a2
F20110216_AABIGP rao_k_Page_69.tif
e51f957a16a0c7c2fd4c8dd2d439817c
05ec45e664d7f815aafc8a922d124794d26b6cd3
3686 F20110216_AABIBS rao_k_Page_49.pro
56eec9a722cf82e24f531c7ebe7b7661
011ef2192d812ea4b849270d8f447b05e86d04a5
3471 F20110216_AABIQL rao_k_Page_49thm.jpg
746c85790097289a0c671205cdbd9b76
14834459592923fae56d3d89eac43efaba94db85
29436 F20110216_AABILO rao_k_Page_40.jpg
e1ad335161f244c59f20d5d91fbd343e
05a86d22769f3d29ccf5c31b76cd8a432de9a5f9
1280 F20110216_AABIGQ rao_k_Page_28.txt
82bb8151d791083de0f37045f7b548d2
b779e602d9666f5c8272a8574178b51f6e9aa201
66995 F20110216_AABIBT rao_k_Page_08.pro
a3c5b9eaba471332df11c9dc8e07016d
8d976cff20e320d568c234df1edbd55594db393c
2166 F20110216_AABIQM rao_k_Page_16.txt
2ab9d3b2a348419f14ea1eaa2d80fb25
43595d2ead724a79b9fd5465679c39a4aab896b3
964 F20110216_AABILP rao_k_Page_75.txt
e905d2da9af0d54f65d02e38c139bd86
8dc7b596abe0d051decc5c7d547b1e7b88e779bb
F20110216_AABIGR rao_k_Page_25.tif
6435cdcd84022268505bc6bc03d1760d
c202e78d58892e0e097ca51315c8371d528ce675
56854 F20110216_AABIBU rao_k_Page_67.jpg
37c31f86888e79d65c2744692a920f16
89eed8260d575c64540140885b9c6f9e46c9bae8
F20110216_AABIQN rao_k_Page_50.tif
68af65fb45b4c5e37f2572cdcffa9199
f6fdde0dbd32780e15f40858953b1b87d3c2acfc
2280 F20110216_AABIGS rao_k_Page_14.txt
8eeaeb295c4300dcd9918310e47edc46
0a8215a042134d2825c493f4ca641337530a29d9
8207 F20110216_AABIBV rao_k_Page_61thm.jpg
24c232fcdae6c0f225dda228dc298367
33a2f24e958e84cc29e971a75d49576d29cca7a5
F20110216_AABIQO rao_k_Page_65.tif
ce00dfba100443bc9dd312743f978e11
199135e4b5c018989a4a5b11b4f3a3b39209d0ff
69510 F20110216_AABILQ rao_k_Page_45.jp2
1f83fb446814ac83d0c80d703352fd23
4b0560bde1d0659237f898842ccbaa9f3188a29d
33609 F20110216_AABIGT rao_k_Page_65.pro
2d5b7d1b4f23835dbdc5bc5f1d4e8f94
7f6a2b387534f2f13cf8b84c5fa0b96294248756
F20110216_AABIBW rao_k_Page_72.tif
1ea84acd3b5349bfeee7c50dfe532f46
eec08729ad900a67fbf935fe7ba0d49028465b3d
44595 F20110216_AABIQP rao_k_Page_33.pro
c86438313863a34d7618512729aaad3b
5da3995cebfaefeea6cf6103ca41afc7a1598fdc
90294 F20110216_AABIGU rao_k_Page_43.jpg
83138552feeb2f6beb97b821ffe784e3
639b65786b5b278d4e2985949581d9378bc3a956
105136 F20110216_AABIBX rao_k_Page_74.jpg
aea31ae8742b86a6d97ffc5be01ae014
4314b58a8faa86528a1d55c40a5d559e811ffd9a
56351 F20110216_AABILR rao_k_Page_57.pro
5af280b4958920a5095f1ddca152de4f
4ec1fed3c41e021db4b5dd5428c2b06846579f1f
12997 F20110216_AABIQQ rao_k_Page_68.pro
a708ce312519c239bd5586bdd208bc23
e26417b69acdb39c260381273258be54916e8f15
4643 F20110216_AABIGV rao_k_Page_56.pro
7fa1c01dc2a0b1cf42449a90566709a9
d51265f9a789a87567596b9b95e21fe88b0a2066
7935 F20110216_AABIBY rao_k_Page_36.pro
093d680135bb97218ae70d7b71b568a4
58658f391886c89836fd921d61f17e8613356828
34727 F20110216_AABILS rao_k_Page_62.QC.jpg
3a5152995fd761bdcdaba59104234806
cdbcbf42a1cc6df2f11b8a35c042a66759c719fa
1018040 F20110216_AABIQR rao_k_Page_72.jp2
08bec1eb9861d975da4f2d0953964064
f56152c197c6d246d712740c640c9fbbf413383f
8617 F20110216_AABIGW rao_k_Page_26thm.jpg
7e170e529a025b63ec2bd688ca9270cf
28c16737eb12338de946035681566c3d445f722b
F20110216_AABIBZ rao_k_Page_55.tif
be1f875d730a6e875d1693cb62388f64
c9e9334ca47dc826b019a08fce5b21041141fcab
7559 F20110216_AABILT rao_k_Page_32thm.jpg
f70747dcf85f1d0f158a62df02d4a1be
859471c1eb888114a8764a76e3ed2b8a57a51276
37993 F20110216_AABIQS rao_k_Page_14.QC.jpg
b4cf28e4d57868ea27748bad8297606e
71b844da1a340e81b6b455f81b0abff6e50eabf6
707 F20110216_AABIGX rao_k_Page_04.txt
cd19ad8cb8aea820a19b5a3dcbf309ff
432eeee856d3f45338a9148f97daf11635e587b0
4106 F20110216_AABILU rao_k_Page_53thm.jpg
02d2dd676154589d7f6897ddda33b860
6b71f20762714deff30c00795bf633b730528e65
104588 F20110216_AABIQT rao_k_Page_64.jpg
0acafbec39bfd094f666d4e6744fb0c6
6f7cd1d5328616091fe57ffdc21926d227b30182
484 F20110216_AABIEA rao_k_Page_41.txt
9f9894890ceb524bc313a83223c0107d
4f818de9d16b57edff6172e3fac740d90a216745
F20110216_AABIGY rao_k_Page_17.tif
c859716e7906ca09e2f5249d42a842ce
30363fbf5c81fd1b1099f8a39eafc480380c4af3
65480 F20110216_AABILV rao_k_Page_05.pro
6ba68ea724c83c601a476ea2d5e589eb
d771548a031bb308b35130b896c32a08e82551ce
55928 F20110216_AABIQU rao_k_Page_58.pro
d7c582dbf526e97a1600eb65ea021e38
914f2dae076e4de0ed4a4ec188939e5d6df1bf93
4394 F20110216_AABIEB rao_k_Page_52.pro
828670ce502a3850fa654e315e4d7fd2
d36d18560fe46dda586f672b44f4a25c8974e2cd
48134 F20110216_AABIGZ rao_k_Page_11.pro
e1e7ef754f52d24ef69fb1d4e84de4a4
86601566a3aaaa8e7c300d1e39c5c359782709c6
F20110216_AABILW rao_k_Page_42.tif
b939faed927c57a518d19f88633017c1
51b57d184a788be97254b8547de832fdf0cc42bd
31762 F20110216_AABIEC rao_k_Page_34.QC.jpg
9b6a17253b78cbd947d59c3a0f778b6e
ef4ccfc7f22f78405b30c5d9212bec7d12f6967a
8737 F20110216_AABILX rao_k_Page_76thm.jpg
9793b8e3f77cddd69222704d72516553
daaecbd8a8b9316a30abc502318bf91cbc750db9
8088 F20110216_AABIQV rao_k_Page_23thm.jpg
24ef6eade44fa7d9ee864e1e84feddac
2b1a55776629a9204fa8d343915686caba030a16
14372 F20110216_AABIED rao_k_Page_52.QC.jpg
5308134373cb5eeb5dec0ee662dcdd31
5d64bfb7a53399082e4bb87481573e3455f1e333
487 F20110216_AABIJA rao_k_Page_70.txt
11b9f42c93146600904557c1ac06cdbd
1c1078d9ba4118eaccb058407c72457af22c44fc
22806 F20110216_AABILY rao_k_Page_18.QC.jpg
93a75b97e857d3dcbc13d801efa4fd4b
a595abbe8512b064d5bb9f1a20a5d8f6cc2e6bc4
2019 F20110216_AABIQW rao_k_Page_22.txt
7f724358c6bba1649e21d58b5e9c06bd
fd0f7f5f49b31292d4d74d8944209d8792e00763
64999 F20110216_AABIEE rao_k_Page_20.jpg
a5719f03f8bde1444cb304af76e3d01c
196b5767358b36f081907b9f0d76e4e6b1192af2
2037 F20110216_AABIJB rao_k_Page_03.pro
818ef071e8569c76c9e2660c92badfb2
18944e6c77b2319ee431b1198ae92da85a97e38c
5527 F20110216_AABILZ rao_k_Page_54.pro
06144e83577cc96df3c529eede321729
29348ef36e09600b95aeff73c27f1a061f8cb552
69560 F20110216_AABIQX rao_k_Page_65.jpg
8918aeb2766bcaccd10a872cb5df827b
cb64b6687e1faf67dc5bfc51bbfb0bcdc69b62fd
9008 F20110216_AABIJC rao_k_Page_57thm.jpg
264f0b9e0a0cc4fc5810413ea1477df5
04ee50663977727a76c9e550f2c6c31a085c191c
8290 F20110216_AABIEF rao_k_Page_71.pro
33cb23768755009ca1b8e9cebc440dd6
1e515c881020e074dc9b873011a59b17c70c290e
F20110216_AABIQY rao_k_Page_18.tif
bcb3fe2b6dd893829383dfa6a6c686c0
875290080e27a61ad205826529e748cd57211898
95021 F20110216_AABIOA rao_k_Page_25.jpg
e139f8434dd725c4cce07f6074b7fb56
949f065bd31d3b4eb46381a37704b030fb14ee20
112904 F20110216_AABIJD rao_k_Page_57.jpg
e02cb6b3a3d02c2ef06183a9f2bec154
991145f8118a74f6b8d45d810f6f3af2bfc15f67
7272 F20110216_AABIEG rao_k_Page_69.pro
473c3944964f2ff28e7810a4e43e3991
d2c1c3fbb4ded946a7533ada0ac9a76bc11c3597
116384 F20110216_AABIQZ rao_k_Page_59.jp2
2bbeff376d47ee43809f28bce752633e
50892a134a74a5ef127dbf7d722a7d1ab86ec946
2260 F20110216_AABIOB rao_k_Page_17.txt
7a7243e6055314fecb4b174b901a755d
5958d84030f8a01189185abf2b3ac659cce9d80f
F20110216_AABIJE rao_k_Page_21.tif
908be0d95ba1e53e650706c9efa64f16
06486848c330286a9631de8bea1b5066d9ad3d27
5068 F20110216_AABIEH rao_k_Page_10.QC.jpg
686de4e9068c461549e20f44e542484c
a6ba65ef112773f105477efcb78bc16c71eb47c4
1032 F20110216_AABIOC rao_k_Page_06.txt
86a3c703eec7bcac60979b62381958f2
a538922bf68e190255cce3c0f187f32e1e650cc6
85325 F20110216_AABIJF rao_k_Page_12.jpg
189608ccde784b3cc2d411860da29cfc
9936e889085876c14cdb362f35cffa5e13dc746a
32598 F20110216_AABIEI rao_k_Page_30.QC.jpg
ad3dc8776dbb01b0c5eb1eb74a9b3396
865663f5fc940ef8d950ca6befbb3092b59b75da
14332 F20110216_AABITA rao_k_Page_10.jpg
d39a58c4dae2fd1aa655a116f8f725c6
e0209f4e73a909703fb4caf7494b5c2b32a3903e
21081 F20110216_AABIOD rao_k_Page_66.QC.jpg
f9c5fa195528dde07fde5cfa8a4369bc
d5836b41224fef57a572b20ad440c053525f6466
50108 F20110216_AABIJG rao_k_Page_48.jpg
9ea6347f570cb2cb5ee7799507c67342
6df27658ea8c8833501d4cb54fdf5349b2e529a0
31919 F20110216_AABITB rao_k_Page_11.QC.jpg
b4383c8ba1bc605a947934acb013da74
579feb21b2fe58155297d87823277c08dd22dd38
33016 F20110216_AABIOE rao_k_Page_61.QC.jpg
62a79695e1faba764193fc84fd23018a
6786d79217b57e5f6d2061c3d2c96c19887f79a5
F20110216_AABIJH rao_k_Page_46.tif
4c1bd529c4addd9ae7cad46226f968fb
c3be059e8a7e66765479b9f7e4a2657fcde16db4
94562 F20110216_AABIEJ rao_k_Page_29.jpg
4294a9b7bb6911599edabbe660478d7b
46b78ae470f46b4323ca0b64e57569d77bf83bb9
109720 F20110216_AABITC rao_k_Page_16.jpg
89894a881a6ebea2fa8ff77e3e2619e6
ef80cfdf3af2eff1837e786b7a2c349cbb2cbca5
36288 F20110216_AABIOF rao_k_Page_76.QC.jpg
f5e74ff67a0fe25b1a50bfff9cd586ae
63a08c91cf4a3a41c2c715ee41200e614419426a
F20110216_AABIJI rao_k_Page_24.txt
bc4a60e94416947c1d69fd4408ce82d2
84d3b217bc8bb3203401589a7e1f9548d919c7c8
F20110216_AABIEK rao_k_Page_33.tif
9d8d67dbb58d954ce2b671470fd5c30b
142cdb828487f624bea06f55ed6e324b5a83a604
37442 F20110216_AABITD rao_k_Page_17.QC.jpg
c5ca609762b8e212bd234f1c095d378f
4b6f373a09d95206fbaa1842e5af9e44279bd9b5
7679 F20110216_AABIOG rao_k_Page_20.pro
d8430b5b1a12e7edad92244909e2affc
1fc291c7f6a7d5a007297d18387be5a37e5ad8b5
128844 F20110216_AABIJJ rao_k_Page_76.jpg
7f6cadbeab0c98d5ac8ecd115190c479
a499af141e0a8caca85ddb719d3b5be11c0390ea
2018 F20110216_AABIEL rao_k_Page_23.txt
f738e0ed0d8d041ae58a25b83fe73cf9
8a75d05a8640298105f1e9c0cd02fcc28517e44e
36643 F20110216_AABITE rao_k_Page_24.QC.jpg
6a5085bdf36838a877fd67a8039eec35
7b368ead3dc062dfd6cbea37c67086e04cf34fa3
3330 F20110216_AABIOH rao_k_Page_04thm.jpg
c4af792a93ff3228a44a1397905762e4
87fec2af88e84ea6cb969ae6e0f3a09043f369cc
8740 F20110216_AABIJK rao_k_Page_42thm.jpg
7128aa81a0f573e1129a6dc8356b0db2
132fd4adf9070043f35c79f52787899e188e1286
923 F20110216_AABIEM rao_k_Page_07.txt
40316ec66f37879f70ee16c66add03bd
b6b7e7d4245a08ef7fe7543d93e57ac2a6f5f841
31122 F20110216_AABITF rao_k_Page_29.QC.jpg
dc7a03b8d67a15cc784edc1260bd3890
f9202dc69d15104190362ba26d1dd20f0e3dff0a
52800 F20110216_AABIOI rao_k_Page_42.pro
a1385d02dfd2e7602777bb32eedf17e1
d8b810f3e3ce4329301a6ed8ee270efa15ea9225
31183 F20110216_AABIJL rao_k_Page_25.QC.jpg
0f9d6d9f07a491689445757a27cd55b7
f965564f73b290c2ebf46be7d444e4dde5088d99
2084 F20110216_AABIEN rao_k_Page_34.txt
9a7bb25caf14bf0185806e9d7d68a253
d1686f70d4bed2fed6df1c47f218eb2206a1f4f8
32710 F20110216_AABITG rao_k_Page_31.QC.jpg
2408997d55aef91ad3a27d2046f7ba69
b4f34f101ca65fd0dddafcab1adb1ed1374fad48
96824 F20110216_AABIOJ rao_k_Page_60.jpg
dfeed3409f323672f4b81a7ad5134228
fe448a6a39d423f3e23dfeed413cfcca788ae228
648 F20110216_AABIJM rao_k_Page_38.txt
738ce470068de899922f1987906b0c9d
18035509f2d9d10705a27ca9a2692fa21b8a3a9d
8531 F20110216_AABIEO rao_k_Page_16thm.jpg
f8c720d73e1db66c2e90ddfc1009c4b0
7c97da75304f6c290639384f2f3c7228359bada5
83126 F20110216_AABITH rao_k_Page_32.jpg
da777e66115bb8be495eaa581933536c
bfc611c0da207218ff8ceeffea9e9f163e659981
2185 F20110216_AABIOK rao_k_Page_42.txt
f706ccb354ddac2bfabe9163f4b6cf89
a5e56538de92ec706ca3524dc75627bf51dca13f
130323 F20110216_AABIJN rao_k_Page_76.jp2
38f14b0499f8c7cc33f1d999227e7b21
b1e4c0b9d89078a8dc14fe0cb53068ce0068e89c
13746 F20110216_AABIEP rao_k_Page_07.QC.jpg
caa751a7dcfb59c8e5d7eb99689f0817
9d483c24ec7dd559f951458bd87db969b0601829
92541 F20110216_AABITI rao_k_Page_34.jpg
1e602d0b48d080e554a97b74403afd7a
d173973fc81667307ac489168d0931ffd4d51644
8230 F20110216_AABIOL rao_k_Page_44thm.jpg
d5a49c596f13cbc27a09388e1d9657a5
069621d80e0edbddfa3ad04f4e6cbf4973f1205a
8882 F20110216_AABIEQ rao_k_Page_27thm.jpg
1b59dd4e4eaa8061c461eca9f5cb98b0
18dfd852d60b6bc9243868e473d477e86e73739f
20375 F20110216_AABITJ rao_k_Page_35.QC.jpg
d108b2be8767e1ded8b3a404f1f8e073
954bec15d31e3528cb257ca54f2c28b91142e0b6
818486 F20110216_AABIOM rao_k_Page_06.jp2
5ac4f22b350720ec6b2a1cfc4143ee0e
21289f06ec83c441a2fc0a5f9ccd080f1de1f33d
F20110216_AABIJO rao_k_Page_77.tif
a09ed9509cd125197a6ccefded042bf6
3e284ec5e13547f15c6f1e7418982ffc46cd56cb
2040 F20110216_AABIER rao_k_Page_13.txt
8a54d7171fbea7719b758b2e9150ecb9
6d76bf3a13ec8ea8b326e60cda88e53d2e2b6697
39767 F20110216_AABITK rao_k_Page_36.jpg
03c6eecafd9045140aa8d09a94f2a12e
de3adc0e7f69caa314e8a0f1f22be6369b00228b
652266 F20110216_AABION rao_k_Page_19.jp2
0aaed819285c9733da89723ef56c7fc9
9cdd02bfce47a36fca8f02bfca41065957916672
62806 F20110216_AABIJP rao_k_Page_76.pro
dcfc000bf4fcab43119934a153e3f9fa
c95c14c9f0327d06fb56e2ae87958c0e27fa0bc8
55677 F20110216_AABIES rao_k_Page_15.pro
34edd7e5c1735e05c7a017c772ff8e82
31bf20a9d491b3e375a868d4b4d4ce69c497bcdc
33262 F20110216_AABITL rao_k_Page_37.jpg
7005b2fbd336347db44a44329fdd3d16
cd45697807ec5297828da301dba3469fcaf10f94
11421 F20110216_AABIJQ rao_k_Page_55.QC.jpg
e3797e2055a3b8007176d947a4e64a30
a4912d18d657c3060f76744fbabf6c134df1523c
3880 F20110216_AABIET rao_k_Page_71thm.jpg
9d9acf1ea20a3b1a0556d5f33ae58337
a548c535ce1bcbcc70ce0796bf7122d3cc575053
F20110216_AABIOO rao_k_Page_16.tif
827b42099180576b392e01132658939a
62602fcfa773e882a75a61c801077ce09b41963a
38770 F20110216_AABITM rao_k_Page_38.jpg
ee714c886d1e6a0358754fbc27728fa5
b61345019be0fecabf282673c1323d203fd7e9e3
23214 F20110216_AABIJR rao_k_Page_65.QC.jpg
e8dbf09caeb7bc63f0d5a1e186749da9
69a1fbf065686bbf2ebdaae04bc8dbe9ed826839
54976 F20110216_AABIEU rao_k_Page_24.pro
ffcba43dcec1d9e3b63535469efbc774
b24853a398f356a80a9d48b38254cda7b3f26b81
8090 F20110216_AABIOP rao_k_Page_39.pro
cb74a28c55d4dc220fd8338fbc0238bb
91259c1f62e9264b919ca79ca538533313785a5b
37402 F20110216_AABITN rao_k_Page_39.jpg
acec417d9dc094982e34339ae5f71fcc
0e8b18224610f84ec979d1644b345069ed3b91df
62655 F20110216_AABIJS rao_k_Page_70.jpg
bdbb7912807efbd2f421ea75bc4f2e87
68d4d2f21f3a96e659a5c4d3551ad497d3f26868
1372 F20110216_AABIEV rao_k_Page_65.txt
80b678b5cd213e6135e80fb6e89fabd7
b4e742f11ea032d29477d2c53429cf6758f563fe
F20110216_AABIOQ rao_k_Page_75.tif
c5ec2fabe1f42a27911881e4eccd978f
c7bbf9fb1fb5cdb19373abcb04e950e0841077fb
13665 F20110216_AABITO rao_k_Page_39.QC.jpg
98abe5aac6836323e45a6f1125543e00
9646fb47930a806ebccfd606e3a83377e38f5276
7765 F20110216_AABIJT rao_k_Page_41.QC.jpg
3d24cc520997e708e52f04e1af8a15a4
36eb12b2dccf791665b65d6856f1522d6263c666
40290 F20110216_AABIEW rao_k_Page_55.jpg
2cfa5627328c4b3c4034b90445d4b85a
4adf6ea6573e84cefd7715aa21d3e341058a1437
3508 F20110216_AABIOR rao_k_Page_40thm.jpg
c9cec91192ef5d36bf16ca704bcbc67c
f4ddbe3a732cc28fd972c2e214717b1c07d173c4
105537 F20110216_AABITP rao_k_Page_42.jpg
9480ec8bee389631368e10495ad7000f
bcc2dffb3948c2130f7516e379247da62c7dda8b
2169 F20110216_AABIJU rao_k_Page_63.txt
e82104cd55d142d471f92b3b6b71d71b
974ed4bf1362b51b0a326f66d43c23477d3988d4
2093 F20110216_AABIEX rao_k_Page_30.txt
999ed764c5059291b64c8947111f8fc2
a51184b9b66a91ad63cf01e80e8bb73763ed2b71
1985 F20110216_AABIOS rao_k_Page_31.txt
c593ace5d5991757b54d5b724eaaeeff
784cf3727205fe91f95fa91a04729e10aa77db94
9676 F20110216_AABITQ rao_k_Page_51.QC.jpg
f97fae040f4e61195cdae728c13be547
c9c4e7cc7bd81cef87ac5b9277a1536b83491498
F20110216_AABIJV rao_k_Page_22.tif
0101bf7a6820c84ae84e1b56a2b0a96d
7e94a336da697fb6167e202ec58593f29910fbab
20514 F20110216_AABICA rao_k_Page_47.QC.jpg
f60d0a72fd09fb76db7799cc89db0835
4631d42851c52e30170712dbcafc4c0618fc9262
379168 F20110216_AABIEY rao_k_Page_39.jp2
8d440de0a334a0d93ab3a09cec3cd690
8f1c1d2d78f067fa4dceb00ffe699818f92780f0
12634 F20110216_AABITR rao_k_Page_56.QC.jpg
6189a1e3f60d15345f9ad3c4c9ad3c99
b080dee9dfed1c8f94dd9bc7659ecc0feb3e3cf6
F20110216_AABIJW rao_k_Page_43.tif
20a2d18b3e45abf7de04d000ed888427
621d50cc63d300829856836f0f251f56d0bb9b5b
111511 F20110216_AABICB rao_k_Page_59.jpg
28b04e7a6ad150312f14616edcde0477
79c66f1cc48bbf39b25f640901cef11f8c35c087
2034 F20110216_AABIEZ rao_k_Page_25.txt
29a5c6d33d9ef8fb95fe28ef18413749
4a43ed396c2c6aa18be9da04773d8c3a4f7b5461
653989 F20110216_AABIOT rao_k_Page_54.jp2
d3a5569374c98466e67f205954ae6e92
c68e7ac6518f40ef9fc1216d4e037645b7ed6408
36410 F20110216_AABITS rao_k_Page_57.QC.jpg
bafa86cd2e0555c4a2eceb7a0502192c
821e560f9c6824a2468de8844511fa05f6595ab1
4903 F20110216_AABIJX rao_k_Page_47thm.jpg
82533083baf68cd97cbd9efca509076b
5d727553f71aeb338235fa121f6fa850939460a6
48911 F20110216_AABICC rao_k_Page_44.pro
066d80185d07ad8de8bbea2b52a167c7
396a46b266e670c6ef342e049e4dc45537756946
109397 F20110216_AABIOU rao_k_Page_30.jp2
093d43f80184e23b54c8b0f863c2b7bb
f21eabb084eacd52f9ab3d89cd3576fde72c0550
34150 F20110216_AABITT rao_k_Page_64.QC.jpg
215502044a34079707d1e5b3803ee48b
0a5433ab6e0f23dce44d776f74c868ab403715c1
2190 F20110216_AABIHA rao_k_Page_15.txt
3103caedb9b5963d7d971d9586e6d7cd
827fc9b662cf26242c5cc8874ff10e4281e7fce7
5020 F20110216_AABIJY rao_k_Page_02.jp2
d23097d9f4ea114e9866d923930307c4
5836b96cf4b96253e379c5d747e74de8dcea0cb7
40946 F20110216_AABICD rao_k_Page_04.jp2
b5c503a9808d5a9c9e83ba01826080fc
1b9efa94b7012ecd3fe4ce1b391472494e3f77a4
334 F20110216_AABIOV rao_k_Page_36.txt
6c945c9d59a79b386817ec30bed641ba
e5dfe2bd52cf1b9158188448c769bc3f19193fbb
72297 F20110216_AABITU rao_k_Page_66.jpg
7a9c2232e9b143094d2c6aad73c6f7ff
a2d87980fc7aef9ad3540f07022a2ec0c29a3070
10134 F20110216_AABIJZ rao_k_Page_40.QC.jpg
8381730eb29623a9d52e40bce22a39a8
a5a436a66987976cd5e4e76a028bb837fc18146c
11365 F20110216_AABICE rao_k_Page_54.QC.jpg
c80f37dd89b858828f319aa087ec42f9
11c0ea34fc1744fd94963e71450de523a536f356
3781 F20110216_AABIOW rao_k_Page_07thm.jpg
0a315c182b89c5ba65737d75e57be19f
94becefe9a072c70ba0f2cf04f40fa378ebeacf1
67109 F20110216_AABIHB rao_k_Page_35.jp2
3c044cb77e0080fe381d104de1d3d44e
da9a2a32676bf4899437239d526451191b98bf1e
49419 F20110216_AABITV rao_k_Page_69.jpg
331791d5658216b0866338d7aed63eda
7ae10c886d8014c7d133d8af185bdac4f909cd3e
425 F20110216_AABICF rao_k_Page_52.txt
72a959327f974fc55c98bf7db312fde8
abc58faae146c8b86919bbc5cedd5c3307926312
21942 F20110216_AABIOX rao_k_Page_45.QC.jpg
148d3f56ee32a74204e1a1348d3fc216
1cc30c0feddd5164d074484e6b7b51987a7dcf64
47995 F20110216_AABIHC rao_k_Page_29.pro
1816972caf97482bef0239d5a61869f4
1f880f0fd7a602c027243cff00c4cb9d93633a75
97676 F20110216_AABITW rao_k_Page_73.jpg
f08e00d22686b1afecb3cb7cd02d72e8
b14b2ccfc6c938cf81566cf458eaa1e342f61b01
100607 F20110216_AABICG rao_k_Page_25.jp2
ce946bd124c2175a92ddf895b6bde0b8
ee68b3405b2e0e8cfa1c25f88dfaa6e29c0b986b
F20110216_AABIMA rao_k_Page_63.tif
2c23ea4e892c305cd6d4db45bdace971
a3ccca7587cd46f58dc54bda490bbac3a64735ae
53977 F20110216_AABIOY rao_k_Page_72.jpg
c868fe1898e911551dbdff73474adec7
f6af1322e3435215369334b57d5b67a5a486aa05
F20110216_AABIHD rao_k_Page_64.tif
645612c98963657263feca6736630fe2
49052f7003b4d713e3e2cd7d676cc5c9034d1e33
31158 F20110216_AABITX rao_k_Page_73.QC.jpg
523d3784ef08da833a2725ce72e774a9
061a3e9b2c522203d9dcebddb9d4bb69f73a0d77
33349 F20110216_AABIMB rao_k_Page_21.QC.jpg
83f38516e51b795c939584509d723ae8
791c03aa0537fdcd9f79a52d0402678e2d7f7082
1980 F20110216_AABIOZ rao_k_Page_01thm.jpg
25019c050f3379eec693ef7990972b04
e2ab46f4cb798c40fe50e22e1737007368e4e936
9016 F20110216_AABIHE rao_k_Page_14thm.jpg
6a400ebd21588ddb46ac8d583213d782
e415946f4796b96a461aaeee077dd9c10b6356b9
100621 F20110216_AABICH rao_k_Page_30.jpg
44bd4d60dfaf9ed6f83c82d1dd63da02
39bc3cb5438ea08562deca5227401bcc35923995
8809 F20110216_AABIMC rao_k_Page_74thm.jpg
d31caf0052a7c15e2257c81f3023bde1
81f7fb4d457af8b889b1a74fd795803a4e670649
64593 F20110216_AABIHF rao_k_Page_47.jp2
5f66d591b58f91138584ecce32f664f9
dce6981e912fd31f9c0dcf775d36b6936c7ac5b0
103837 F20110216_AABITY rao_k_Page_13.jp2
a91380499297a790dbefc947d1e5a87e
496962cca8bb668418dd9d789212c7be69dad0e0
371 F20110216_AABIRA rao_k_Page_39.txt
1ea9e426575b5b1e332c19cfc77b6432
04f9d2ed434c03ad3dc3a69ed738998ea5cba215
4447 F20110216_AABICI rao_k_Page_48thm.jpg
171e67b1be8476b0e8f9e86a0226e3fc
a091acb635d2d17b064b197ff5fa72c64cc26343
538 F20110216_AABIMD rao_k_Page_02thm.jpg
a9ddae3fd0650f02d01131a37b7c26c7
24f04e610d26e93ee9db5bd2cb10f17084eb2495
105421 F20110216_AABIHG rao_k_Page_22.jp2
0af20f0e01414c2b8bfd6a879c261ab7
f3fde013506f5e9dc205593d9e577a6644f22bd0
273467 F20110216_AABITZ rao_k_Page_40.jp2
98cd456cc068f51f85ffbae167116a78
eac7ce1e539d87f227a4daef34dbb98252c8301c
F20110216_AABIRB rao_k_Page_12.tif
bd4524ace343ea44931f50f5ca8aec03
eca733e1c0334597f536bbd7da373411489010ab
589 F20110216_AABICJ rao_k_Page_72.txt
45629fa8310839b678a1eb43edabb21a
ff928fbd1b47959a656f2de4d883d801612b7085
36164 F20110216_AABIME rao_k_Page_15.QC.jpg
07a417bc80f10a2759e54dad963d9f94
f8c239ca531586a1a8aaa86f69bc4a643b22762a
10965 F20110216_AABIHH rao_k_Page_77.QC.jpg
d1e2a35d63a09c93bb878dc61fcce451
982223d790eea34af8a59d4d74a6712e8a9a0533
49482 F20110216_AABIRC rao_k_Page_22.pro
efc677ee94caf4ee481cc04d18facc4e
fbbcf3bd2e0c41493a0ee2913b2189dafe4b4b14
41954 F20110216_AABICK rao_k_Page_32.pro
98c08cded40a73ac2857865d0db0a06c
c21fe181d61e6dcefe40d83a41ee4f318d846c7c
9221 F20110216_AABIMF rao_k_Page_38.pro
6aa8fdd5f72212d4329e6f056139ee65
c26f40084a0d889163aff532b76cd2a2c44c34c4
4352 F20110216_AABIHI rao_k_Page_75thm.jpg
e384a43d1cb9199f725bd9445c173261
d6548c7011cd93e41c03ec48117f4aa54b0d9c62
7674 F20110216_AABIRD rao_k_Page_43thm.jpg
f32a124057245236ae6ab9f2067f5aff
bdf8c800c737566413b72b61ce1bed9ea9de4600
F20110216_AABICL rao_k_Page_38.tif
0d1412300da5919f4b386489ef3bc373
0411619167931f0746c86d709380c8ef5d2bf34a
48302 F20110216_AABIMG rao_k_Page_23.pro
fd53e47ea61c197e3713ace38dae3b60
acd2e284b1c362b37272e559c1d290f369c50829
7708 F20110216_AABIHJ rao_k_Page_11thm.jpg
da337bb92190a921a1612a92bac91115
01b4bdc9406d66230731384168943fe802cc2819
F20110216_AABIRE rao_k_Page_67.tif
f6287021e36a1cc300a6c5af7c7b3536
9cbbc1c0a150a1fb3c9eebd1ae0044e0b8c91cf2
34911 F20110216_AABICM rao_k_Page_26.QC.jpg
2fcbc4e0856d40bb69a26e5fdcf2f271
27d78b6d48d027b968ee9b92b9bdc2d8091762e2
8568 F20110216_AABIMH rao_k_Page_08thm.jpg
fe2a8ec8a3678e481e3e8f8834814fe0
344f130edd31c64458eec0a02858e9dfbbbd5ac6
90414 F20110216_AABIHK rao_k_Page_33.jpg
bdabf90660cb93c0a29c1a693be4ba3d
509ea8d0130dbda4498b74750386e5254cd72726
F20110216_AABIRF rao_k_Page_53.tif
33f2e954aaff03236e3411010601e99a
226dad1e83c25c3b7001e3fa5f88e2d303eb19d6
3647 F20110216_AABICN rao_k_Page_50.pro
4ebbf69c19aefe899ec3882be6f767f5
a21a44584255a756068fa488f2d1acab1a66cc39
2309 F20110216_AABIMI rao_k_Page_59.txt
9bcebd4ee7f8d0dd049b41dd38eca0e8
854e688f335b04d2f2e291821e19f0f30bf44203
53605 F20110216_AABIHL rao_k_Page_75.jp2
ce9c519707e3f32a49c8668e4be76277
b6fba37fe28c57111142e413c0da1a311d3fee2b
123640 F20110216_AABIRG UFE0025106_00001.xml FULL
6a7a05efa7d129b431a67a97e1773ab4
6738719c478d56cb2972f4a09f545bbf36af274b
41432 F20110216_AABICO rao_k_Page_12.pro
32e310c2d1708e6bc7ba0ab27cbeba7e
4c6e6cc24fff0ee70a0943fb2a46307e5fd89ce8
61712 F20110216_AABIMJ rao_k_Page_35.jpg
69e672eab45b122f88c23669e10813e9
371b6c762293cc9a093ec5cabbd9c53224e8d269
107904 F20110216_AABICP rao_k_Page_31.jp2
37cd3fa9fd5ce9501751e046c523470b
17f3e65dc36fc07102554c799e737ca8dddbfee5
99729 F20110216_AABIMK rao_k_Page_11.jpg
dacc6d8c4a9e664e85edb1886e56e4ef
85a3506ab4aaa4beeac46d03a5d0dfa5b2113bc5
30696 F20110216_AABIHM rao_k_Page_66.pro
12b8841f0bd759f529c3c0ab600228fe
353c032b1b12812981780768bacd6fd222f672ef
1357464 F20110216_AABICQ rao_k.pdf
eb44179acecfc005ad98cea15eb58a5b
17251ff16deefa796329184c43517aeb684ff6bb
36093 F20110216_AABIML rao_k_Page_63.QC.jpg
fabb3817acb8d0a4a204ea7461e2ced5
31b2868d9b5bc6fb0ba72eb3903123f1c2b50341
7640 F20110216_AABIHN rao_k_Page_46thm.jpg
7aee75ce36fb92ca6a1e7ebfe94dc577
439ba6517507ab4dbae3e810d80a1e1fdecd1a18
F20110216_AABIRJ rao_k_Page_01.tif
795a8d6ecc3a6aa6aeb741c07ffb138b
417db403262405af950f0639831b4b1896453242
1388 F20110216_AABICR rao_k_Page_66.txt
b3dfc29c8cf71ebf81fa8f488877c1eb
382df421650cbba860e3b8b617b22d61cb713e0d
4217 F20110216_AABIMM rao_k_Page_52thm.jpg
a5a508763243e471c2478e7074ae8f7c
faa78a49ed4d529d571d2c58db942106656f7645
54205 F20110216_AABIHO rao_k_Page_16.pro
d44699482132cac99eefa0c5447e3359
ebc3f75d7920faff557a124b6d0edcce8df2252d
F20110216_AABIRK rao_k_Page_04.tif
5c66fae61f53392466d4d12d8a1a67fb
da7f84a86c9772822af9b23320b744b706460aea
103225 F20110216_AABICS rao_k_Page_23.jp2
ca0316f57c96d11ac06493fb54865766
cb7163681509fe538b8d5d10245189a71a5721d5
38583 F20110216_AABIMN rao_k_Page_54.jpg
37f4d7cd4e813f2c0f4b31f2d146b3f7
fe54235ec385a8933df1a510287d3da8b9ba8c02
100822 F20110216_AABIHP rao_k_Page_21.jpg
cf5600a10093057ef43b31ca86f769ca
77ca022e01550895d181114a78d8e324e23d74c9
F20110216_AABIRL rao_k_Page_09.tif
eaf1d8fdf519449a7b46d076fd4b9010
4e23a818812710ee207edb508ad0aff741478e8c
2753 F20110216_AABICT rao_k_Page_05.txt
946edff279804663c649fdf8f748232e
e9ce1d8f139e1b721a18f85dbcae37c970601bfa
205 F20110216_AABIMO rao_k_Page_56.txt
497b7f418d1a686b66912e4055dac3be
d7bcdda1b84ec171a5601990212c7aac012ad6ba
11108 F20110216_AABIHQ rao_k_Page_06.QC.jpg
7ec36bba10f0204b941eae3d0507bc0a
63784ac104c5de3fbdc8a959faf35befb8460eae
F20110216_AABIRM rao_k_Page_13.tif
c98b9699573438b2717f1c69c31df677
caedf75cabd1c0bd08532d4a0a9a48310d64e159
25290 F20110216_AABICU rao_k_Page_06.pro
2621ff0802fd6c9bd8a90e5224ba563f
1589c6e0311c014a14e5f5f91c598da74d22b9af
47213 F20110216_AABIMP rao_k_Page_52.jpg
62068c3fdc91f204593d6d90aff27628
a1b33b301184434b0fddf2f351c2481207344e8e
2165 F20110216_AABIHR rao_k_Page_27.txt
300dc21e614006856287142ea8f0641c
261ff721bfab7b2dd9295919581de536d024eb66
F20110216_AABIRN rao_k_Page_15.tif
d649b45e06e1ddabc4e688e151282410
0323919e4b1ce3942ad8bf7b422f32be35ca6acf
3495 F20110216_AABICV rao_k_Page_50thm.jpg
e1f4695d2dc0b6397f1321f5ff526264
68f08a47188e0dab7c9d6469febf22a6cd270413
296 F20110216_AABIMQ rao_k_Page_51.txt
34334792397646ecfb8e006a4ed2cdaf
e95d9321e4e9b4babe36848301f32c022ecc2598
5645 F20110216_AABIHS rao_k_Page_65thm.jpg
edb47bfd3848ec396d11445d12078110
18e88d463a0735809dd806fcc48b1393948ca4c3
F20110216_AABIRO rao_k_Page_23.tif
56b8c3f8da9abcb9bf99575c105641c4
c49251f58c3b7e221d58df13969f084d0323f972
27644 F20110216_AABICW rao_k_Page_32.QC.jpg
634fdd16b45be88dd68610389b201b82
8e1cb2fb40ed8d91eb66575ed7c1ce9d3e52de43
462 F20110216_AABIHT rao_k_Page_01.txt
c8db0a1a4845de479c3dfa2c8065ef03
40e4ff617d2c48f56018fbd2a8192531e922222a
F20110216_AABIRP rao_k_Page_26.tif
744b7edcb83c768fb900a594a431f3af
bb319e64a927e441f743adb216615a0eefb45b84
23750 F20110216_AABICX rao_k_Page_01.jp2
3b822fffff72ae3d8783309b33885d2d
e495c1de416d18d2173473ff06c10194f81c4eda
F20110216_AABIMR rao_k_Page_64.txt
ff91cf94e53a35786cec080ad83b8959
67a7de596049be38a19b1fd77b374a986730853b
10414 F20110216_AABIHU rao_k_Page_68.QC.jpg
843a5e8839837cf526a6be1d78b25415
83b199f91270e628ee0aa7a1753f543dae291ea4
F20110216_AABIRQ rao_k_Page_30.tif
55900e9b73d8f56d7e0524d03407f495
19195420e4318da0792ece2af2bfd86eb079fe32
1373 F20110216_AABICY rao_k_Page_45.txt
0abdbb4ad5a98289ceb0ec00b1c8a131
83ef8d7b2b373b0037e12acf5be3b5b2452f6285
3444 F20110216_AABIMS rao_k_Page_51thm.jpg
addc00e6c05e518e0b4ce64ca2a053e9
3a226adddf4269219debb1af324e5f97341b93f5
8435 F20110216_AABIHV rao_k_Page_31thm.jpg
c161addcf5b1fb9e61b320b4a40c402d
4cc569263b3fd347ec122677db596de10359e4ac
5383 F20110216_AABIAA rao_k_Page_70thm.jpg
222f8946613a34f03a0818c68cf9952b
7cbf954a2fdb23dfdea9a40b230491d7920a4c0b
F20110216_AABIRR rao_k_Page_36.tif
a7f0b7d26133ed2313917958e8506183
f63cc7011a20345a6f913632a05d31b5b29c0023
66236 F20110216_AABICZ rao_k_Page_28.jp2
839408b25cb7d5606135875bd7d45819
d628004db4f740878f7aba1fd876ed2e7d94c467
5557 F20110216_AABIMT rao_k_Page_35thm.jpg
4376cb19b255d5faadc48e7142964b1e
122353e7eb02cb9e781012528613e2562c404011
F20110216_AABIHW rao_k_Page_07.tif
c9e35262f478f642c68e0f3dd5be9f7e
b1ed4d2152f5d610a58201b943c07654f309ac29
52449 F20110216_AABIAB rao_k_Page_19.jpg
a19aa36bec8434fc46b409cbe2704269
8a393b1cfa404237b3053619d3c3fefd64d0d8cc
F20110216_AABIRS rao_k_Page_52.tif
610dc8553ee7295bab18a1bfbfb66e1a
2186185b410d696121caaf102b2cbf93371d2148
5506 F20110216_AABIMU rao_k_Page_19thm.jpg
9ecf86a58bc8e9993936bfe3673b4617
946c9ab2f7064f34af6b3a2a42301259f4ac9ed1
31991 F20110216_AABIHX rao_k_Page_23.QC.jpg
3f203cdba3de849a0fe9e15f942ac57f
b4455b9408176fe354e5a34ae836760bbd3f61b3
F20110216_AABIAC rao_k_Page_05.tif
3d5828f73a75a128fcf4afa87921dc4a
4d9746af658bb1b3f7a958421d88ac5a11b2c067
F20110216_AABIRT rao_k_Page_60.tif
dfda6729b6b01ff250ce56730666af83
682d24d755817406cf97011b07a23e87f9825325
102474 F20110216_AABIMV rao_k_Page_29.jp2
69176d54ff6d2f55311dddd0e85b767c
eab7059c2baa7cb9e3a1fbbcf58329c5c7a0ac7b
6003 F20110216_AABIFA rao_k_Page_19.pro
fd98c24c668cd2b5aaae3f14fd5e48a3
dbdebaee7150d9d0d50498995a26077abfb2f5e4
115774 F20110216_AABIHY rao_k_Page_63.jp2
2d918edab8ec9f30239eb0ae6ac2ef2e
25a376d5a0cf2528fc21b0553953f8d16897ec62
24332 F20110216_AABIAD rao_k_Page_01.jpg
d00fdbe0101d5488c5be2a672b760dcd
4cad818c907240a2333e3e096b07cb7f6a9d13ba
F20110216_AABIRU rao_k_Page_62.tif
e9bfedc95ac5c33ce3a329d6d6ad3dce
35ecab7840086c7cf32d8eabbed67abb09117c78
110980 F20110216_AABIMW rao_k_Page_15.jpg
fffb54d58415958660b7e856256fddde
34b4b61d48df626500ebde379e3ceeff836bf7a3
2075 F20110216_AABIFB rao_k_Page_29.txt
7430c7d8d318ea21f4c502eb973910d9
188b1521cc9b32c7dad1c9faccba6e0dd0f6f75c
8429 F20110216_AABIHZ rao_k_Page_21thm.jpg
f5febd89b5277c790810a9bbc32ad257
7f06aded9a9cc59fc7bac9260a827e4da16e03ce
110308 F20110216_AABIAE rao_k_Page_58.jpg
c750ff132dafa9e39d025993bf256202
9b50c65d171b0bc08b337c30e1971024950ec383
F20110216_AABIRV rao_k_Page_66.tif
fd5cf0e74a44ac0b3613136831c849e2
7fa6c3e04ec9216e9b11d1a6f0ad77970cda19e6
4463 F20110216_AABIMX rao_k_Page_36thm.jpg
c1720c4fce2f4c48234358f9788b1b1d
36d33b56b2d9a4e76a96f0db3887480a3489d6dc
79476 F20110216_AABIFC rao_k_Page_09.jpg
c0b24549df0ab3c17a8d44ac67691261
fe3c1a4540a9444142f0c0613bb0fa1d0b234af6
91787 F20110216_AABIMY rao_k_Page_12.jp2
40155a5ed6d9fb0e9a8a2029f8bf2385
fbcf3c2d1963333bf9bceadf30f4a3fd0b9bfbf4
36172 F20110216_AABIFD rao_k_Page_59.QC.jpg
02442df27072684c7b4202dd6e1e51e0
92865179d5dcc21e5fbb2049e6a52ace246e175c
111685 F20110216_AABIAF rao_k_Page_24.jpg
089ca73d8f842b33d4251d62df6516ff
c44dcab84aa56e7841412af54ad604a0f9bc0f02
105155 F20110216_AABIKA rao_k_Page_11.jp2
c96b85880184f52741b084a173cf4918
50368dc70fde8480ac64e8ca2e3132286359961b
F20110216_AABIRW rao_k_Page_73.tif
2dff779a8a9fb00e414cd1bc78f1d9db
118b8f56f751cc411f2915077f14e49f9776fb62
5883 F20110216_AABIMZ rao_k_Page_69thm.jpg
54a26d1f36f21cb395f6ffbeba7eaf09
c35c27b3951fb27c2639916cd02c2d5400fd0b4b
F20110216_AABIFE rao_k_Page_48.tif
863120a965415b76f8998e74c9150025
a87e7e7e1452295f5f5aba894501ada8efbff25f
49605 F20110216_AABIAG rao_k_Page_61.pro
387485536d2a2b3de5bdd8cc3cbbf88b
1b5244c294d1688bbba9f63a88397558998c3ece
390 F20110216_AABIKB rao_k_Page_40.txt
ff304887964bff3731a4f80b2234b84e
81f126787f1a373ce54e2bd2291cbb4d4181619c
F20110216_AABIRX rao_k_Page_76.tif
b43407e69e7be3fcc437298a6fd2bbbc
776119ddc58ffe77487dec4cfb20830d772abcde
106913 F20110216_AABIFF rao_k_Page_21.jp2
81b287ac3c259a9464d03b52d6c3c630
87c43e727ce9f663ed5a34b73713b873bf1fcfc2
117164 F20110216_AABIAH rao_k_Page_24.jp2
ab56c7e5e15e3e3671fceeab73ce0d77
c68748a3b31ed105aae9334a499f639ba72637b6
F20110216_AABIKC rao_k_Page_40.tif
dfe0ccfb18b6dd9a495ddc51087c2876
0fd0d23ff8b6fe5451b61ec0e517ba914b88b150
2787 F20110216_AABIRY rao_k_Page_08.txt
3e17d454955af0892bb3280712d03eb3
218396a61d372e2ba7848dcef713b192900b466b
F20110216_AABIFG rao_k_Page_10.tif
a7e3fa58f276c48f873e13ad243c1df5
c408258983e2cd9f352c3cfdb19419d744a6f389
32577 F20110216_AABIPA rao_k_Page_60.QC.jpg
982d59922d3958b582c056f384df7e33
64ef9a974702c47c3eb1f5ce6878c9e079ba4f97
2074 F20110216_AABIAI rao_k_Page_43.txt
cecd6a4d7a6a73afe070eacba9dfd99f
9e77ab2ad31b77fb1597c04d1cb8b1f85534ffd1
3351 F20110216_AABIKD rao_k_Page_37thm.jpg
863a62d758a068ab0e6267bc4bdb205a
2bc3375b326744db6966b06d7acedc6831d4656b
1537 F20110216_AABIRZ rao_k_Page_09.txt
8afd984b48743c329de9c83976b8ddcf
ad224dc6b07e8018894c34e3414838373b19b978
8910 F20110216_AABIFH rao_k_Page_58thm.jpg
88021bb2063f5facd0910cc472561a2a
5a69a355bfa265b0d52a86ecbfd506feb405fc51
221 F20110216_AABIPB rao_k_Page_10.txt
448128825e2ff778d180e3712b8d94f5
a69fe0f727177c455f896b765ca29f816f06af04
F20110216_AABIAJ rao_k_Page_58.tif
a1ad67a5db4d2fca7f5e8e9248e61c9d
bc61b649e5d328b554a3f838cd837a8cc03a0ac8
13890 F20110216_AABIKE rao_k_Page_77.pro
d05c6bc5791c39c59bab299de6dd49d5
ba2d0527fa90de9846798d0c54df52a7b91e84ae
61110 F20110216_AABIFI rao_k_Page_47.jpg
7335bf544d23dee774d7dfaa497ebe43
563a07bcca094c4bb0907f3a3b6570e0c6a3abe5
118822 F20110216_AABIPC rao_k_Page_15.jp2
5bbb90c052ce6c4f0c2d8f974ee6bb9b
8a978acbdc141bdb455f727127107f7b4a3c3ed1
55232 F20110216_AABIAK rao_k_Page_63.pro
788cf597e709c83e32f960f6d9c52d1a
b1104ebbe54af13773fff7533fce363451856f00
21161 F20110216_AABIKF rao_k_Page_20.QC.jpg
388e6d48c9ce1383569f1b03168ddcec
da7fbeb3fd58c6a26b554cf9b44a7cef133d5b10
93209 F20110216_AABIUA rao_k_Page_46.jp2
268ac87ecd0812b22e428daa0e67ea55
dcc98c748543d6287b3db2ac5e79a2407caaa659
53283 F20110216_AABIFJ rao_k_Page_74.pro
41b5a69fa6916bc0ec2952bcd3018ab3
db77af252c9aff0a31bd47f9eeae7903ffa0e6b3
134248 F20110216_AABIPD rao_k_Page_41.jp2
1efa05841eb26b2bc104a8b2833608b7
f1370733c4e1c690e76beaaea23a945baac0cbc3
14818 F20110216_AABIAL rao_k_Page_10.jp2
26cdd59791f0ffb5a67c85c1d719ea4a
c45f4dd22dd6b94b568f8ffc7288af9bfa0877f9
19649 F20110216_AABIKG rao_k_Page_41.jpg
2c25178ace80f5d8cab3e5c4e0179856
167928c032eea8c5a2b4a4b6230cdb03f99bcfdd
565603 F20110216_AABIUB rao_k_Page_49.jp2
bee67795658dab5d5e0144d7ce286030
dca7d7bbf8d77a8029029e37c26d46cdbf5dcba8
21050 F20110216_AABIPE rao_k_Page_28.QC.jpg
bb4391da101447634300482b7a3d9c7d
d867e5e96bfb81c6f62d69cd2fad010670170210
5983 F20110216_AABIAM rao_k_Page_28thm.jpg
0693fcf6cc91f3d460c1639ec62e73d9
0e38293382ec3258737d5d9b1a3e9cc2963a4d05
99435 F20110216_AABIKH rao_k_Page_61.jpg
965283f7d71e4142f7963b566773298e
c0ad431f288015648dc94dbab3659904cc876ac9
576521 F20110216_AABIUC rao_k_Page_50.jp2
29fd550ef7ede7344b3edfb3862eacdd
7a6c32fb55c7d5eea52780df52ae379f31b6d91d
F20110216_AABIFK rao_k_Page_70.tif
89b5b8e44fd128b09774f21fbd16c12e
4e968b259b869b762d6f9e0c8162dc60f5d79965
38112 F20110216_AABIPF rao_k_Page_09.pro
ed74834d7ca67becbba1a858e4aadaa7
6dee37d2e5f460e4e18eb2166551bcd413ed4688
F20110216_AABIAN rao_k_Page_20.tif
f0225ed6357fcdf83d69fb1134dd7f74
47e87e9be0ba8aa2b48059f470c99069d6ae122a
115464 F20110216_AABIKI rao_k_Page_27.jp2
84397782ef5be21eb958d535aaca5434
addba8f1e51e6c9840ff768617be1d179e00bf4f
803867 F20110216_AABIUD rao_k_Page_52.jp2
19bae189253cd9f92e315dfa6bac680f
f37792e99bac6c80293875b0c7f15e5db67712d0
F20110216_AABIFL rao_k_Page_71.tif
1c9c4f994854badf3b7406c0eab018a5
b3877bc65d25d69a4ee6d3fa6450f18073626031
131376 F20110216_AABIPG rao_k_Page_08.jpg
252b3a7ec1d807252780938002d57d64
8389550684bafbb9836684db2a71c73ba7db972f
32742 F20110216_AABIAO rao_k_Page_44.QC.jpg
fa357e7d81a9a0e4feea2e318a2a02f3
28df6d90f302dbeb169a2cfb392cb2443a1b6bce
101816 F20110216_AABIKJ rao_k_Page_31.jpg
dc7886e4df7aa485711e95eb662dd284
61d8c9d5582b7c35eabc854f979d7380af7537c9
116080 F20110216_AABIUE rao_k_Page_58.jp2
86b037e57f3e0ce4e784dd8e7fcc3bb4
c8c16ede3516cf244002dbc54f4d7763dff77907
2071 F20110216_AABIFM rao_k_Page_03.QC.jpg
6815b4265b7f9bb5ea8a237e34b8abcf
3079e345125fbb17561fb0f047597e564baa38cd
F20110216_AABIPH rao_k_Page_06.tif
d22be28d83a8f56305fedd5163389906
5d38cb836de6cef8ce1ea3daa4dd6feac3fced59
2014 F20110216_AABIAP rao_k_Page_61.txt
363ace71fcff7559fb12b3cfd87f2c47
28cd015f3e0edb1e21b921f2e6543de06a4056fe
33745 F20110216_AABIKK rao_k_Page_42.QC.jpg
0745aa918f71191ba2d771970a42f376
c7c4cce7cb12b65cccfc426262ed8e95e3d6e99d
103972 F20110216_AABIUF rao_k_Page_61.jp2
12c244fadcd10269934a7ad6c66f8f9f
6e392aa0c97359675ac0e7b9808b381d9e2d6371
32539 F20110216_AABIFN rao_k_Page_22.QC.jpg
377090b569638909a07e4ec4e8cbf2b4
5c3679c4a6c44f6779225780de0e11b39d603330
31189 F20110216_AABIPI rao_k_Page_13.QC.jpg
a4957f0de31c5b250f73db4db35d326d
2da9c6c11461c017c3845f603d96eb4878a2ddb8
702 F20110216_AABIAQ rao_k_Page_55.txt
03ba1aa1b14c3b8dce8dc9df16a2f47b
98648aab9dde1f789c2d536dcc1e5216642163ee
3692 F20110216_AABIKL rao_k_Page_54thm.jpg
1119b5b539ed27022c5784027786c075
8d03ece7e4256491a05ec7c925ab59c8635811f5
108786 F20110216_AABIUG rao_k_Page_64.jp2
8307f9fe317b3ede24cb5544fd0c1f10
fdf427a38085aa460b22639147f413adadb35c84
8235 F20110216_AABIFO rao_k_Page_29thm.jpg
f4bf2d21146368c64d1b2a175ff5929d
9a6d0cb9d72ec2d5b6759201d0cd5a6b188c05d8
96739 F20110216_AABIPJ rao_k_Page_13.jpg
c98e68dea0497911409ae0bb6fcc6e3d
3d21eaebf55879e987a66a291bf7393a48b593b3
108599 F20110216_AABIAR rao_k_Page_63.jpg
593dfcaca8ab2fc696f5ba86685acc8a
69206e11393c99d6d3c217d5a4c1c49b68cec1cd
99170 F20110216_AABIKM rao_k_Page_34.jp2
17a6dd8792e54e6f38f12ca919146bbb
60a1c3136a7297ad28d10a98bdd95719c0c4d17d
1051862 F20110216_AABIUH rao_k_Page_70.jp2
05a9fb1155dbca13ac70305954158309
62f989cc78b391789de1d7f319c0ece1f1a1ac6d
312 F20110216_AABIFP rao_k_Page_49.txt
4b48d36e52cbdd04853ec6255487218a
1c9843f120ad870687cea0e66f7a571cb6772d82
776457 F20110216_AABIPK rao_k_Page_53.jp2
12b1f7d150109ce6060681c8479e35fa
c2baa17310f128e7705396991bfbea583b437009
721546 F20110216_AABIAS rao_k_Page_71.jp2
c338ce1abbd042cb63c8903380752671
36724b2a7e0c205d47cb67568cb32e1f6d3bdd22
30032 F20110216_AABIKN rao_k_Page_35.pro
675b44b9a49d14c9f2ee4abe21f00aa5
6dc2c0a016b8f8dbc8a0163e52a68d549346f408
102535 F20110216_AABIUI rao_k_Page_73.jp2
337a7319a31ede6c55cf8a686a8fff2c
8a349543cba9087e771571a639ba260ed0b15f1e
108640 F20110216_AABIFQ rao_k_Page_27.jpg
b083332487444479b6bc83f7f5f3cab7
8abc45d702c354cbc53cfd6768be05e9e64ce825
12678 F20110216_AABIPL rao_k_Page_38.QC.jpg
75d31e0c7979adab0f8caaa8e7d26650
3e43a287b83793ff5f0065c2d0207f25026d8e30
3063 F20110216_AABIAT rao_k_Page_06thm.jpg
bd848324976cee635baf9efcc5cec825
a2dd40a362f54f39af135ec25f703629a2d131bd
F20110216_AABIKO rao_k_Page_31.tif
40cc06cd4b92d9e38a757cf1cbf2c589
27db5eec7ee9b207b90943d1af1b79f19d7c3e93
761 F20110216_AABIUJ rao_k_Page_03thm.jpg
5c22997e34c402ad76ad710468888e64
de9336320afae5d84f9a6722a18d41a858ff0e52
1051948 F20110216_AABIPM rao_k_Page_09.jp2
94a65a4994d462885c881682676678cc
da71a652c57ec9d96955ae5f3a634774566cb019
98056 F20110216_AABIFR rao_k_Page_23.jpg
dd5a762eb332f02fd5b2e57273474e44
b8d66da92d4dfabc282cbfdae148e0a3e8f4dfe8
15857 F20110216_AABIAU rao_k_Page_71.QC.jpg
1fe0f19e5eb2bd4dde0a852b16e1d4b7
d325044ae53727a5dba88db48a04ef5c60421daa
5634 F20110216_AABIUK rao_k_Page_09thm.jpg
3f44f50aace9d8da841f9246c9ede1b7
6e941919052051b420e4bf6aa10d8e89143f7206
15836 F20110216_AABIPN rao_k_Page_72.QC.jpg
0ad69f85004818e9157dc55337133d34
2c246c85f48fa1e097c9ac155af820cf8eb6fbdc
103866 F20110216_AABIFS rao_k_Page_62.jpg
165deaf5fc969c969f11c33ddc57d2c5
ae618ed9e3218d474a48ff54b465461c44ca0a70
F20110216_AABIAV rao_k_Page_60.txt
96b25538284bf0c6b1bb3a7a30ecb8b8
cdf0dd302294b066382181041bfcda30c45d0617
8014 F20110216_AABIKP rao_k_Page_60thm.jpg
0d1277b11b4d040ffa2ea56997c1f19d
0e8a34d309fd888a51815eb29cb6fb08742ef937
8712 F20110216_AABIUL rao_k_Page_15thm.jpg
d0ae3e30ef5689cecd3247c9c41a9757
c9bb276acb497ca1f06545a3ad50937d8f1f2069
46432 F20110216_AABIPO rao_k_Page_34.pro
eac05bc053b1c298959fe34b1c8acc9c
61a9c0e179edb8247c7faf125bf3d627298b15ac
44376 F20110216_AABIFT rao_k_Page_46.pro
adda34eb5ac8318eabefbed23877c90b
835948e8e7e30aa21ae2bcb60831036626079726
108206 F20110216_AABIAW rao_k_Page_26.jpg
913bea3e2e48e918d36cbc96dc10a271
1a5ac66267d9e62849a8e547453d274d5d935ca4
6843 F20110216_AABIKQ rao_k_Page_20thm.jpg
7648ab993affb5ce8f10239025d742e8
f6e58b090fe9c79cb4ca4ef63ea6bb93f6978951
8611 F20110216_AABIUM rao_k_Page_30thm.jpg
43cd107266e609267c9efd2825f7e788
a27a2eb0a0c1f89fc844e74221721d3704d6f171
943026 F20110216_AABIPP rao_k_Page_48.jp2
26035de5a15ffa21bbaf64878a14d633
ecc757acd582f9eab8ea7dea1a9442d2c7059954
F20110216_AABIFU rao_k_Page_61.tif
d56cab06ebeb7710e85793042685e29b
9c6c9db494075e08249adc6dbf5c53140271a28e
95109 F20110216_AABIAX rao_k_Page_33.jp2
9cc0522421887db47a84025227a6cdbb
a7abfd94631557553dfad2e993fabc34ca926735
32245 F20110216_AABIKR rao_k_Page_68.jpg
b829b184c3861d824fe491e506029dde
ad90354b8e3384a35beda6c8cf9ab30273ec6870
7891 F20110216_AABIUN rao_k_Page_34thm.jpg
998a998f0e1a810619a5c5055709e0f6
c34227ce816e704306fa99a9288454865fb276a5
8906 F20110216_AABIPQ rao_k_Page_63thm.jpg
4e2d55bcace329a3105d9dedee53cc48
9da9248bdad46896c0c073d6fdd0afef90323b14
64176 F20110216_AABIFV rao_k_Page_28.jpg
97de816bc6014bf7533ea6b833c4cfe1
0db6e1f9dc5b1201d617b5c4f4c18b9ce8d6ac9f
99994 F20110216_AABIAY rao_k_Page_22.jpg
e97a9e861dc727826e2b1ecfaa251009
6d1c81ed4a5f2843e44a7250fd892110c111d7d8
50602 F20110216_AABIKS rao_k_Page_75.jpg
331a0b3cbed1102b87d3e4a47fbe2932
4a9989835d2824c95e69b3f987a648d53105a274
4095 F20110216_AABIUO rao_k_Page_38thm.jpg
eba90f71e92489009945d79d7bce01ac
2b139bb2c2b0d4ead9e117a4ddc3ac3ae854ce90
120403 F20110216_AABIPR rao_k_Page_14.jp2
23221531e51b8693857799d19422a8e1
7e76dcc4042851bb1bc14e0babb7cc044dabecf6
3635 F20110216_AABIFW rao_k_Page_55thm.jpg
f6b1b72454e4bd48bd3f726c2303960a
c7aaabe0c070975642b66c72864b8ef5db96e3ae
16923 F20110216_AABIAZ rao_k_Page_67.QC.jpg
30269c3dcc9fb73051fd03f9a09802c3
5e25361460d4be4fa8977c2fa24f727e49210c62
18397 F20110216_AABIKT rao_k_Page_19.QC.jpg
140c585726521f9a05cc4d4e3290364e
cfb73edb20d983d54aaae7d9627517c543a2979b
6096 F20110216_AABIUP rao_k_Page_45thm.jpg
245a5d479b8f80f25e0b78772faa2d93
fc697bd7adf80d7bc9fe0c2088928e9087897bcc
118329 F20110216_AABIPS rao_k_Page_57.jp2
5df3f8a38048f3db5185513b5c350699
6cccf232bd05b120535079d717dac490e3b891e4
73289 F20110216_AABIFX rao_k_Page_65.jp2
3477f4020d653b47f4c9e462a20f02d3
26bc9c72f3e99047bcb0dc386b818c1b18b50fa6
4459 F20110216_AABIKU rao_k_Page_67thm.jpg
24c23e1551dfa4545e578e911b7cfaf0
04afe1d5d8561581e8b452c3c51941166cf8bf53
4115 F20110216_AABIUQ rao_k_Page_56thm.jpg
5250ba1c7d4ff29b103a7edab0f47cbb
551ad8008fee3272ec4418416427713ee90076dd
63080 F20110216_AABIPT rao_k_Page_45.jpg
15de76877ce8d49cca6fe9a7a62d03a6
e74ca19d6b5e44178f37b590fdae4e369fc834e7
F20110216_AABIFY rao_k_Page_74.tif
26ad1b6500983addba5e074b601e7f10
792bce76e2eef6649508790942da99b1a588a748
136 F20110216_AABIKV rao_k_Page_03.txt
f12e70e51bd4bc9bbd30c97338e074d6
11faaecbba43adad5292276ef9552ec0eed07652
111742 F20110216_AABIDA rao_k_Page_42.jp2
d139b4387ed6d161ec4c936717deb907
42469fd389934bf64c10c93f4198fb9be1a5ef51
8797 F20110216_AABIUR rao_k_Page_59thm.jpg
5a9f8d662760ebf413504dcabe2cc056
4796c89e6309845c1808cdd213f074c1589eb365
8516 F20110216_AABIKW rao_k_Page_64thm.jpg
4b3901f55ddab66bbd12259bd9a47e42
e97187a2e858fb30adaeec98ad9a53667f77359e
97 F20110216_AABIDB rao_k_Page_02.txt
b259ee13cc463b4aac1b394937e22503
7fb4ae799280aa877b1d259d08b8887fd76aa4b1
101231 F20110216_AABIFZ rao_k_Page_60.jp2
45c6692addc26ea424a5fe50585613e5
2637b69c31524fe26d27a734af5732232e7bc911
8663 F20110216_AABIUS rao_k_Page_62thm.jpg
9ac39ffbaa1ac72e9a52f2d7dc91d59b
b1cef984903ec5d9f2551d0c8bb7012744c1243b
799 F20110216_AABIPU rao_k_Page_68.txt
7bf2278c40bedf14db904c654082eed0
b09326f79bf89827d995b3a1ba1ce6fb82631a09
3651 F20110216_AABIKX rao_k_Page_53.pro
78f63df0301c6764dbd7c9ae3abf969d
c062cf05a7882377a0aa472d1a7c354cad37c706
34993 F20110216_AABIDC rao_k_Page_77.jp2
ef39f7d83bf6372b24adbe9b43bf1731
b4feeb77faa84b3e0f753861db082db3e42ecf7f
5550 F20110216_AABIUT rao_k_Page_66thm.jpg
21b5f3a2af090627a16a249987aa9ec2
1b6b92ab576a1e7c2a15023d984b24f9b86abf86
55290 F20110216_AABIPV rao_k_Page_67.jp2
161864372bd41f44f8eca48c6f495165
90acbec7630f26557fcba91bc5d8fca7dfe67da5
F20110216_AABIIA rao_k_Page_47.tif
fd956477db67fb8a3e0218ca84c0033c
25f290ee1f3615cbbc201dc101083250d89e5592
12508 F20110216_AABIKY rao_k_Page_72.pro
66ec9227c4e3c0a6fefe62091d04f2f4
45f748bc889c906f9a24cfb6fed254fedcaf9650
105072 F20110216_AABIDD rao_k_Page_44.jp2
6af107c8b0e8710dc2f3d3689d712826
4a5712f449e1e72346787efdd652c8b11804d816
3856 F20110216_AABIUU rao_k_Page_68thm.jpg
89b3294fbd23871f636c2ffa489a32ab
0fb6fc7a0953e6a3ac06a51468bfffc12a263fe5
9679 F20110216_AABIPW rao_k_Page_50.QC.jpg
8a28b179bc19137e363f52b0b308db31
6a3a06a5b7cde27084ccb26fff1db8a4cea44ed8
55557 F20110216_AABIIB rao_k_Page_59.pro
8e261a9c7fc6211a12bc237592447ae6
0d9a3013b3302683de322d26025596e119f7a44f
439586 F20110216_AABIKZ rao_k_Page_37.jp2
077049c61ad3fe40145c24aa3ba70a57
bcce55f46c79f6e4ec28750b3aa0c93fa2c0f5b5
570419 F20110216_AABIDE rao_k_Page_51.jp2
46001dfd89136185c0a8b2d9cbf1eb4e
acc58db6449f1a1848acb6c0d9fbde0c27ee1d35
5105 F20110216_AABIUV rao_k_Page_72thm.jpg
22af7d38d8a7fc507b08d421461226c7
b97b1365616592b332467a74c06b357725459206
2072 F20110216_AABIPX rao_k_Page_33.txt
5ad45bdb5c79ea4d2c53068babe919e7
0c012f3aaca638f95748963491a48650e057a2a2
6117 F20110216_AABIIC rao_k_Page_40.pro
a0a2bdd195df63ece40a50803e48f292
9a8a47e1dfab39d16eb90e8da5f31d18d66fae73
30282 F20110216_AABIDF rao_k_Page_43.QC.jpg
9e32d7a393de4327dd94b203b55e58ed
452f95d22cf84789413660dc084f62d10151f8be
7585 F20110216_AABIUW rao_k_Page_73thm.jpg
66cb9d1ff7947e443e14e7a568802c7d
963245406b14c3504047da0a2862091bc26bac3a
24829 F20110216_AABIPY rao_k_Page_05.QC.jpg
2cfa8e4caba5b717e9455a54de3e3403
a919079d8f260ce3df79f60d0fefe3ded9cbac26
3961 F20110216_AABIID rao_k_Page_02.jpg
7b0f4f1cc7bbeb469d520de5203004ba
642036d6c63246e46ff23e95e924b703e3443d69
44067 F20110216_AABIDG rao_k_Page_43.pro
04c885b621d3908e9efeae42fdb45bbf
e0a0332401d65e034c89ad958b62cbdc265c1ff5
117620 F20110216_AABINA rao_k_Page_16.jp2
494376e7ae35ea5d3745dadbae4f9716
a01be153d5d87a4ec7f8bdb33497e7fdd7aa133d
90980 F20110216_AABIUX UFE0025106_00001.mets
b47237af0873c9cbc98d7a7f2fe4f43a
b0a9ac8aa6013ac6b678d9bf5cd3c1b998be508a
2276 F20110216_AABIPZ rao_k_Page_57.txt
06723906ad6d22ffc78bf25f66917277
bfce74b32949e2056a0eee6067de55b67ab464a8
73858 F20110216_AABIIE rao_k_Page_66.jp2
2ecb1afdc6232b825495dd1dca8e99f0
26284b2a535df1701b54d6144bf26c1ff757a445
42133 F20110216_AABIDH rao_k_Page_56.jpg
288d3529d096fa2de1b63b6537ad2d09
0f2903bef6c2fa810cefa4c29dcb7518cee90c0f
47874 F20110216_AABINB rao_k_Page_73.pro
93e7465cb9e498308fae9abfde51ade2
d19bc4027edd0f4dfc2727432de1ec94fb9f3a16
12033 F20110216_AABIIF rao_k_Page_04.QC.jpg
33c66e0f369baa78d2d148816f5ade0f
f48844c6a1f781e9ca902ccd721e799887598c2b
8132 F20110216_AABINC rao_k_Page_22thm.jpg
4b882aa53506b9e07037b71069c35456
473bc7fa79b99baf07ac78f447653050dd866e65
2102 F20110216_AABISA rao_k_Page_11.txt
31c3ecd75d93394dcd3426b4b442152a
d650eed4d7fa99a567d9e8e9bba4b12525d8559a
F20110216_AABIIG rao_k_Page_32.tif
8bcaffc2917e901f4433f3dd9d97db73
e28a0064e7fe40fe90bad809033b2df3e1ca0f8c
4229 F20110216_AABIDI rao_k_Page_39thm.jpg
35dfd00939358621155e767b473a7db6
f885d1b4eed05cd529478e055819864d17883dda
2718 F20110216_AABIND rao_k_Page_77thm.jpg
cb3c453da4aa8927b71cd079fe6a87c1
0d54713db64317ff0a4ac2f719c0dcd49e8d4a26
1641 F20110216_AABISB rao_k_Page_12.txt
5b9dee3c62117d06d8df155300b9c4d9
02a0ce15bd4e2d73a8cacd4404db1e1c2cdc1a21
F20110216_AABIIH rao_k_Page_11.tif
e56f09520fedefd9f49d273fcfdf4ce8
3886536669c2893f7fd9a30323f2e325239e3822
1471 F20110216_AABIDJ rao_k_Page_10thm.jpg
718e405eb743152a4261a154f1159403
846d757e2fd4c1ab1a96efd96905a3f3e09dd5fb
17255 F20110216_AABINE rao_k_Page_69.QC.jpg
abdcf068931d2e62eb9516d3efb62f77
a3f1d36f658f75baa9d7371dfd32e8acd03ea369
302 F20110216_AABISC rao_k_Page_19.txt
d661342e35e7abaf91ebc166d80762da
9753648d5c574f03a0cacb9131c8c5d22fb30002
F20110216_AABIII rao_k_Page_28.tif
432532144a0321d23ccdee410c730cf9
1366c6337f5de546ba5340f5278b906bea7a4892
98343 F20110216_AABIDK rao_k_Page_44.jpg
4c341effb8e6ede0185203ae653483cd
df9f4e25a22040a62ca44df5374c85ad6a3e00da
F20110216_AABINF rao_k_Page_49.tif
92a378af6f05146303731c4fb6e306c1
b1a310f3beac41fb4313b40c95d34f40ae3d0cc1
F20110216_AABISD rao_k_Page_26.txt
90f7ed0c3731f59f16c6cc4688f266cd
68ac8e2943b1fbc0ea6d46e45c3637014df018dd
373546 F20110216_AABIIJ rao_k_Page_36.jp2
ff139d8b24407147cabcb9ade46a738e
6a00c9ee1cedc697dc3a1e3d6fa911f65e34fefb
1051967 F20110216_AABIDL rao_k_Page_05.jp2
80e6c4043b9f8bc44fc7df857f36553f
f9fca6b92a0a97967631c1ccdf2d9cc8e33434d0
120596 F20110216_AABING rao_k_Page_17.jp2
a74d590f08ad13b86a4a04273abf439e
5c65803e0e25fe42d44b9c9147e4d3735bd8b221
2224 F20110216_AABISE rao_k_Page_32.txt
76530d1437105b42441bf44bd16dc454
c8c41ccb7006fa745cc922c75d3cee70dd344ce6
1171 F20110216_AABIIK rao_k_Page_47.txt
a539c49e0e5fbc615845f60ded19c0bd
dc5cb1b445f396ef5459dfe1c76c53ea2e887ae1
36755 F20110216_AABIDM rao_k_Page_58.QC.jpg
0647438c658d35cd7a850ff1d0c4c85d
5171c712976873a58c07244ee1ac6f846387aac5
14339 F20110216_AABINH rao_k_Page_36.QC.jpg
327aad15021e0c396aedf5c2aeeb0d10
496806a6ca2150a29da7a4a629e5f0e67431ba67
267 F20110216_AABISF rao_k_Page_37.txt
4ec59075926609d10fefd9ec1db3e4d3
cd4c6eaf704cdb305df193e7e22737d35a731854
F20110216_AABIIL rao_k_Page_08.tif
c7f3fe84b4974ee8e1dd41591fb238e1
6edf341d0b41de6048d6fe6aab0c51c5e28b5628
F20110216_AABIDN rao_k_Page_17thm.jpg
4515eef43ad65cb2448f37d99a03dee2
dd152223fcd71e6857d5596c61dbad7ff57e2972
7633 F20110216_AABINI rao_k_Page_33thm.jpg
6e787192b6d1457211646cd6a6ebae21
4621253fba075347dcedb0ee01ab332e2e1bdd9b
1867 F20110216_AABISG rao_k_Page_46.txt
8582665a83a5c21183c9c1d7170cbd82
875de7cbf24205ef66c2e1d77a4e918d3258a348
112511 F20110216_AABIIM rao_k_Page_14.jpg
3ac29b4b28c7042754989412fbd3af52
2ecf2925b0900548ea10a4a871a32bbe8022c529
32381 F20110216_AABIDO rao_k_Page_51.jpg
79bc185802468092c882599dbae203dc
db58e970c335da32ab63ddc9f86229efa08aa26c
28080 F20110216_AABINJ rao_k_Page_46.QC.jpg
f8bbe47c21b12199d97670e99ab6628a
14ac051058b4525367d53afec7e55f122a172eac
320 F20110216_AABISH rao_k_Page_50.txt
087fb387380232ab6e509f1f0968865d
16cef6d0cc834d79991ff52e26549451ebcd67aa
503 F20110216_AABIDP rao_k_Page_69.txt
b40d28bec6bd1538b43f462b62d291b6
731f9ce5b599e1801a72fd3039586cfcde4af1cb
F20110216_AABINK rao_k_Page_34.tif
6fa99852906f732fa759538b0fb8d208
b0f2311fc620ef321bf06df69921095c589ab004
448 F20110216_AABISI rao_k_Page_54.txt
ddd9cbd120db708f9a4dd3abe200e3a3
85f93617a3f5836bb78bff5a43645e44f06a6dad
F20110216_AABIIN rao_k_Page_37.tif
40ce2a93bc1e05caac20d0542f004b10
07d4f79e1e309a33dfdabd7049917368164d0ce3
50292 F20110216_AABIDQ rao_k_Page_31.pro
0faac234367d99203f09321b3d42dff6
53faf46c7b2f36f157ffd2c0ebaca7ce8be61e60
13259 F20110216_AABINL rao_k_Page_53.QC.jpg
602168d494bf6761bdf002f165fc28b9
8ea696389e9895b39a1879653489aa0d95c17f7f
2222 F20110216_AABISJ rao_k_Page_58.txt
2dcf5a21216fed954baae45c60589337
9e0be9ac3c08e6d86612c808e88ba23cea102716
F20110216_AABIIO rao_k_Page_24.tif
ba5e3637bbbdd0f0a86e29b8df397ca6
b43f2f47ac2e485686c54661624cddc3ca92adc7
33233 F20110216_AABIDR rao_k_Page_50.jpg
96e48bd1383541ce529f34be1fa4b684
45cb0cce42d70faf5773571e99344ca017e2fe7a
1168 F20110216_AABINM rao_k_Page_02.QC.jpg
b280afb16d341559083fe98014466314
d097b8aaf4b22a8bb9d35e93e0bf1dc6d4e99648
2439 F20110216_AABISK rao_k_Page_76.txt
5e7b7602d6a4390008a79e4e9d6d162e
f7110c67983303d9716606eab7353d70d0871fbf
6819 F20110216_AABIIP rao_k_Page_12thm.jpg
6833e457e603ddad8cece5d2e03aec48
26aed33aed948c92dbb71fe10bde9999bbb6ec7a
19259 F20110216_AABIDS rao_k_Page_70.QC.jpg
a2937f9064a9fa5dcf2ea57be6fe6c93
7900dcc8571ef07c0fca3dd6de2a2070264b82f6
113549 F20110216_AABINN rao_k_Page_17.jpg
9cb79b1db52bdacf9c18154b044b73ab
a1764ea4039ba017cf33ee02b66bf925d208a132
595 F20110216_AABISL rao_k_Page_77.txt
632ef7a211a3a351a333e052bc032ed2
137338b294647e664bf1850a8fe87d42b901267a
26086 F20110216_AABIIQ rao_k_Page_28.pro
c352c4d97017da0b1c4c0281ead77d36
006fbc0e902d6ab7a7643f615211de1d5d3d9ad7
F20110216_AABIDT rao_k_Page_59.tif
551bc2ad454d8751a85796a046f9c208
e364959ca86e1d2a9bb5f88699e3a66156e108aa
6607 F20110216_AABINO rao_k_Page_05thm.jpg
30b78c19077d99d237cc6e492e4372be
b3140ca3a03073e029d0e90eaea5a4da716e7b3d
852 F20110216_AABISM rao_k_Page_02.pro
7d65ee25acc5d8701f188f7a7fe41e39
f5008d04fb1d9450081a1bfe63e2b534476a9cb2
5462 F20110216_AABIIR rao_k_Page_55.pro
16ddf6cce34a2e1332a2c161fc7b0a30
3d9df85630512ca635e5be0c2aa7a0e256a95242
F20110216_AABIDU rao_k_Page_39.tif
60ab5375cc45bd1b91712f71cde63cd8
a3b37be98ab21c03c2e5742893c70fc622a756eb
7941 F20110216_AABINP rao_k_Page_03.jp2
a96fc089b6a6b75d43f1d89838b84ec1
c03d6f0b2d5e70b1ce05928a7e7f55b1d9a70c34
57259 F20110216_AABISN rao_k_Page_14.pro
45afee4eb9abe4fe2bb856ceab2356d7
f945d56c57b675df29d496ebe9759a87012afb7b
691992 F20110216_AABIIS rao_k_Page_56.jp2
7baab8641f226cda934120351413a477
b84fbca5a21bc91e7ba4b61cb1ac98738e244d81
109903 F20110216_AABIDV rao_k_Page_62.jp2
7b5539f02a59c1e3dd94c966033727b0
e4af8e793f5c2658842d22b10ff9fa3b33d182a1
F20110216_AABINQ rao_k_Page_44.tif
00b176c86311a14083e2f8fc6ef45d3c
47d473ecb02d0cf122b785840e5f990b0ea831b8
57557 F20110216_AABISO rao_k_Page_17.pro
447e8c6ab7fc9e1649833f5e0faac9ea
92110ae67ecf2699d991f205408e140a03b56691
429 F20110216_AABIIT rao_k_Page_20.txt
868fa1893feb0ff903b451c13239b675
a23a552d5231880acfe014753d89904fefdd2fcc
F20110216_AABIDW rao_k_Page_03.tif
8c1e1e8fe6f78b020f32a3c8864d35e6
da4db1678e926a0699274b74f801612a9da01e42
95150 F20110216_AABINR rao_k_Page_43.jp2
aeb92143fe55565e111d9be5a94facd0
202573fad56748d63039611d03832889338fdd9c
47017 F20110216_AABISP rao_k_Page_25.pro
3f0985f1915883bbcd13279d38a16bae
295aa47aff01107c60f6e80a22c26002c3c6e758
35672 F20110216_AABIIU rao_k_Page_27.QC.jpg
d949bc6738ff001bbc5c49d0797a7494
d347aac83c80f5e8016a85593a21a68e19cf4437
6105 F20110216_AABIDX rao_k_Page_37.pro
54ac73ff6a119bbb99d6126592b669a8
e146522d673e7fa9bc3791554c0f7468e402c2f1
54030 F20110216_AABISQ rao_k_Page_26.pro
f693f120c5f5fe8cf3c074af211a9588
0d8018c02a43a3952b7ddb10302b9b2b72053bfa
4566 F20110216_AABIBA rao_k_Page_10.pro
eb2a7d2cf594e771933c1f07e2c0a6eb
42af8079d818e2be12e949d5b1e9e5643a581ee1
F20110216_AABIDY rao_k_Page_29.tif
359bfd9ec4471856443f40b60ca98fa3
d9563c06040b9a80af11f29e168a81ae34abeab2
F20110216_AABINS rao_k_Page_68.tif
b2d2d5b202ec7405e732dbc9526b0735
e9429d0de60cbfcfb8d8e759ffe05f757f034d9c
35459 F20110216_AABIIV rao_k_Page_74.QC.jpg
841f7412b916c3346ec1b63af373f3ce
3b955dbf45ff9331829fc345172294f0090cdd50
55208 F20110216_AABISR rao_k_Page_27.pro
87ead15f6796afdbd69ce8a71463caa5
3e122259cc8d18c9abe483a80404e9bd82ff76d8
1051986 F20110216_AABIBB rao_k_Page_08.jp2
aeef5570b9a5725f8ee0ffb4b9ad880b
54538e7fb2ef4692f2ef0a360613930944ed9f86
792800 F20110216_AABIDZ rao_k_Page_20.jp2
fec45fca060abdaea14b49b671fc0717
78d05294f6d628a4db625b5ab8b0723820efa959
112931 F20110216_AABINT rao_k_Page_26.jp2
82a7d8c6920d19f321e5ad8331548e3d
11b8692236aa87f0bde7e3747b8a084cfa73e0a2
47252 F20110216_AABIIW rao_k_Page_60.pro
5fc094ce97594c28d3ddfb5282901c15
22af64ef3bb1003c9e093e15f8bc225660e9a4d9
52027 F20110216_AABISS rao_k_Page_30.pro
88cf50f08cd5a7d27528bf508d8e2cba
50cad19b691f8f9f17bcae04dcc7863173e472dd
F20110216_AABIBC rao_k_Page_41.tif
cae87fb0476d4c2080792dc9f22488a9
ae8173f0fb8475abccc51a26ee00a754b76cd398
113836 F20110216_AABINU rao_k_Page_74.jp2
738b584397dfcb78c7876680c2a4842d
e81e6ce1b8f376e61a27f12b95600ed603fc1ed1
3503 F20110216_AABIIX rao_k_Page_41.pro
6f51d7dd84b4400dd9d4b5081c63186c
7f35e0411393658cc8aa8cd797185a4b1e232704
30027 F20110216_AABIST rao_k_Page_45.pro
9fd556349941281c9c911ac5bbbb7dab
fcfb9d042a7d3496d02763075e93bb6aadc54dd3
7727 F20110216_AABIBD rao_k_Page_25thm.jpg
8e8c9557807400524225b827689035a5
80bd014f919e6dee047f042b3bba72a438235976
39197 F20110216_AABINV rao_k_Page_04.jpg
6a4250f7903e3de276c9d9d87b71e403
cbead0b24e6ffe310f53f968db92bfdfd5353a47
8799 F20110216_AABIGA rao_k_Page_24thm.jpg
73320e14f8c1503919b2f4108a80ff80
c9d8031b936770daaf4e0afc474828db4fe06154
2064 F20110216_AABIIY rao_k_Page_44.txt
24c3fcb8a9c767fb233fb08113b22e0a
fb047b3d282180ad79ee83a0062bd10eb1cfd122
10006 F20110216_AABISU rao_k_Page_70.pro
6dcd90591a4326538e205640091b6ea4
eb107d9cb1af5a2e245edb44143684fb32438d1d
69022 F20110216_AABIBE rao_k_Page_18.jpg
8ce5d714ffede2d4517bdd326ed7dfa0
0d0ef755cdb97299373f920c66848e33c4e1fad4
355810 F20110216_AABINW rao_k_Page_38.jp2
56206fa5bdee85dd42da64521cb476c6
2fabaff9133d6c61dcbb3c97f154160d5392e259
1327 F20110216_AABIGB rao_k_Page_18.txt
c37a8491f1369b3558f8ba7bfe05b03b
a13a6bd32e78b802eda8191e61dd2393b95f284e
33733 F20110216_AABIIZ rao_k_Page_77.jpg
909ed3775382eb01034efe8c044f6da5
01f2f45ad3d88976466f2c0f08df83cff77aa42d







DESIGN AND ANALYSIS OF FIXTURING METHODS FOR MESO SCALE
MANUFACTURING




















By

KOUSTUBH J. RAO


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA INT PARTIAL FULFILLMENT
OF THE REQUIREMENT S FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2009






































O 2009 Koushtbh J. Rao
































To my parents and my sister
Thank you for always being there for me









ACKNOWLEDGMENT S

I thank my advisor, Dr. Gloria Wiens, Associate Professor, Department of Mechanical and

Aerospace Engineering, University of Florida, for her continuous support and guidance. I would

also like to thank my committee members Dr. Hitomi Yamaguchi Greenslet, Associate

Professor, and Prof. John Schueller, Professor.

I express my deep sense of gratitude to my family members for their perennial moral

support and encouragement. I also record my special thanks to my friends and lab mates (Space,

Automation and Manufacturing Mechanisms Laboratory and Space Systems Group) for making

my stay at UF very memorable.












TABLE OF CONTENTS





ACKNOWLEDGMENT S ................. ................. 4...............


LI ST OF T AB LE S................................ 7


LIST OF FIGURE S ............... .................... 8


LIST OF ABBREVIATIONS ................. ................. 10......... ....


AB ST RAC T ................. ................. 1......... 1.....


CHAPTER


1 INTRODUCTION................ ............. 13


1.1 Background .................................... 13
1.2 M otivation ............... .................... 13
1.3 Focus of Research .................................... 14
1.4 Literature Review................ ................ 16
1.5 Thesis Outline ............... .................... 18


2 THEORETICAL FIXEL MODEL AS SUMPTIONS ................. ...............21...............


2. 1 Problem Statement ................. ...............21................
2.2 Assumptions .................................... 21

3 PRINCIPLES OF THE DESIGNS ................. ...............23........... ...


3.1 D esign 1 ............... ... ............ ................ 23
3.1.1 Introduction to Compliant Mechanisms ................. ...............23.............
3.1.2 Pseudo Rigid B ody Model of a Compliant Mechani sm ................. ................ .. 24
3.1.3 PRBM of the Fixel .................. .... ... ... ......... ... ... ..... ..... ..........2

3.1.4 Derivation of Stiffness of the Fixel at Contact Point (Along y-direction) ...........26
3.1.5 Derivation of Stiffness of the Fixel Along x-direction............... ...............30
3.2 Design 2 ................ ... .......... ..... ... ......... ... ..........3
3.2.1 Design 2a) Mechanical Variable L ................. ...............31.............
3.2.2 Design 2b) Mechanical Variable h ........................... ........33
3.3 D esign 3 ............... .................... 34


4 COMPARISON OF THE DESIGNS............... .................42


4.1 D esign 1 ............... .... .. .................... 42
4.2 Compari son of Designs 2a, 2b and 3 ................. ...............43.............
4.2.1 Design 2a ................. ...............45........... ...
4.2.2 Design 2b ................. ...............45................












4.2.3 Design 3 ................ .... .. .._. .............. 46...
4.2.4 Plots for Design 2a, 2b and 3 .............. .....................46


5 ANALYSIS OF THE FIXEL DESIGNS ............... ....................57


5.1 Static Analysis of Design 2a ................... .. ............... 58
5.2 Analysis of Design 2a Under Harmonic Force ........................... ........60
5.3 Compensation for Transient Dynamics ............... .................... 65


6 DISCUS SION AND CONCLUSION S ............... ....................73


6.1 Conclusions ................... .. ............... 73
6.2 Directions for Future Research ............... ....................74


REFERENCES ............... ....................76


BIOGRAPHICAL SKETCH ............... ....................77










LIST OF TABLES


Table page

5-1 Overview of the runout and length values for various trials obtained from the FE
A naly si s ............... .................... 66

5-2 Calculation of the stiffness values corresponding to different spindle speeds .........._......66

5-3 Calculation of the lengths of the compliant beams corresponding to the beam
stiffness values for Design 2a............... ...................66

5-4 Calculation of the widths of the compliant beams corresponding to the beam stiffness
values for Design 2b ............... ....................67

5-5 Calculation of di stance of point of application of force from the fixed end of the
compliant beams corresponding to the beam stiffness values for Design 3 ........._.._..........67










LIST OF FIGURES


FiMr page

1-1 Conceptual Integration of Fixturing and Mesoscale Machine Tool (horizontal plane
of application) (mMT stage courtesy of J. Ni at S.M. Wu Manufacturing Research
Center, University of Michigan, Ann Arbor, MI) .................... ............... 19

1-2 Conceptual Integration of Fixturing and Mesoscale Machine Tool (vertical plane of
application). (mMT stage courtesy of Thomas N. Lindem at Atometric, Inc.,
Rockford, 1L) ............... .................... 20

1-3 Illustration of run out error ................. ................. 20......... ...

3-1 Design 1 Simple monolithic compliant four-bar mechanism consisting of four
compliant j points ................. ................. 36......... ....

3-2 Example of a C ompliant Mechani sm a Crimping Mechani sm consi sting of three
flexible members (courtesy of Compliant Mechani sms by Larry Howell) ................... .....3 6

3-3 Pseudo Rigid Body Model for the Four bar monolithic fixture ............... .................... 37

3-4 The various configurations of Design 1 during stiffness calculations a) Initial
Configuration b) Symmetric Equilibrium with Base Length L1 c) Displacement Due
to Contact Force with Base Length L1 d) Change in Displacement Due to a Change
in Contact Force with Base Length L1 ............... ....................38

3-5 Design 2 Principles Involved ............... .................... 39

3-6 Implementation of Design 2a with Four Fixels ................. ...............39........... ..

3-7 Implementation of Design 2b with Four Fixel s. ................ ...............40........... .

3-8 Design 3 Principle Involved ................. ...............40...............

3-9 Implementation of Design 3 with Four Fixel s. ................ ...............41........... .

4-1 Fixel Stiffness (L2=L4=45mm, L3=30mm, all k' s are equal) .................... ...............4

4-2 (km;,, /Eb) for Designs 2a, 2b and 3 versus [r,s] ................. ...............49...........

4-3 (km,/ Eb) for Designs 2a and 2b versus [r,s] ................. ................. 50..........

4-4 (km,/ Eb) for Design 3 versus [r,s] ..........___...... .__ ...............51.

4-5 (Ak /Eb) x 10-' for Designs 2a and 2b versus [r,s] ............... .................... 52

4-6 (Ak /Eb) x 10-' for Design 3 versus [r,s] ................. ................. 53.........










4-7 (Average Slope of k /Eb) x 10-4 for Design 2a versus [r,s] ................. ....................... 54

4-8 (Average Slope of k /Eb) x 10-3 for Design 2b versus [r,s] ................. ....................... 55

4-9 (Average Slope of k /Eb) x 10-' for Design 3 versus [r,s] ................. ....................... 56

5-1 ProE model for fixel- workpiece configuration for a milling operation with two
compliant fixels and two rigid fixels. ................. ................. 68......... ..

5-2 ProE model for single fixel workpiece configuration............... .............6

5-3 FE Analysi s to calculate the maximum di splacement of the compliant beam .........._......69

5-4 Variation of length of the cantilever beam for corresponding runout values ........._.._........69

5-5 MSC Adams model for the two fixels-workpiece configuration ................. ................ ..70

5-6 Plot showing the Di splacement of the center of mass of the workpiece with respect to
time where the frequency of the harmonic force is 60000 rpm................ ...................70

5-7 Compari son of the di splacement of the center of mass of the workpiece and the
expected runout values for a frequency of 60000 rpm ............... ........ ............71

5-8 Plot showing the error between the displacement of the center of mass of the
workpiece and the expected runout values in the transient stage............... ...................72









LIST OF ABBREVIATIONS

mMTs mesoscale Machine Tools

FIXEL Fixture Element

PKM Parallel Kinematic Mechanism

PRBM Pseudo Rigid Body Model

TIR Total Indicated Reading









Ab stract of Thesi s Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

DESIGN AND ANALYSIS OF FIXTURING METHODS FOR MESO SCALE
MANUFACTURING

By

Koustubh J. Rao

August 2009

Chair: Gloria Wiens
Major: Mechanical Engineering

The workpiece-fixture and the tool-workpiece interactions are important factors for a

manufacturing process and play a maj or role in the efficacy of the process. They lead to non-

beneficial machine performance in terms of inaccurate dimensioning and improper finishing of

the workpiece. These interactions become even more important for mesoscale manufacturing

because of the higher accuracy requirements. In addition, the type of fixturing used on mesoscale

machine tool system sets a limit on the ability to achieve high accuracy of dimensions. Fixturing

also remains a critical issue impeding the integration and autonomous operation of

micro/mesoscale manufacturing systems. To push beyond these accuracy limits, innovative

work-holding approaches are needed. This study presents an investigation of four fixture element

(fixel) designs for fixturing to be incorporated into mesoscale manufacturing systems. These

fixels help in improving the process accuracy by taking into account the effects of tool runout.

Using compliant mechanisms and components (e.g., monolithic four-bar mechanisms and/or

cantilever beams), fixels exhibiting mechanically adjustable stiffness characteristics are

achievable. Manually or automating the stiffness adjustments, these fixels provide a functionality

for enabling greater control of the response of the workpiece due to runout and variation in

contact forces at the tool-workpiece-fixture interface. The method of implementing these fixel










designs (using four fixels) for common manufacturing operations is suggested in addition to the

designs. Each design has specific mechanical parameters for its fixels and adjustable stiffness

characteristics are achieved by varying these parameters. The monolithic four bar design has two

mechanical variables whereas the three cantilever/compliant beam fixel designs require actuation

of only one mechanical parameter. To quantify the fixel functionality and its dynamic range, the

theoretical models of the stiffness characteristics expressed as a function of these mechanical

variables are presented for each of the designs. Upon establishing a common stiffness range for

the three compliant beam fixel designs, a metric is formed for better comparison between the

designs. This metric is based on the sensitivity of stiffness expressed as a function of slenderness

ratio and an operation range, bounded by a minimum possible stiffness value shared by the

cantilever beam fixture models. The slenderness ratio is defined as the ratio of the minimum

length of the cantilever beam to its maximum width while the operation range is determined by

the ratio of the minimum and maximum possible values of the fixel' s mechanical variable. Using

this metric, results are generated for each of the designs and then compared with one another to

highlight the advantages and disadvantages of each design. Models for one of the fixel designs

are then developed to perform dynamic analysis and understand the behavior of the fixturing

design under manufacturing operation conditions.









CHAPTER 1
INTRODUCTION

1.1 Background

The growing trend towards miniaturization has impacted technologies in virtually every

field, from medicine to manufacturing. Consequently, a dramatic shift is occurring within the

manufacturing paradigm toward the development of complementary capabilities for producing

miniaturized products. Furthermore, this shift has led to research efforts on micro/meso levels,

bridging the gap between the micro and macro worlds. The creation of mesoscale machine tools

(mMTs) that are less expensive and more portable than conventional precision machine tools is

an indication to these efforts.

However, mesoscale manufacturing is still faced with critical issues in accuracy due to the

adverse affects resulting from tool-workpiece and the workpiece-fixture interactions. These

inaccuracies could be due to runout in the tool, mi salignment of the tool/workpiece, workpiece

material properties, etc. These challenges are magnified for the complicated process of creation

of micron features on micro and macro-sized parts, where fixturing and material handling pose a

problem. This is an area with significant importance, yet there has been limited research on

addressing these problems. The obj ective of this study is to contribute towards research on

mesoscale manufacturing by introducing fixturing designs which can facilitate controlling the

tool-workpiece interface dynamics.

1.2 Motivation

In prior research related to mesoscale manufacturing, designs for reconfigurable

manipulators consisting of compliant elements and actuators for part positioning and handling

have been developed. These methods and devices typically manipulate components that are

adhered to the device or a base material and do not have the ability to directly control the tool-










workpiece interactions. By developing fixturing devices which have the ability to passively or

actively control these interactions, the repeatability and precision of mesoscale machine tools can

be enhanced. In the context of controlling the tool-workpiece interface dynamics, mechanical

adaptability is a key factor and can be of advantage to the current situation. The feasibility of

mechanical adaptability at the mesoscale and macro-scale for fixturing has been successfully

demonstrated but has not been implemented. For the work reported in thi s thesi s, the premi se of

the active fixturing can be stated as follows. If fixturing methods which are capable of providing

stiffness adjustments are developed, they will enable greater control of the response of the

workpiece due to runout and variation in contact forces at the tool-workpiece-fixture interface.

Consequently, it is necessary to develop designs and methods of implementation for

mechanically adaptable fixturing devices comprising of fixels (fixture elements). Such a

fixturing system is illustrated in Figures 1-1 and 1-2, consisting of four fixels and integrated with

a mesoscale machine tool. In the system illustrated in Figure 1-1, the fixels are in a horizontal

plane whereas the plane of application of the fixture designs shown in Figure 1-2 is in the

vertical plane. In such a system, gravitational effects will also play a maj or role. Once fixel

designs are developed, a method to compare them would be required along with the analysis of

the designs to validate them.

1.3 Focus of Research

This thesis focuses on designs for fixels for passive/active fixturing which can provide

mechanically adjustable characteristics. In this thesis four candidate designs are presented for

fixturing mesoscale workpieces. The first design consists of monolithic compliant four-bar

mechanism type fixels. The other three fixturing designs consist of cantilever beam type fixels.

These designs must be evaluated in their ability to control stiffness, both in range and direction,

via a minimal set of mechanical adjustments. To quantify the fixel functionality and its stiffness










range, this study presents the theoretical models for determining the stiffness characteristics at

the point of fixel-workpiece contact. These stiffness characteristics are expressed as functions of

the mechanical variables of each of the fixel designs. Through variable stiffness characteristics,

the response of the workpiece due to runout and variation in contact forces at the tool-workpiece-

fixture interface can then be better controlled. In order to determine the capabilities of these

fixturing methods, the modeling and analysis of candidate fixel designs is done.

The stiffness adjustability of each of such fixels can be integrated in a coordinated manner

with other (adjustable) fixels. The fixturing system formed by combining these independent

fixels with the workpiece takes the form of a parallel kinematic mechanism (PKM). In this PKM

setup, the fixels act as the independent kinematic chains and the workpiece is the end-effector,

thus forming a closed loop system. Leveraging these similarities with the PKM, passive/active

fixturing also has the potential of providing a dual-capability of fixturing as well as manipulating

micro/mesoscale workpieces [8], [9]. Furthermore, via stiffness adjustments of the fixels, the

fixturing device will have the ability to tune the kinematics and dynamics of the mesoscale

machine tool and control the tool-workpiece interface dynamics. These characteristics can be

used to reduce the errors caused by tool runout to improve the accuracy of the process and

reduce the effects on the tool properties. This stiffness relationship can then be applied towards

development of efficient control algorithms implemented through actuated adjustments of the

mechanical variables of the fixels.

While this thesis focuses on the mMT application, it should be noted that the methodology

presented herein is applicable to both macro and mesoscale manufacturing. Furthermore,

mesoscale manufacturing is performed using various machine tool sizes that vary from small to

medium sized desk top machine tools to the more traditional sized machine tool.









1.4 Literature Review

As mentioned in the previous section, there has been an increase in the research on

developing more efficient and accurate mesoscale machine tools (mMTs). Some of the mMTs

that have been developed or are being developed are less expensive and more portable than

conventional precision machine tools. This is achieved by developing a thorough study of the

dynamic behavior of the mMT [1], developing different methods for their calibration [2] and

evaluation of the performance of the developed systems in mesoscale processes.

There has been prior research done on better understanding the process parameters that

have a large effect on a mMT system and estimating the optimal operation conditions for such

systems. The parameters such as depth of cut, magnitude of cutting forces, feed rate and spindle

rotation speed impact the performance of a mMT system. The spindle speed is usually high for

such systems and is in the order of 60000 rpm (with certain processes reaching speeds up to

120000 rpm) while the cutting forces are usually in the order of 10 mN [3]. Any design of a

mesoscale system or its components should be developed and analyzed for these values of the

parameters.

In recent years, research is being carried out on using compliant mechanisms for

developing positioners for mesoscale systems which assist part positioning by reducing the

alignment errors. Culpepper and Kim [6] developed a six-axis reconfigurable manipulator

consisting of compliant elements and actuators for part positioning. The characteristics of

compliant mechanisms can then be extended to not just positioners, but also to fixturing systems.

Since the feasibility of mechanical adaptability at the mesoscale [4], [5] and macro-scale [7] for

fixturing has been already been successfully demonstrated (without implementation), research

needs to be done to take advantage of these properties. The use of compliant mechanisms to

achieve variable stiffness characteristics can therefore improve the performance of mMTs.










The performance of a mMT is dependent on many process factors such as the spindle

rotation speed, the runout in the tool, feed rate, cutting force and the depth of cut. Tool runout is

one of the important parameters which directly affects the accuracy of the mesoscale

manufacturing process [3]. Runout, often measured as Total Indicated Reading (TIR, is defined

as the difference between the maximum and the minimum distance of the tool point from the axis

of rotation measured over one revolution as illustrated in Figure 1-3 [12]. Radial runout is the

result of a lateral (parallel) offset between the rotational axis of the tool and the central axis of

the collet/spindle system. Runout can also occur due to variation in the length of the teeth of the

cutter resulting in varying depth of cut. A longer tooth will have a deeper cut as compared to the

teeth of the correct size, thus affecting the accuracy of the machining process. The runout from

the spindle of the system (or from varying teeth dimensions) causes an extraneous force on the

workpiece and results in increased excitations [3]. Even though there exists a relationship

between runout and the resulting contact force, researchers have yet to define the specifics of this

relationship in terms of magnitude, frequency and amplitude as a function of the runout and tool-

workpiece material properties. Non-contact precision capacitance sensors are typically used to

dynamically measure the runout of a tool. These sensors use capacitance technology to measure

the runout and provide accurate readings for frequencies up to 120,000 rpm.

There are many factors which affect the tool runout such as frequency of spindle rotation,

the duration of the operation and the tool-workpiece interactions. The runout value increases for

higher frequencies and these values change depending on the precision of the setup. The typical

values of runouts for high precision systems are in the range of 0.001 mm to 0.006 mm and in

the range of 0.03 mm to 0.05 mm for lower precision systems [12]. Although considerable

research is being carried out to achieve reduced runout, it still remains a challenge. Hence,









alternate approaches are required to achieve control on the adverse affects of tool runout. Using

fixturing methods with variable stiffness characteristics is one such approach.

1.5 Thesis Outline

This thesis is organized in the following manner. After giving a brief introduction to the

growing trend towards micro/mesoscale manufacturing and the challenges faced in accuracy due

to the adverse affects resulting from tool-workpiece interactions in Chapter 1, Chapter 2

summarizes the problem statement followed by a discussion of the assumptions made in

developing the fixel designs. This is followed by Chapter 3 which provides a detailed description

of the principles of each of the four fixel designs and the relationships between fixel stiffness and

the specific design parameters. Chapter 4 describes the metric obtained to compare the compliant

beam fixel designs followed by the plots depicting the stiffness characteristics of the fixel

designs. Chapter 5 deals with the implementation and analysis of one of the fixel designs in a

mesoscale manufacturing process under typical operation conditions. Later Chapter 6 identifies

the maj or conclusions and highlights areas where thi s work needs to be further developed.































Figure 1-1. Conceptual Integration of Fixturing and Mesoscale Machine Tool (horizontal plane
of application) (mnMT stage courtesy of J. Ni at S.M. Wu Manufacturing Research
Center, University of Michigan, Ann Arbor, MI)































Figure 1-2. Conceptual Integration of Fixturing and Mesoscale Machine Tool (vertical plane of
application). (mMT stage courtesy of Thomas N. Lindem at Atometric, Inc.,
Rockford, IL)



Runout


Actual tool motion


Expected tool motion

Figure 1 -3. Illustration of runout error









CHAPTER 2
THEORETICAL FIXEL MODEL ASSUMPTIONS

As stated in Chapter 1, the overall obj ective of this thesis is to characterize the

functionality and capabilities of an active fixturing methodology for mesoscale manufacturing.

This requires a clear understanding of the compliant nature of each fixel and the system of fixels

as well as the dynamic impact of active fixturing on the tool-workpiece interface dynamics.

Chapter 1 first provides a description of the active system and how it may be implemented on

two different typical micro/mesoscale machine tools (vertical and horizontal planes of

application). This is followed by a narrowing of the problem statement to fixel design

development, modeling and analysis in terms of fixel stiffness and dynamics. In this thesis, the

workpiece-tool interface scenario used as a basis for analysis and demonstration is tool runout

and its adverse effects on feature dimensional accuracy.

2.1 Problem Statement

The focus of the current study is to develop fixture designs which have variable stiffness

characteristics and can be implemented on typical mesoscale machine tool systems. After

developing the designs, modeling and analysis is to be performed to validate the implementation

of the design in a typical mMT setup. Four different fixel designs are considered in this study, a

compliant four-bar mechanism type fixel and three beam type fixels with the beam type fixels

being fundamentally similar.

2.2 Assumptions

While the four bar mechanism based fixel design consists of compliant components and

will typically be fabricated as a monolithic compliant structure, the modeling of this design is

done using its Pseudo-Rigid Body Model (PRBM). Therefore, the assumption for the compliant

four bar mechanism is a four bar mechanism consisting of rigid links with spring loaded j points.










Comparatively, there are different methods for obtaining the compliant structure. One method is

to locate all the compliance within j oint centric locations of the device for which the lengths of

the rigid segments are large relative to the lengths of the flexural members. The flexure

members provide localized compliance analogous to a spring loaded kinematic j oint. Another

method is to have a structure with a kinematic topology that yields motion under loading

analogous to a four bar mechani sm. Either of these methods can be modeled using the PRBM

method yielding kinematically and dynamically equivalent designs. In this thesis, the first

method is assumed (details of which can be found in a later section 3.1.2). It should however be

noted that the second method is anticipated the preferred method for yielding the designs to be

fabricated and implemented in the commercial application. This will be the subject of future

research and development.

For the beam type fixels, three different implementation approaches are explored. The

fundamentals of the beam type designs arrive from solid mechanics principles for beam

deflections. As known from these fundamentals, the beam deflections can be expressed as a

function of a given load and support locations [10]. From these principles, the fixel's stiffness

(proportional to Aforce over Adisplacement at the point of fixel-workpiece contact) can be

quantified for each fixel configuration. The Euler-Bernoulli beam theory assumption is made in

deriving the theoretical models, i.e., the length of the fixel beam element (denoted by L) is

constrained to be at least 10 times larger than its cross-sectional thickness (denoted as width h

herein) as shown in Equation 2-1.

L /h >10 (2-1)

This ratio of L to h is an important parameter for the designs which is explained in the later

sections.









CHAPTER 3
PRINTCIPLES OF THE DESIGNS

The principles involved in each of the four fixture designs proposed in this study and the

mechanical parameters involved in each of the designs are described in this chapter. The

relationship between these parameters and the fixel stiffness is then established followed by

developing the variable stiffness characteristics of the designs.

3.1 Design 1

The first fixture design under consideration is one in which each fixel is a monolithic

compliant four-bar mechanism consisting of links and compliant flexure j points as illustrated in

Figure 3 -1. The length of the links and the length of the flexural j points are the characteristics of

this design. To simplify the modeling, an equivalent pseudo-rigid-body model (PRBM) of this

fixel is developed in this study. The next section gives a brief introduction to compliant

mechanisms and the development of equivalent PRBMs.

3.1.1 Introduction to Compliant Mechanisms

Compliant mechanisms are mechanisms which transfer motion, an input force or energy

from one point to another, using flexible members. Unlike rigid body mechanisms, where the

mobility of the mechanism is achieved through the degree of freedom in j points (kinematic pairs),

compliant mechanisms achieve mobility through the elastic deformation of the flexible members.

An example of a compliant crimping mechanism similar to a vice grip is shown in Figure 3-2.

This mechanism consists of three flexible members, unlike in a vice grip which consists of

revolute j points instead of these flexural pivots. The input force applied on the hand grips results

in some energy being stored in the mechanism in the form of strain energy in the flexural pivots.

Some examples of other common compliant devices are binder clips, backpack latches, paper

clips, etc.









The two main advantages of compliant mechanisms are cost reduction and increased

performance [l l]. Since compliant mechanisms are mostly monolithic and do not require

kinematic pair type j points or pins, this results in reduced part count as compared to a rigid body

mechanism. This directly results in reduction in assembly time, improved manufacturing process

reliability and the ability to be easily miniaturized. Since there is no relative sliding motion

between the surfaces of the j points as seen in rigid body mechanisms, friction is generally

assumed negligible in compliant devices. This results in reduced wear and thus reduced

maintenance in terms of lubrication. Such mechanisms also do not have the inherent problems of

backlash and/or joint tolerances which is common in j pointed mechanisms. Thus, these

mechanisms also exhibit increased precision and increased reliability.

3.1.2 Pseudo Rigid Body Model of a Compliant Mechanism

Although there are numerous advantages of compliant mechanisms, there are some

challenges faced with respect to analyzing and designing them. The most well known technique

to design and analyze compliant systems that undergo large deflections is the use of Pseudo

Rigid Body Model (PRBM). In this concept, the deflection of flexible members is modeled using

equivalent rigid body components. The flexural pivots are modeled as kinematic pair type

joints/pins with the stiffness of the flexural members represented by torsional springs within the

joints. For each PRBM, the most important consideration is the placement of the revolute/pin

joints and the value of the spring constant to be assigned [1 1].

A monolithic compliant four bar mechanism consists of four rigid links connected by

flexural pivots. A PRBM for such a mechanism is modeled by placing revolute j points at the

center of the flexible segments and torsional springs at each of these j points. The placement of the

joints is based on the assumption that the lengths of the flexural members are small relative to

the lengths of the rigid segments.









3.1.3 PRBM of the Fixel

Figure 3-3 presents the pseudo-rigid-body model of the four-bar mechanism type fixel

developed as explained in the previous section. Each of these four bar fixels will provide

directional variable stiffness characteristics to the workpiece fixturing setup within its plane of

motion. In this model, the j oint flexural pivots are modeled as revolute j points (2, 3, 4 and 6) as

shown in the figure. The stiffness of the flexural pivots is expressed by adding torsional springs

of spring constants k,, k3, k, and ko, respectively. The base link has a length ofL1 while the rigid

links have equivalent constant link lengths of Lz, L3, L4, and L5, and the contact point of the f ixel

with the workpiece occurs at the end of link Lg which acts as the interface element. The angles

between the links L1 and Lz is labeled as 02(called the base angle), the angle between L1 and L4 aS

04, the angle between Lz and Lg as 03, the angle between Lg and L4 aS Os and the angle made by

link L5 with the vertical is denoted as Os. Due to the geometry of the four bar model, the values

of the angles 83, 84, 85 and Os are functions of 82 and the link lengths and hence, can be obtained

for any configuration of the model. These relationships are expressed using the Equations 3 -1 .

dr36 = Sqrt(L + L~ 2 L, cos B,)
30, = sin-1(L~ sin Bz / d36
03b = COS-1(( +d362 --L )\/23d36)

Oz = Oza + 83b (3-1)
04 = Sin-1(d36 Sin 3b; / L4,
86 = 2xr-8, -8, 84
8, = --~r + 8, + 83

As mentioned previously, the tip of the interface element (link L5) is in contact with the

workpiece. The x and y coordinates (xc, yc) of this point contact with respect to the X, Y frame

affixed to the base link are given by Equations 3-2. Hence, the change in ye in one configuration









to another configuration represents the displacement of the point of contact which will be

required to calculate the (vertical) stiffness constant at the point of contact.

xc =-1 /2L, +L, cos ,+ (1 /2)L co s8, -L,sin 9,

ye = L, sin 8, + (1/ 2)L sin 9, + L, cos8, (3 -2)

For this design, the stiffness characteristics of each fixel as seen at the point of fixel-

workpiece contact is mechanically adjusted by extending the base length (L y) and/or an input

angle (82). Hence, L1 and 82 represent the mechanical parameters for this design with the

stiffness constant being a function of these parameters.

3.1.4 Derivation of Stiffness of the Fixel at Contact Point (Along y-direction)

To derive the stiffness values for the PRBM, the link lengths were chosen such that Lz=L4

and Lg is shorter than the other links. Also, all flexures (torsional springs) are assumed to have

equal stiffness values. During the equilibrium configuration of the fixel, there is no net force

acting on the contact point and hence, the j oint flexure' s torsional loads are calculated using

angles measured from this unloaded equilibrium. At this equilibrium configuration, all the link

lengths and the respective angles are at their initial values and the link Lg is horizontal. At this

configuration, the length of base link is denoted as L1o and 8 is denoted as 82o as shown in

Figure 3 -4a (the remaining angles are also denoted with similar sub scripts). When one or both of

the mechanical parameters of the design (L1 and 82) are varied from their equilibrium values,

assuming contact is not lost, it results in a contact force (Fe) acting at the workpiece-fixture

interface and a corresponding change in the y-coordinate of the point of contact (yc). When the

base length and/or the base angle are again changed, it results in a corresponding change in Fe

and ye (denoted by hM and Av ) as compared to the previous configuration. The ratio of this

change in contact force to the change in the position of the point of contact is defined as the









stiffness of the fixel at the point of contact. Hence, to derive the stiffness value, the

corresponding values of AFc and Ave are calculated. The procedure to calculate this stiffness is

described next.

The base link length and the base angle can each be varied within particular kinematic

ranges with the initial value of the base length operation range being L1o. The range of values for

base length can be chosen according to the footprint of the setup. To calculate the range of values

of the base angle, a particular value of base length is chosen from the range and the unloaded

equilibrium configuration for this value (as shown in Figure 3 -4a) is first determined. Each of

these unloaded equilibrium configurations, result in a four-bar mechanism and the two toggle

positions (motion limits) for this four bar model are then determined. These limits provide the

corresponding kinematic range of base angles that comply with the design at that particular base

length. This process is repeated for all possible values of the base length and the individual range

of values of base angles is obtained for all the base lengths. From these different sets of ranges of

base angles, the values which are common to all the sets represent the final operation range for

the input base angle 82.

To calculate stiffness, the base length is first changed to a particular value in the operation

range (say L1) using a pri smatic actuator shown in Figure 3 -3. Due to the geometry of the four

bar model, this change in the base length results in the base angle changing from 87o to BR as

shown in Figure 3 -4b (with the remaining angles similarly changing to 83o', 84o'anYd 8o'). Then,

the base angle is increased to a certain value (say 82,) within the operation range using the

revolute j oint actuator shown in Figure 3-3. This increment of aO will result in a change in the

angles of the model (i.e. change in 8,1, 84, and Os,) along with a displacement of the link Lg and a

contact force (say Fer) acting on the link Lg. The contact force occurs on the link Lg so that the









model remains in static equilibrium. The configuration of the fixel at this instant is shown in

Figure 3-4c.

At this configuration, a static force analysis is carried out on the model and the equation

shown in Equation 3 -3 is obtained.




c,,i 1 0 1 0 1 0 0
32xt (3-3)
F32 -0 0 M43M,4 0 0 V
F 51 5, ,3 54 5,6 5
F ~6;56,6 6
F34


where F;;; i s the force acting on link 1 by link m, sub scripts x and y indicate the

corresponding (xy) components of the forces Fa;; and the subscript i denotes the current

configuration. In addition, the values of V4, Vs, and y for the corresponding torsional springs are

obtained using the value of the spring constant and the change in angle at the particular joint as

shown in Equation 3 -4. The elements of the inverse matrix in Equation 3 -3 are functions of the

link lengths and the angles and are given by Equation 3 -5.



V4 -k i k 8 -8



V6= -k6l~ 6 i k4 4









M4,3 = L2 *Sin(82i) M4~,4 = -L2 "COS(2i)
Mj~ = -L5 *cos(est) M5,2 = -L5 *esin(8st)
Lr.*ins i L3 (3-5)
M5,3 Sin(5i) 5,4 = COS(5i)

Ml = *sin(8st) M5,6= CS(8i>
M6,5 = L4Sn04) M,6 = 4CO(4


Using the above equations, the value of the contact force at the current configuration

( Fc, ) is calculated. The value of y-coordinate at this instant ( ye~ ) is then calculated using the

expression for ye shown in Equation 3 -2 and the current values of the angles.

Keeping the base length fixed, the base angle is then incremented to another value (say

62,47 Within the kinematic range) which results in a corresponding change in the remaining angles.

The new configuration of the model at this instant is shown in Figure 3 -4d. A static force

analysis is again conducted on the model in this configuration and equations similar to Equations

3-3 to 3 -5 are obtained. Using these equations, the value of the contact force at the new

configuration ( F ~~) and the value of y-coordinate at this new configuration ( yeg ) are then

calculated. The difference in the contact forces at the new configuration and the previous

configuration is calculated and is denoted by (a4)l,, followed by calculating the displacement of

the point of contact, (@yc),,, (difference in the y-coordinates of the two configurations). The ratio

of (a4),,l to (@c,),,, gives the numerical approximation of the stiffness at the point of fixel-

workpiece contact corresponding to the current configuration (L1 and 62,4 '

This procedure is then repeated by keeping the length of the base link fixed and varying the

base angles through the kinematic range of values to obtain different stiffness values for the

different configurations (L1 and 62 S). Another set of stiffness values is then obtained by









changing the base length to another value (say L2) fTOm the known range and repeating the

procedure for all possible base angles. These stiffness values correspond to the new set of

configurations (L2 and 62 S). By repeating this procedure for all base lengths and all base angles

within the allowable range, a final set of all the possible stiffness values (along y-direction) is

obtained. Thus, using this fixel design, adjustable fixel stiffness can be achieved by varying the

two mechanical parameters the base length and the base angle of the fixel.

3.1.5 Derivation of Stiffness of the Fixel Along x-direction

The stiffness of the fixel along the x-direction for all possible values for LlandBz in the

kinematic range of values are calculated by following the procedure similar to the derivation of

the stiffness along y-direction. The value of the contact force along the x-direction ( Fc~ ) is


calculated using Equation 3 -3. The value of x-coordinate of the point of contact (x-cc, ) is

calculated using the expression for xc shown in Equation 3-2. Repeating all the steps followed in

section 3.1.4, a final set of all the possible stiffness values (x-direction) for all base lengths and

all base angles within the allowable range is obtained.

Hence, the Design 1 provides directional adjustable stiffness characteristics in X and Y,

and has two mechanical variables for adjustment.

3.2 Design 2

Fixel Design 2 is a cantilever beam fixel with unidirectional variable stiffness

characteristics. This design is a basic cantilever beam (beam fixed at one end) with contact

between the workpiece and the fixel occurring at its free end. For a cantilever beam, the

minimum value of k, or stiffness, occurs at the free end of the beam. For a given material, this

minimum k i s a function of the length L of the beam measured from the fixed end and its

moment of inertia I. For the designs shown herein, the fixels are assumed to have rectangular









cross-sectional areas with breadth b and width h. Thus, it is understood that k is a function of L

and h and this dependency is advantageous in designing fixels. For the cantilever beam type fixel

designs proposed in this section, the mechanical adaptability is achieved individually varying L

or h. Thus, for these designs, there is a single mechanical parameter (single actuation) unlike

Design 1.

3.2.1 Design 2a) Mechanical Variable L

The cantilever beam type fixel for Design 2a is shown in Figure 3 -5. The integration of the

four fixels with the workpiece setup is shown in Figure 3 -6 which consists of the

cantilever/compliant beam, the fixture base module and the interface element. To decouple

residual motions from occurring when making changes in the mechanical variables, the fixel

contact with the workpiece is acquired via the interface component. One end of the interface

component has direct contact with the workpiece and the other end is connected to a point on the

cantilever beam (near the free end). The contact force Fe between the workpiece and the fixel

acts along the interface element and occurs at this point on the cantilever beam. This contact

force will vary depending on the loads involved in the machining process. The fixture module

clamps the cantilever beam at the base, with the length of the beam L being defined as the

di stance between thi s fixed/constrained end and the point of action of the contact force (near the

free end of the beam).

Using the cantilever beam equation for stiffness, the principle involved in fixel Design 2a

is to adjust the stiffness k at the point of fixel-workpiece contact (near the free end of the

cantilever beam) by modifying the beam length L. Mathematically, the stiffness can be expressed

as shown in Equation 3 -6 where for fixel Design 2a, a = 3EI = Ebh /4 and is a constant, and the

mechanical variable is L.










3El Ebh3
k = (3-6)
L3 4L3 L

In response to the fixturing needs, the length of the cantilever beam could be controlled by

unclamping the beam from the fixture module base and then moving it by the required length;

thus changing the effective beam length L (as illustrated in Figure 3 -6). In this design approach,

the value of k given by Equation 3 -6 can be adjusted passively during fixturing set up as well as

actively during the process.

For the Design 2a' s fixel, the minimum value of k would occur when the length is Lmax and

maximum value of k corresponds to Lmzn as shown in Equations 3 -7 and 3-8.

Ebh3
kmi (3-7)
4max3 Lmax3

Ebh3
kmax (3-8)
4L~mn3 Lmn,3

For the purpose of comparisons between fixel designs, a new parameter r is introduced

which i s defined as the ratio of minimum to maximum values of the fixel's mechanical variables.

This ratio provides a dimensionless measure for quantifying the mechanical variable' s range of

motion, facilitating the comparison of the different fixel design types. For Design 2a, r is the

ratio of Lmzn to Lmax and substituting r into Equation 3 -7 results in the Equation 3-9 which

provides a relationship between the design constants (material and geometric parameters), kmzn,

and the range of motion of the mechanical variables r.

Ebh3 3 a3
kmi (3-9)
m 4L~mn3 Lmn,3

Hence, for Design 2a, Lmzn and r are the two unique parameters. The selection of Lmzn and r

determine the value of Lmax and by varying L from Lmzn to thi s calculated Lmax, a set of values of k









that are achievable for each given fixel design are obtained. For investigating the mechanical

adaptability of the fixel, the design variations of Design 2a are explored for different sets of the

design parameter r for a given L;;,,, and ks,,,.

3.2.2 Design 2b) Mechanical Variable h

The cantilever beam type fixel for Design 2b is shown in Figure 3-5. The integration of the

fixel with the workpiece setup is shown in Figure 3 -7 which is similar to Design 2a (four fixels).

In this design, the length of the cantilever is held constant and the adjustable fixel stiffness is

achieved by varying the width h of the beam as shown in Figure 3-7. This can physically be

achieved by adding multiple strips of same length and breadth to the cantilever beam thus

changing the effective width h '. The effective width can be calculated using the Equation 3-10

where n is the number of thin strips and Ah is the width of each thin strip.

h' = h +n(Ah) (3-10)

Since Ah has to be increased discretely (unlike L in Design 2a), a very small value of Ah can be

chosen such that the change in k is equal to the least count of the system (the least value of

change in k required).

The value of h (and thereby the number of thin strips) needed to obtain a required k can be

calculated using Equation 3-10. The width could be changed from a minimum value of he,,, to a

maximum value ofh;;; which would correspond to a minimum and maximum value for fixel

stiffness k given by the Equations 3-11 and 3-12 where for fixel Design 2b p would be a

constant defined as P = Eb/4L3 and the mechanical variable is h.

Ebh
km mm =p flm (3-11)
4L3 i

Ebh
k =mx= phm;x (3-12)
mx 4L3 a









Similar to Design 2a, introducing r as the ratio of hmzn to hmax and sub stituting r into

Equation 3-11, results in Equation 3-13.

Ebr3h 3
k =max = fl3hm 13 (3-13)
mm 4L3 a

Here, hmax and r are the two unique parameters of Design 2b and fixed values for these

parameters determine the value of hmzn. By varying h from hmax to this newly calculated hmzn, one

obtains the set of values for k that are achievable for each given design. Similar to the previous

design, the variations of Design 2b are explored for different sets of the design parameter r for a

given hmax and kmzn. These results will be compared with the designs of other fixels to further

delineate their effectiveness.

3.3 Design 3

The fixel for this design consists of a beam fixed at both ends as shown in Figure 3-8. The

workpiece would be held via contact with one end of an interface element (similar to Design 2a)

where the interface element connects at some point along the beam. Figure 3 -9 is an illustration

for the implementation of Design 3 using four fixels to hold the workpiece. Each contact point

would be at a distance x from a fixed end and the contact force Fc acting along the interface

element will be exerted on the beam at this point. By moving the beam along the direction of its

axis and by allowing the end of the interface element to roll on the beam, the contact force point

of application can be changed without inducing any motion of the workpiece. The new value of x

obtained by this change would correspond to the value of k required at this instance. This value

of x can be calculated using the relationship between k and x given by Equation 3-14 where L is a

fixed length and 7 is a constant defined as y = 3EI = Ebh3/4 .


k = L3(3-14)
x3 (L -x)3










Thus, in this design, the stiffness k can be controlled by varying the point of application of

the contact force (Fc), i.e., by adjusting the mechanical variable x. The distance of the contact

point can be varied from a minimum value of xni to a maximum value of xna where xna will be

half the fixed length L of the beam (due to the symmetry of the fixel system). The corresponding

minimum and maximum values of k are given by the Equations 3-15 and 3-16.

2Ebh3
km = (3-15)
mx,

Ebh3L3 YL
k =(3-16)
4xm 3 L xmm, xmm X,,L xmm )

Introducing r as the ratio of xnz to xna and substituting r into Equation 3-16 results in the

Equation 3-17 below.

2Ebh3
kma (3-17)
r' (2-r xm

Here, xna and r are the two unique parameters of Design 3. Also, similar to the previous designs,

fixed xna and r values will determine a unique value ofxnzn. Again, similar to the previous

designs, for a fixed xna and for different sets of the design parameter r, different variations of

Design 3 are obtained. Each design variation results in a set of k values that can be achieved

from varying its mechanical variable (x) from xnin to xna.










Compliant joints


Figure 3 -1 Design 1 Simple monolithic compliant four-bar mechanism consisting of four
compliant j points


Figure 3 -2. Example of a Compliant Mechanism a Crimping Mechanism consisting of three
flexible members (courtesy of Compliant Mechanisms by Larry Howell)









F-Contact Pona S orkint
between Fixel.


T


Y
Base Pointf

Revolut e ,
Joint 2- Ii
Actuator


Revolute
Toint 4


PrismaEtic Joint Actuatorc


Figure 3 -3. Pseudo Rigid Body Model for the Four bar monolithic fixture












xL
lo, A





L1 B

lFc




L1 c






L1 D
Figure 3-4. The various configurations of Design 1 during stiffness calculations a) Initial
Configuration b) Symmetric Equilibrium with Base Length L1 c) Displacement Due
to Contact Force with Base Length L1 d) Change in Displacement Due to a Change in
Contact Force with Base Length L1








Varying thickness


Fixed length
compliant beam


Varying length
compliant beam






Fixurebas :


Design 2a)


Fixture base



Design 2b)


Figure 3-5. Design 2


Principles Involved


It


Compliant beaml


F~ixtur~e base
mllodule


Figure 3-6. Implementation of Design 2a with Four Fixels


























Col~impla~nt~ ~ bea
Fixtur~e base
muodlule


Figure 3-7. Implementation of Design 2b with Four Fixels








Varying point of
application of force


Fixed at I
bo3th ends


Figure 3-8. Design 3


Principle Involved




















WTorkpiece
Fe Fe


Interface
elemllent

Fe
Beaml fixed
at both ends



Figure 3-9. Implementation of Design 3 with Four Fixels









CHAPTER 4
COMPARISON OF THE DESIGNS

After establishing the principles of the designs and the relationships between fixel stiffness

and the mechanical parameters of the designs, the fixel stiffness at the fixel-workpiece contact

are plotted for Design 1. This is followed by a comparison of the stiffness characteristics of the

similar compliant beam designs 2a, 2b and 3.

4.1 Design 1

For the Design 1 proposed in section 3.1, results were generated for the fixel design with

constant parameters Lz=L4=45mm, L3=30Omm and all torsional springs having equal spring

constants. The link lengths were chosen to exhibit the same relative footprint as Designs 2 and 3.

The fixel stiffness as seen at the point of fixel-workpiece contact for the four-bar type fixel is

plotted as a function of the two mechanical variables (L,,02) by following the procedure

mentioned in the section 3 -1 The kinematic range of values for L1 is chosen from 45mm to 55

mm and the range of values for 82 for these link lengths was obtained to be from 40 degrees to 60

degrees. This plot for fixel stiffness with respect to L1 and 82 is shown in Figure 4-1.

In the Figure 4-1, a dynamic symmetry is observed for the mechanical variable 82 for each

value ofL1. This symmetry can be attributed to the design constraints ofLz=L4, joint flexures of

equal stiffness and to the method of calculating the range of 82 from one toggle position to

another. Because of the dynamic symmetry, it is observed that the trend in the fixel's stiffness is

independent of the length of the base link. This is apparent in the Figure 4-1 by observing the

different contours in 82 for different L1. However, the magnitudes of the base link length impact

the actual instantaneous stiffness values and thus the dynamic range of stiffness achievable for a

given range of the base angles. For example, an L y=45mm fixel configuration exhibits higher

range of stiffness values achievable as compared to L1=55mm fixel configurations.










It is also observed that for the L1=45mm fixel configuration, a local minimum about its

unloaded equilibrium configuration (82=50 deg) exists and therefore deviations from this

configuration will take less effort. While for L =55mm fixel configuration at its unloaded

equilibrium configuration, the fixel exhibits a peak stiffness value and will be resistant to

changes from this configuration. It is also observed that the range of stiffness values achievable

is lesser for smaller base length values. Due to the presence of these variations in the trends, it is

observed that Design 1 will provide for greater control diversity in tuning the fixture dynamics

via actuating the base length (L y) and input angle (82) either individually or together.

4.2 Comparison of Designs 2a, 2b and 3

Arbitrarily considering values for each fixel's fixed and unique design parameters, would

not yield a discernable comparison of the different types of fixel designs. Thus, to compare

Designs 2a, 2b and 3, a common range of values of k achievable for all three designs is

determined. It is known from the previous section that each of the three beam type designs have

a particular knzn value for given r and fixed design parameters. Hence, for more equitable

comparison between the Designs 2a, 2b and 3, this minimum value of k for each of the designs

with same specified value ofr were made equal. Equations 4-1 to 4-3, shown below, are

obtained by equating ks,,, and r values of each of the designs.

Ebh r3 Ebrh hma 2Ebh3
kmi (4-1)
4 mn2eign 2a dein2b mx dsg


k h~r r~h 2h3
Ebn 4Lin 4 xma



r =Lmm hmn xmi (4-3)
Lmax ma Xmax










By restricting that all designs have the same minimal fixel stiffness (kmzn), the value of kmax

at r equals 1 is also found to be the same for all three designs. Note, when r equals 1, the range

of motion of the mechanical variable is restricted to be zero, thus resulting in the loss of the

variable stiffness feature of the fixel. With this common window of constraints on the stiffness

values formed by the kmzn restriction, the values of kmax, Ak (equal to kmax kmzn) and the average

slope of k (Ak by change in the varying parameter of the design) for the intermediate values of r,

the different beam type fixel designs can be compared more directly.

For Equation 4-1 to yield a viable comparison, it was assumed that the breadths (b) of the

beams of each design are equal and that the material of the beam is also the same. Furthermore,

the constant beam width h of Design 2a was made to be equal to hmzn of Design 2b, and the

constant beam width h of Design 3 was made equal to r times hmzn of Design 2b. Similarly, the

constant length L of Design 2b was made equal to Lmzn of Design 2a and xmax of Design 3 was

made to be twice Lmzn. This leads to the different beam type fixel designs having the same lower

bound on the slenderness ratio.

The slenderness ratio (s) of a beam is defined as the ratio of the minimum length of the

beam to its maximum width. For the different beam type fixel designs, the following

relationships shown in Equations 4-4 and 4-5 yield the lower bound on s.


s = m for Designs 2a and 2b (4-4)


2x
s = mx for Design 3 (4-5)


where, from the assumptions of Euler-Bernoulli beam theory, s should be greater than 10.

To determine Lmzn, the size of the workpiece has to be considered. Assuming that the

workpiece has a size of 10mmx10mmx10mm, a safe estimate for Lmzn is given by Equation 4-6.










Lm = 10 inn (4-6)

From specified values of Lnz, r and s, the parameters of comparative fixel designs and the

corresponding k values can be obtained by the following procedure:

4.2.1 Design 2a

Equation 4-6 gives the value of Lnz

Using the specified value of s, the value of hna can be obtained from Equation 4-4. Then

using Equation 4-3 and the specified value for r, hnzn can be calculated to determine the constant

beam width h as shown by Equation 4-7.

h = h,, (4-7)

Using Equation 4-3 with the same value of r, the value of Lna can be obtained from the

known value of Lnz.

For each fixel design variation defined by specified values of s and r, a range of values for

L varying from Lnzn to Lna is obtained which correspond to a range of values of k.

4.2.2 Design 2b

Equation 4-6 gives the value of Lnz which would be equal to the constant beam length L of

this design as shown by Equation 4-8.

L = L,, (4-8)

Values of hna and hnzn for specified values of s and r are the same as that obtained above

in Design 2a.

A range of values for h varying from hna to hnzn is thus obtained which correspond to a

range of values of k.










4.2.3 Design 3

As previously explained, x;;;a is twice the value of the L;;,,, obtained from Equation 4-6

which can be expressed as Equation 4-9.

1718X 171 2L,, (4-9)

Also, the constant width of the beam would be r times he,,,. Using the known value of he,,,

from Design 2a calculations, the fixed width can be calculated using Equation 4-10 shown

below.

h = rhniz (4-10)

The value of x;,,, for specified values of r can then be determined by substituting the value

of x,,; into Equation 4 -3 .

For specified values of s and r, a range of values of x varying from x,;; to x,,,,, is obtained

which correspond to a range of values of k.

4.2.4 Plots for Design 2a, 2b and 3

For each of the above beam type fixel designs, ks,,,, ks;,, Ak and average slope of k were

plotted as a function of r and s (r varies from 0.1 to 1 and s varies from 10 to 30). Figures 4-2

through 4-9 are the normalized plots of the stiffness quantities (ks,,,, k;;;a, Ak and average slope of

k) with respect to (Eb). These results provide the trends for each design independent of the

fixel's material properties. Figure 4-2 demonstrates that all the three designs have the same

instantaneous ks,,, values, shown as a function of r and s. Figures 4-3 and 4-4 demonstrate the

variation of k;;;a for various combinations of [r, s]. It is found that all the designs have the same

maximum possible k;;;a value (which occurs at r= 1 and s=10). However, the variation of k;;;a as

a function of r and s is different for Design 3. Figures 4-5 and 4-6 represent the variation of Ak

(difference between k;;; and ks,,,) for the three beam type designs. Figures 4-7 through 4-9 are










the plots for the average slopes of k, where the average slopes are calculated as the ratio of Ak

over the range of the change in the mechanical variable corresponding to each of the designs.

From these plots, it is observed that the range of k's achievable reaches a maximum at

around r 0.8 for all slenderness ratios. Also, a smaller s (shorter or thicker beams) yields a

greater range in k. For Design 2b, a smaller value for s results in a larger hmax value for given Lmzn

which has the effect of increasing the range of the mechanical variable h and thus increasing the

range of k. This increase in hmax effects the h of Design 2a which in turn has a similar

amplification affect on thi s design' s resulting range of k. The Ak are also found to be higher for

Designs 2a and 2b than that of Design 3 (by a factor of 7). For an overall measure of k variations

between kmzn and kmax, Design 2b trends were found to be a factor of 10 higher in average change

in k, as compared to Designs 2a and 3. This indicates that Design 2b has greater mechanical

adaptability for a given range in variable (defined by r).










dFy /dAY (N/mm) versus [82(degrees), L (mm)]


L1 (mm)


82(degrees)


Figure 4-1. Fixel Stiffness (L2=L4=45mm, L3=30mm, all k' s are equal)












kmin for designs 2a), 2b) and 3 vnersus [r,s]


20
15

s from 10 to 30


Figure 4-2. (k,nzn /Eb) for Designs 2a, 2b and 3 versus [r,s]











kmax for designs 2a) and 2b) versus [r,s]


0 302
15
10
s from 10 to 30


Figure 4-3. (kmax /Eb) for Designs 2a and 2b versus [r,s]











kmax for design 3 versus [r,s]


0302
15
10
s from 10 to 30


Figure 4-4. (k,na /Eb) for Design 3 versus [r,s]












delta k for designs 2a) and 2b) versus [r,s]


25 ~-0.6 0.4
0.8
30
r from 0 to 1
s from 10 to 30


Figure 4-5. (Ak /Eb) x 105 for Designs 2a and 2b versus [r,s]













delta k for design 3 versus [r,s]


v.o
30 r from 0 to 1


s from 10 to 30


Figure 4-6. (Ak /Eb) x 105 for Design 3 versus [r,s]












Average slope of k for designs 2a) versus [r,s]


4



'11


1 0.8

00 0.6
25 0.4
200.
15
10 from 0to 1
s from 10 to 30



Figure 4-7. (Average Slope of k /E~b) x 10-4 for Design 2a versus [r,s]












Average slope of k for designs 2b) versus [r,s]






x 103













0.8

0.6

030 .4
25
20 0.2
15 10r from 0 to 1
s from 10 to 30



Figure 4-8. (Average Slope of k /E~b) x 10-3 for Design 2b versus [r,s]












A\Rrage slope of k for design 3 \ersus [r,s]


x 10


0.8


0.6
r from 0 to 1


s from 10 to 30


Figure 4-9. (Average Slope of k /Eb) x 105 for Design 3 versus [r,s]









CHAPTER 5
ANALYSIS OF THE FIXEL DESIGNS

In this chapter, a fixel design is incorporated into a machine tool setup and the performance

of the design under operation conditions subject to runout induced errors is analyzed. The fixel

design and the number of compliant fixels (of that design) used in a setup will depend on the

type of manufacturing process. For example, in case of a meso scale drilling process at least four

fixels of Design 1 are required. The (X, Y) directional stiffness characteristics of fixels of Design

1 integrated in a coordinated manner is best suited for achieving the preferred fixturing

capabilities and accurate drilling. In the case of a milling process, two fixels of any of the four

designs are theoretically required to achieve fixturing capabilities. Because each of the designs

have the same principle of utilizing adjustable stiffness characteristics, the following analysis is

done without loss of generality for only one of the designs for a particular meso manufacturing

process. The results of this analysis can be extended to other designs and processes. For the

purpose of this study, Design 2a is selected for analysis and is implemented in an end milling

machine tool system which has runout in the direction perpendicular to feed. Another example of

such a system could be a burnishing setup where a mesoscale workpiece with micro features is

being burnished for improved finishing. When the tool is burnishing the surface of the workpiece

which is close to the feature, the runout in the tool (perpendicular to the feed) inadvertently

affects the dimensions of feature during the burnishing process.

One more mesoscale manufacturing setup which can have runout in the perpendicular

direction to the feed is a peripheral milling setup with runout error occurring due to variation in

teeth lengths. Considering that in such a setup, one of the teeth of the cutter is longer than the

other teeth, the excess length of the longer teeth corresponds to the runout error. Such runout

errors in a peripheral milling setup results in varying depth of cut and this error can be









compensated by following the same procedure as that for the end milling setup which is

explained in Section 5.1 below.

5.1 Static Analysis of Design 2a

The fixture workpiece setup for an end milling operation with two fixels of Design 2a) and

two rigid fixels is shown in Figure 5-1. The fixture-workpiece is fixed to the platform of the

milling machine such that the rigid fixels are aligned parallel to the direction of motion of the

platform. These rigid links counter the cutting force during the milling operation. Assuming for

the current setup, that the tool on the milling machine has radial runout in the direction

perpendicular to the cutting direction, the compliant fixels are aligned in this direction. Due to

this runout, the thickness of the slot being machined into the workpiece is larger than the

required value. The perpendicular runout of the tool in such a system results in a force

(extraneous to cutting force) acting on the workpiece which then acts along the interface element

of the compliant fixel. Since the interface element is connected to the variable length compliant

beam at its free end, it leads to the bending of the beam which then results in a corresponding

displacement of the workpiece. The rigid fixels slide freely in this direction without their contact

with the workpiece slipping so that it does not introduce any resistance to the compliant fixture' s

corrective displacement of the workpiece. The obj ective of the current approach is to improve

the accuracy of dimensioning in spite of errors resulting from runout by constantly controlling

the beam deflection such that it is always equal to the runout value. This eventually results in no

relative motion between the tool and the workpiece from its theoretical nominal cutting

traj ectory, thus mitigating the adverse affects on the dimensional accuracy of the finished

workpiece. The value of beam deflection x will be a function of the contact force Fc (created

because of the tool runout) and the stiffness of the beam k which in turn depends on the length of

the beam L as shown in Equation 5-1.










Ebh3
F = x (5-1)
4L3

If the runout value (and the corresponding non-proportional contact force) changes during

the process, the length of the beam can be adjusted so that the deflection of the beam is now

equivalent to this new runout value thus maintaining the desired thickness of the slot. The runout

values can be dynamically measured using Non-contact Precision Capacitance Sensors.

To prove this concept of achieving beam deflection equivalent to the runout, a ProE model

ofa single fixel and the workpiece along with appropriate constraints is modeled as shown in

Figure 5-2. The mesoscale size of the workpiece is chosen as 5 mm x 5 mm x 5 mm. A set of

values of the required runouts are initially chosen (similar to values found in literature [12]). To

calculate the range of values of the contact forces, a value of 8 mm is chosen for L and the

expected runout is chosen (from known range of values) as 0.06 mm. Using Equation 5-1, the

value of contact force is calculated to be in the order of 25 mN. Now, the value of the contact

force is varied in the range of 20 mN to 30 mN and the values of L corresponding to a range of

runouts are calculated using Equation 5-1. The contact force is not necessarily proportional to the

runout and keeping this in mind, different combinations of contact force and runout values are

chosen to calculate the L values. To validate these estimated lengths using the model, the length

of the compliant beam in the model is set to one of the calculated values of L and the

corresponding contact force is applied to the workpiece. A force analysis is done on this model

and the deflection of the beam is obtained as shown in Figure 5-3. The value of the beam

deflection is then compared to the value obtained from the formula. Since, the formula is

obtained after making certain assumptions, these values vary by a small amount. Similar values

of L were obtained for different sets of runout and contact forces and are tabulated in Table 5 -1

which shows the possibility of achieving different runout compensation by varying the length of









the compliant beam. This is again shown in the form of a graph in Figure 5-4. These values of

length and corresponding stiffness values may or may not be the same for the analysis under the

influence of a harmonic force.

5.2 Analysis of Design 2a Under Harmonic Force

The next step is to analyze the design under the operation conditions of a mesoscale

milling setup. Assuming that there exists runout in the tool of such a system, it results in a

harmonic force F (as shown in Equation 5-2) acting on the workpiece which in turn acts on the

compliant beam. The frequency of this harmonic force will be equal to the frequency of spindle

rotation. This setup is modeled in MSC Adams by approximating the cantilever beams as springs

with attached fixel masses which in turn are in contact with the workpiece as shown in Figure 5-

5. The rigid links which are aligned parallel to the cutting direction are modeled as constraints in

this model. The springs are pre-stressed such that they are always compressed thus constantly

applying a force on the workpiece. This system is equivalent to an undamped system under

harmonic force excitation and hence, the displacement of the workpiece will now be a function

of its mass and the frequency of the harmonic force in addition to the stiffness of the fixels (e.g.

cantilever beam) and the maximum amplitude of the contact force. The maximum amplitude X of

the particular solution of such an undamped system under harmonic force is obtained from

Equation 5-2 shown below, where Fo is the maximum amplitude of the harmonic force, k is the

stiffness of the springs (in parallel) in the system (cantilever fixels), m is the mass of the

workpiece and a is the frequency of the harmonic force [13].

F = Fo sin(a t)

= F, (5-2)
X= o ~









For an undamped system under a harmonic force, the phase of the particular solution is

same as the phase of the harmonic force only when the ratio of the frequency of the harmonic

force to the natural frequency co, of the system lies between 0 and 1 as shown in Equation 5-3

[13].

O < a / <1 (5-3)

But the natural frequency of the current system is a function of the spring constant of the

cantilever beam and the mass of the workpiece (ignoring fixel mass). Hence, to cancel the errors

induced by the runout, the fixels must move the workpiece in phase with the runout. Therefore,

the stiffness of the beam has to be chosen such that the natural frequency of such a system is

greater than the frequency of spindle rotation.

Now, to study the behavior of the workpiece-fixels system under a harmonic force, the

Adams model is to be analyzed at different spindle speeds ranging from 60000 rpm to 90000

rpm, which is the typical range of spindle speeds for a mesoscale system. The value of the runout

in the tool is assumed to be in 0.0012 mm and a nominal value of25 mN is chosen for the

maximum amplitude of the harmonic contact force. It should be noted that the value of runout is

chosen as 0.0012 mm since it is the least runout value usually observed in high precision systems

(found in the literature) and to verify the model for this low value. The value of the harmonic

force is chosen to be same as in the static analysis so that there is consistency with the previous

analysis. Using Equation 5-2, the value of the stiffness of the compliant beam required to achieve

the chosen runout value at different spindle speeds is calculated and tabulated as shown in Table

5-2. It should be noted that the resulting design values ofk per fixel depend on the number of

fixels and their arrangements, yielding an equivalent k value for Equation 5-2.









To validate the performance of the design, the frequency of the contact force in the Adams

model is initially set to 60000 rpm and the value of the spring constant is set to the value from

Table 5-2 corresponding to the spindle speed and the model is run for 0.5 seconds. For this

analysis, a plot for the displacement of the center of mass of the workpiece with respect to time

is obtained as shown in Figure 5-6. It is observed from this figure that the system has an initial

transient state for a very small duration (order of 0.01 seconds) following which the maximum

amplitude of the displacement remains constant. This magnitude of the maximum amplitude is

measured from the plot and is found to be 0.0012 mm which is equal to the expected runout

value.

To better compare the expected runout value and the displacement of the workpiece, these

two quantities were plotted together as shown in Figure 5-7. It can be clearly observed from this

figure that the di splacement of the workpiece i s not equal to the runout value for a small duration

(close to 0.01 seconds for this example set of conditions) at the beginning of the process. But

after this initial transient period, the displacement of the workpiece is equal to the expected

runout value along with always being in phase with one another. Thus, the displacement of the

center of mass of the workpiece is made equal to the expected runout by varying the length of the

compliant beam to achieve the required stiffness values.

To analyze the workpiece-fixel system at various frequencies, the frequencies of harmonic

force were varied from 60000 rpm to 90000 rpm and the Adams model is run for these

frequencies. During each of the runs, the spring constant is changed to a corresponding value as

obtained from Table 5-2. From the Adams models, it is observed that the systems reach steady

state quickly and the values of the maximum amplitude of the displacement are equal to the










expected runout values in the steady state. The displacement and runout values are also in phase

with another for each of the systems.

To study the error between the di splacement of the center of mass of the workpiece and the

expected runout value (0.0012 mm in each case) in the transient state, these parameters are

plotted for each of the frequencies as shown in Figure 5-8. As can be seen in this figure, the error

between these two parameters varies with the frequency of the harmonic force. Also, the time the

system takes to reach a constant value is also dependant on the frequency of the force. For higher

values of frequencies, the system reaches steady state more quickly as compared to lower

frequencies. Also, the maximum value of this error for all the frequencies is measured to be

approximately 0.001 mm. Hence, higher frequencies of spindle rotation are optimal for mMT

systems. While not shown in Figure 5-8 due to time scale shown for only 0.02 seconds, all cases

converge to a low value of error.

To estimate the physical size of the cantilever beams required to obtain the calculated

values of spring constants, the stiffness values from Table 5-2 are used to calculate the

corresponding lengths of the beam considering the breadth to be 1.25 mm and the width of the

beam to be 0.75 mm. These values have been shown in Table 5-3 and it can be observed from

this table that the required lengths are comparable to the footprint of the mMT setup. It can also

be observed from this table that since the minimum and maximum stiffness values tabulated in

Table 5-2 are nothing but the kmzn and kmax values respectively of this configuration of design 2a,

the values for length obtained in Table 5-3 are the corresponding Lmax and Lmzn values. The ratio r

of these Lmzn to Lmax is calculated to be of the order of 0.8 which from the ob servations of Section

4.2.4 is the optimum value of r for achieving maximum range of stiffness. It is also observed

from the table that the slenderness ratio s (ratio of Lmzn to h) for thi s configuration i s greater than










10, thus satisfying the assumption of the design. It should be noted that different combinations of

breadth and width of the beam and length of the beam can be used where the critical aspect is

acquiring the desired fixel settings.

These stiffness values can also be obtained for designs 2b and 3 for the appropriate values

of the parameters of the designs. To estimate the values of these parameters for design 2b, the

same stiffness values as before from Table 5-2 are used to calculate the corresponding widths of

the beam considering the fixed length of the beam to be 10 mm and the breadth of the beam to be

1.25 mm. The thickness values have been shown in Table 5-4 and are comparable to the mMT

setup. It is observed that, similar to design 2a, the minimum and maximum values of width

obtained in Table 5-4 represent hmzn and hmax of the current configuration design 2b. The ratio of

these minimum and maximum widths is calculated to be in the order of 0.8 which again is the

optimum value of r (from Section 4.2.4). Thus the parameters of the current configuration are the

optimum values for design 2b. The slenderness ratio s (ratio of L to hmax) in this case is also

greater than 10.

Similarly for design 3, the values of the di stance of the point of application of force from

the fixed ends (x) required to achieve the stiffness values from Table 5-2 are calculated and

shown in Table 5 -5. To calculate these x values, the fixed length of the compliant beam is

considered to be 5 mm, the breadth to be 0.5 mm and the width to be 0.25 mm. It can be

observed from Table 5-5 that the values of x obtained are achievable for the particular chosen

configuration. Similar to the previous designs, the ratio r (ratio ofxmin to xmax) is same as the

optimum value of 0.8 and the slenderness ratio s (ratio of L to h) is also greater than 10. Hence,

by performing dynamic analysis on the fixel-workpiece system it is established that runout










compensation is achievable using variable stiffness fixels (of designs 2a, 2b or 3) for different

frequencies of the spindle rotation.

5.3 Compensation for Transient Dynamics

As observed in the previous section, under the influence of harmonic force, the maximum

displacement of the workpiece varies initially before reaching a constant value. This transient

state exists for a very short duration but its effect on accuracy depends on the feed rate of the

process. For small feed rates, the length of the workpiece machined in this small duration will

not create excess inaccuracies. But in the case of high feed rates, these inaccuracies cannot be

negligible. If the effects of the transient dynamics are not negligible, to counter this variation it is

proposed that sacrificial material be used for the initial duration of the process. The sacrificial

material is to be obtained by using a longer (in case of milling) or thicker (in case of drilling)

workpiece and discarding the excess length/thickness after the completion of the operation.

In the determination of the appropriate fixel stiffness, the workpiece mass value will need

to include the mass of the sacrificial material in the calculation of the length of the cantilever

beam corresponding to a runout value.










Table 5 -1. Overview of the runout and length values for various trials obtained from the FE
Analy si s


Runout values obtained Length of the cantilever
fr-om the ProE model beam, L in mm
0.15255 9.999
0.13903 9.962
0.11989 9.742
0.09795 9.307
0.08245 9.053
0.06642 8.628
0.05142 8.137
0.03647 7.999


Runout values in mm


0.1488
0.1354
0.1182
0.0957
0.0813
0.0645
0.0492
0.0381


Table 5-2. Calculation of the stiffness values corresponding to different spindle speeds
Spindle rotation Maximum amplitude Expected runout Calculated stiffness
speed in rpm of the contact force in mN value in mm values in N/mm
60000 25 0.0012 29.645
66000 25 0.0012 33.683
72000 25 0.0012 38.106
78000 25 0.0012 42.913
84000 25 0.0012 48.105
90000 25 0.0012 53.681



Table 5 -3. Calculation of the lengths of the compliant beams corresponding to the beam stiffness
values for Design 2a
Calculated stiffness Breadth of the Width of the Length of the compliant
values in N/mm compliant beam in mm compliant beam in mm beam in mm
29.645 1.25 0.75 9.725
33.683 1.25 0.75 9.320
38.106 1.25 0.75 8.945
42.913 1.25 0.75 8.597
48.105 1.25 0.75 8.276
53.681 1.25 0.75 7.979










Table 5 -4. Calculation of the widths of the compliant beams corresponding to the beam stiffness
values for Design 2b
Calculated stiffness Breadth of the Length of the Width of the compliant
values in N/mm compliant beam in mm compliant beam in mm beam in mm
29.645 1.25 10.00 0.771
33.683 1.25 10.00 0.805
38.106 1.25 10.00 0.838
42.913 1.25 10.00 0.872
48.105 1.25 10.00 0.906
53.681 1.25 10.00 0.934



Table 5 -5. Calculation of di stance of point of application of force from the fixed end of the
compliant beams corresponding to the beam stiffness values for Design 3
Length of the Breadth of the Width of the.
Calculated stiffness Value of x in
.ausinNm compliant beam in compliant beam in compliant beam in m
mm mm mm
29.645 5.00 0.5 0.25 1.387
33.683 5.00 0.5 0.25 1.319
38.106 5.00 0.5 0.25 1.257
42.913 5.00 0.5 0.25 1.200
48.105 5.00 0.5 0.25 1.149
53.681 5.00 0.5 0.25 1.101













Complantr Oxel


Wn~orkpiece

PLagid fixel











Figure 5-1. ProE model for fixel- workpiece configuration for a milling operation with two
compliant fixels and two rigid fixels








Fx.0 Fy.-16.~F FL.0(Wr
Nbc:O My:0Mr 0(WCS


Workpiece
Dz-0 Dy:- Dz:O) (WCS
Rx:0 IRy:0 Rz:D (WCS


Dr 0 Dy Dr 0 (W"
Rix 0 Ry 0 Rcz 0 NVCS


Interface elecment


Ox 0D0 r.Dz:O (WCS3
Rx:o0Ry:o Rz:o twc9
Fixture base
module


ComphIantb eam


Figure 5-2. ProE model for single fixel workpiece configuration









































_II ~i


Figure 5 -3. FE Analysis to calculate the maximum displacement of the compliant beam


10.5











85



7.5

0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16

Run out value: in mmn




Figure 5 -4. Variation of length of the cantilever beam for corresponding runout values

















___~


Wosrkpiece


Sprin with
attached mass


H~armronic Folrce


Figure 5-5. MSC Adams model for the two fixels-workpiece configuration




model fixture_9th


0.00


0 .0,

5 0.0

E

- 0.00


Time (sec)


Figure 5 -6. Plot showing the Di splacement of the center of mass of the workpiece with respect to
time where the frequency of the harmonic force is 60000 rpm
























0.0-




-0.001- I




-0.002
0.0 0.005 0.01 0.015 0.02
Time (sec)


Figure 5-7. Comparison of the displacement of the center of mass of the workpiece and the
expected runout values for a frequency of 60000 rpm


emMEA_1


2 00.0


- Displacement of the center of mass of the workpiece
Expected runout values












Error between displacement of the center of mass of
workpiece and the expected runout values in the transient stage


1.50E-03


1.00E-03


5.00E-04


0.00E+00
0.01

a 5.00E-04


-1.00E-03


-1.50E-03


Time in seconds

84,000 rpm
- 66,000 rpm


- 90,000 rpm
72,000 rpm


78,000 rpm
;;;;;;; 60,000 rpm


Figure 5 -8. Plot showing the error between the displacement of the center of mass of the
workpiece and the expected runout values in the transient stage









CHAPTER 6
DISCUSSION AND CONCLUSIONS

6.1 Conclusions

The motivation for investigating the fixel designs presented in this thesis is driven by the

conj ecture that through the provision of mechanically adaptable stiffness characteristics in

passive/active fixturing one can potentially improve the dynamic response of the workpiece

during manufacturing of micron features. This study has provided the theoretical modeling of

four such fixel designs and a comparison of what stiffness trends one may extract from their

design.

It is observed that four-bar mechanism type fixel offers greater versatility in that it exhibits

two mechanical variables for making adjustments. But this adds to the complexity of the design

by requiring twice as many actuators for controlling these variables. However, the results

indicate that by controlling one of the parameters, the fixel can still be tuned to control the

dynamic behavior. In addition, the trends observed in the stiffness as a function of the

mechanical variables exhibit local minimums.

For the beam type fixels, Designs 2a and 3 offer continuous adaptability fixel stiffness

while the Design 2b uses incremental adjustments in changing its stiffness. For these three

designs, it was found that similar trends occurred in terms of k versus available range of input

variable. Additionally, the lower bound on the slenderness ratio contributed to an amplification

of the rate of change in k with respect to change in input variable, as well as on the range of k.

Design 2a and 2b yield the greatest of all values for range of k while Design 2b yields the

greatest value for the average slope ofk. But since design 2b is potentially hampered by the

discrete nature of its design, its practical application will be difficult. Hence, considering both









the stiffness characteristics and ease of application, Design 2a is the best among the fixel

designs.

To analyze the performance of the proposed designs, one of the designs (Design 2a) is

analyzed by implementing it in an end milling mesoscale machine tool system where the tool has

runout along the direction of the fixels. The fixel-workpiece system was initially modeled in

ProE and a static analysis was performed, from which it was established that runout

compensation can be achieved by adjusting the length of the compliant beam to obtain beam

deflection equal to the runout error. To simulate a mesoscale tool system, a MSC Adams model

was developed and analyzed under the influence of a harmonic force and other typical mesoscale

system operation parameters. It was observed that runout compensation can be achieved in this

case as well, but the length of the compliant beam was now a function of the frequency of the

spindle rotation and the mass of the workpiece in addition to the stiffness of the compliant beam.

To conclude, four different fixel designs have been presented that exhibit the mechanical

adaptability of their stiffness characteristics and analysis has been performed on one of the

designs to prove the implementation of the design concept for fixturing for mesoscale

manufacturing. A provisional patent for the designs proposed in this thesis has been obtained.

6.2 Directions for Future Research

For further research, these designs need to be tested experimentally to further quantify their

ability to control the dynamic response of the tool-workpiece-fixture interface. In addition,

implementation issues in terms of actuators and sensors and their real-time control and effects of

friction in the fixture-workpiece platform (support surface) need further study. It should be noted

that the research in this thesis focuses on application of the fixel designs within the fixturing

system by using two or four fixels. The determination of the specific number and optimal









locations of the fixels required, which is workpiece and process dependent, has been left for

future work. Also, the determination of the optimal contact force has been left for future work.

It has been observed from the analysis of the fixel designs that the workpiece-fixture

system under the influence a harmonic force experiences a transient period before reaching the

steady state. The effect of this transient nature of the displacement of the workpiece on the

accuracy of the process will be dependent on the feed rate values. A methodology to reduce the

duration of the transient state needs to be developed which will be done during further research.

Also, the sensitivity of the results of the analysis to variations in the mass of the workpiece has

not been studied. This will be studied in future work to quantify its significance on part accuracy

for such scenarios in which material is being removed.










REFERENCES


[1] S. W. Lee, R. Mayor and J. Ni, "Dynamic analysis of a mesoscale machine tool," Joumnal
of Manufacturing Science and Engineering, 128 (1), 194 203, 2006.

[2] A.H. Honegger, S.G. Kapoor and R.E. DeVor, "A hybrid methodology for kinematic
calibration of micro/meso-scale machine tools (mMT s)," ASME Joumnal of Manufacturing
Science and Engineering, 128 (2), 13 18, 2006.

[3] H. Li, X. Lai, C. Li, Z. Lin, J. Miao, J. Ni, "Development of meso-scale milling machine
tool and its performance analysis," in the Chinese Journal of Mechanical Engineering, 42 (11),
162 167, 2006.

[4] J. Bronson, G. Wiens, J. Palmer and J.R. Meyer, "Feasibility study of mechanically
adaptive microfabrication process tooling and fixturing," 1 st International Conference on
Micromanufacuring, ICOMM 2006, University of Illinois at Urbana-Champaign, IL, 2006.

[5] G.J. Wiens, R.M. Wagner, A. Valdes and J.R. Bronson, "Mechani cally-adaptive fixture
element for directional control of contact interface forces," International Conference on Micro-
Manufacturing 2007, Clemson University, SC, 2007.

[6] M.L. Culpepper and S. Kim, "Six-axis reconfigurable nanomanipulator systems for
positioning in nano-scale manufacturing and research," International Symposium on Nano-
Manufacturing, ISNM 2004, Daej eon, Korea, 2004.

[7] U.A. Tol, G.J. Wiens and J. Schueller, "Mechanical adaptability of active fixturing
system," submitted for refereed journal publication, 2009.

[8] B.S. El-Khasawneh and P.M. Ferreira, "Computation of stiffness and stiffness bounds for
parallel link manipulators", Intemnational Journal of Machine Tools & Manufacture, 39 (1), 321 -
341, 1999.

[9] G.J. Wiens and D. S. Hardage, "Structural dynamics and system identification of parallel
kinematic mechani sms," in Proceedings of the International Design Engineering Technical
Conferences, DETC2006, 2006, pp. 10.

[10] W.F. Riley, L.D. Sturges and D.H. Morris, Mechanics of materials, John Wiley & Sons,
New York, NY, pp. 475 484, 1999.

[1l] L.L. Howell, Compliant mechanisms, John Wiley & Sons, New York, NY, 2001.

[12] D.L.Martin and R.O.Wilcox, "Spindle runout effects on postional accuracy," in the
University of Wisconsin Symposium on Small Hole Techonolgy, Palm Springs, California,
1990.

[13] S.S.Rao, Mechanical mibrations, Pearson Eduaction, Inc., New Jersey, NJ, pp 221 225,
2004.









BIOGRAPHICAL SKETCH

Koustubh Rao was born in India, in 1984. He received his bachelor' s degree in mechanical

engineering from the Indian Institute of Technology Guwahati, Guwahati, India in May 2006.

Mr. Rao will have received his Master of Science degree in mechanical engineering from

University of Florida in August 2009. At University of Florida, he worked in the Space

Automation and Manufacturing Mechani sms Laboratory in the Department of Mechanical and

Aerospace Engineering under the guidance of Dr. G. Wiens.





PAGE 1

1 DESIGN AND ANALYSIS OF FIXTURING METHODS FOR MESOSCALE MANUFACTURING By KOUSTUBH J. RAO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MA STER OF SCIENCE UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Koustubh J. Rao

PAGE 3

3 To my parents and my sister Thank you for always being there for me

PAGE 4

4 ACKNOWLEDGMENTS I thank my advisor, Dr. Gloria Wiens, Associate Pr ofessor, Department of Mechanical and Aerospace Engineering, University of Florida, for her continuous support and guidance. I would also like to thank my committee members Dr. Hitomi Yamaguchi Greenslet, Associate Professor, and Prof. John Schueller, Prof essor I express my deep sense of gratitude to my family members for their perennial moral support and encouragement. I also record my special thanks to my friends and lab mates (S pace, A utomation and M anufacturing M echanisms Lab oratory and S pace S ystems G roup ) for making my stay at UF very memorable.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 7 LIST OF FIGURES .............................................................................................................................. 8 LIST OF ABBREVIATIONS ............................................................................................................ 10 ABSTRACT ........................................................................................................................................ 11 CHAPTER 1 INTRODUCTION ....................................................................................................................... 13 1.1 Background ........................................................................................................................... 13 1.2 Motivation ............................................................................................................................. 13 1.3 Focus of Research ................................................................................................................. 14 1.4 Literature Review .................................................................................................................. 16 1.5 Thesis Outline ....................................................................................................................... 18 2 THEORETICAL FIXEL MODEL ASSUMPTIONS ............................................................... 21 2.1 Problem Statement ................................................................................................................ 21 2.2 Assumptions .......................................................................................................................... 21 3 PRINCIPLES OF THE DESIGNS ............................................................................................. 23 3.1 Design 1 ................................................................................................................................. 23 3.1.1 Introduction to Compliant Mechanisms ................................................................. 23 3.1.2 Pseudo Rigid Body Model of a Compliant Mechanism ........................................ 24 3.1.3 PRBM of the Fixel ................................................................................................... 25 3.1.4 Derivation of Stiffness of the Fixel at Contact Point (Along y-direction) ........... 26 3.1.5 Derivation of Stiffness of the Fixel Along x-direction .......................................... 30 3.2 Design 2 ................................................................................................................................. 30 3.2.1 Design 2a) Mechanical Variable L ...................................................................... 31 3.2.2 Design 2b) Mechanical V ariable h ...................................................................... 33 3.3 Design 3 ................................................................................................................................. 34 4 COMPARISON OF THE DESIGNS ......................................................................................... 42 4.1 Design 1 ................................................................................................................................. 42 4.2 Comparison of Designs 2a, 2b and 3 ................................................................................... 43 4.2.1 Design 2a .................................................................................................................. 45 4.2.2 Design 2b .................................................................................................................. 45

PAGE 6

6 4.2.3 Design 3 .................................................................................................................... 46 4.2.4 Plots for Design 2a, 2b and 3 .................................................................................. 46 5 ANALYSIS OF THE FIXEL DESIGNS .................................................................................. 57 5.1 Static Analysis of Design 2a ................................................................................................ 58 5.2 Analysis of Design 2a Under Harm onic Force ................................................................... 60 5.3 Compensation for Transient Dynamics ............................................................................... 65 6 DISCUSSION AND CONCLUSIONS ..................................................................................... 73 6.1 Conclusions ........................................................................................................................... 73 6.2 Directions for Future Research ............................................................................................ 74 REFERENCES ................................................................................................................................... 76 BIOGRAPHICAL SKETCH ............................................................................................................. 77

PAGE 7

7 LIST OF TABLES Table page 5 1 Overview of the runout and length values for various trials obtained from the FE Analysis .................................................................................................................................. 66 5 2 Calculation of the stiffness values corresponding to different spindle speeds ................... 66 5 3 Calc ulation of the lengths of the compliant beams corresponding to the beam stiffness values for Design 2a ................................................................................................ 66 5 4 Calculation of the widths of the compliant beams corresponding to the beam stiffn ess values for Design 2b .............................................................................................................. 67 5 5 Calculation of distance of point of application of force from the fixed end of the compliant beams corresponding to the beam stiffness values for Design 3 ....................... 67

PAGE 8

8 LIST OF FIGURES Figure page 1 1 Conceptual Integration of Fixturing and Mesoscale Machine Tool (horizontal plane of application) (mMT stage courte sy of J. Ni at S.M. Wu Manufacturing Research Center, University of Michigan, Ann Arbor, MI) ................................................................ 19 1 2 Conceptual Integration of Fixturing and Mesoscale Machine Tool (vertical plane of applicat ion). (mMT stage courtesy of Thomas N. Lindem at Atometric, Inc., Rockford, IL) .......................................................................................................................... 20 1 3 Illustration of runout error ..................................................................................................... 20 3 1 De sign 1 Simple monolithic compliant four -bar mechanism consisting of four compliant joints ...................................................................................................................... 36 3 2 Example of a Compliant Mechanism a Crimping Mechanism consisting of three flexible member s ( courtesy of Compliant Mechanisms by Larry Howell) ........................ 36 3 3 Pseudo Rigid Body Model for the Four bar monolithic fixture .......................................... 37 3 4 The various configurations of Design 1 during stiffness calculations a) Initial Configuration b) Symmetric Equilibrium with Base Length L1 c) Displacement Due to Contact Force with Base Length L1 d) Change in Displacement Due to a Change in Contact Force with Base Length L1 .................................................................................. 38 3 5 Design 2 Principles Involved ............................................................................................. 39 3 6 Implementation of Design 2a with Four Fixels .................................................................... 39 3 7 Implementation of Design 2b with Four Fixels .................................................................... 40 3 8 Design 3 Principle Involved ............................................................................................... 40 3 9 Implementation of Design 3 with Four Fixels ...................................................................... 41 4 1 Fixel Stiffness (L2=L4=45mm, L3=30mm, all ks are equal) .............................................. 48 4 2 (kmin / Eb ) for Designs 2a, 2b and 3 versus [ r s ] .................................................................... 49 4 3 (kmax / Eb ) for Designs 2a and 2b versus [ r s ] ........................................................................ 50 4 4 (kmax / Eb ) for Design 3 versus [ r s ] ........................................................................................ 51 4 5 ( / Eb ) x 105 for Designs 2a and 2b versus [ r s ] ............................................................... 52 4 6 k / Eb ) x 105 for Desig n 3 versus [ r s ] ............................................................................... 5 3

PAGE 9

9 4 7 (Average Slope of k / Eb ) x 104 for Design 2a versus [ r s ] ................................................. 54 4 8 (Average Slope of k / Eb ) x 103 for Design 2b versus [ r s ] ................................................. 55 4 9 (Average Slope of k / Eb ) x 105 for Design 3 versus [ r s ] ................................................... 56 5 1 ProE model for fixel workpiece configuration for a milling operation with two compliant fixels and two rigid fixels ..................................................................................... 68 5 2 ProE model for single fixel workpiece configuration ....................................................... 68 5 3 FE Analysis to calculate the maximum displacement of the compliant beam ................... 69 5 4 Variation of length of the cantilever beam for corresponding runout values ..................... 69 5 5 MSC Adams model for the two fixels -workpiece configuration ........................................ 70 5 6 Plot showing the Displacement of the center of mass of the wor kpiece with respect to time where the frequency of the harmonic force is 60000 rpm ........................................... 70 5 7 Comparison of the displacement of the center of mass of the workpiece and the expected runout values for a frequency of 60000 rpm ........................................................ 71 5 8 Plot showing the error between the displacement of the center of mass of the workpiece and the expected runout values in the transient stage ........................................ 72

PAGE 10

10 L IST OF ABBREVIATIONS mMTs mesoscale Machine Tools FIXEL Fixture Element PKM Parallel Kinematic Mechanism PRBM Pseudo Rigid Body Model TIR Total Indicated Reading

PAGE 11

11 Abstract of Thesis Presented to the Graduate School of the Unive rsity of Florida in Partial Fulfillment of the Requir ements for the Degree of Master of Science DESIGN AND ANALYSIS OF FIXTURING METHODS FOR MESOSCALE MANUFACTURING By Koustubh J. Rao August 2009 Chair: Gloria Wiens Major: Mechanical Eng ineering T he workpiece -fixture and the tool -workpiece interactions are important factors for a manufacturing process and play a major role in the efficacy of the process They lead to non beneficial machine performance in terms of in accurate dimensioning and improper finish ing of the workpiece. The se interactions become even more important for mesoscale manufacturing because of the higher accuracy requirements. In addition, t he type of fixturing used o n mesoscale machine tool system sets a limit on the abi lity to achieve high accura cy of dimensions F ixturing also remains a critical issue impeding the integration and autonomous operation of micro/mesoscale manufacturing systems. To push beyond th ese accuracy limits, innovative work holding approaches are needed. T his study presents a n investigation of four fixture element (fixel ) designs for fixturing to be incorporated into mesoscale manufacturing systems. These fixel s help in improving the process accuracy by taking into account the effects of tool runout Using compliant mechanisms and components (e.g., monolithic four bar mechanisms and/or cantilever beams), fixels exhibiting mechanically adjustable stiffness characteristics are achievable. Manually or automating the stiffness adjustments, these fixels p rovide a functionality for enabling greater control of the response of the workpiece due to runout and variation in contact forces at the tool -workpiece -fixture interface. The method of implementing these fixel

PAGE 12

12 designs (using four fixels) for common manufa cturing operations is suggested in addition to the designs Each design has specific mechanical parameters for its fixel s and adjustable stiffness characteristic s are achieved by varying these parameters The monolithic four bar design has two mechanical v ariables whereas the three cantilever/compliant beam fixel designs require actuation of only one mechanical parameter. To quantify the fixel functionality and its dynamic range, the theoretical models of the stiffness characteristics expressed as a functio n of these mechanical variables are presented for each of the designs Upon establishing a common stiffness range for the three compliant beam fixel designs, a metric is formed for better comparison between the designs. This metric is based on the sensitivity of stiffness expressed as a function of slenderness ratio and an operation range, bounded by a minimum possible stiffness value shared by the cantilever beam fixture models. The slenderness ratio is defined as the ratio of the minimum length of the can tilever beam to its maximum width while the operation range is determined by the ratio of the minimum and maximum possible values of the fixels mechanical variable. Using this metric, results are generated for each of the designs and then compared with one another to highlight the advantages and disadvantages of each design Models for one of the fixel designs are then developed to perform dynamic analysis and understand the behavior o f the fixturing design under manufacturing operation conditions.

PAGE 13

13 CHAP TER 1 INTRODUCTION 1.1 Background The growing trend toward s miniaturization has impacted technologies in virtually every field, from medicine to manufacturing. Consequently, a dramatic shift is occurring within the manufacturing paradigm toward the development of complementary capabilities for producing miniaturized products. Furthermore, this shift has led to research efforts on micro/meso levels, bridging the gap between the micro and macro worlds T he creation of mesoscale machine tools (mMTs) that are l ess expensive and more portable than conventional precision machine tools is an indication to these efforts However, mesoscale manufacturing is still faced with critical issues in accuracy due to the adverse affects resulting from tool -workpiece and the workpiece -fixture interactions. These inaccuracies could be due to runout in the tool, misalignment of the tool/workpiece, workpiece material properties, etc Th ese challenges are magnified for the complicated process of creation of micro n features on micr o and macro -sized parts where fixturing and material handling pose a problem This is an area with significant importance, yet there has been limited research on addressing th e se problems Th e objective of this study is to contribute towards research on m eso scale manufacturing by introducing fixturing designs which can facilitate controlling the tool -workpiece interface dynamics 1.2 Motivation In prior research related to mesoscale manufacturing designs for reconfigurable manipulator s consisting of compl iant elements and actuators for part positioning and handling have been developed. These methods and devices typically manipulate components that are adhered to the device or a base material and do not have the ability to directly control the tool -

PAGE 14

14 workpiec e interactions By developing fixturing devices which have the ability to passively or actively control the se interactions, the repeatability and precision of mesoscale machine tools can be enhanced. In the context of controlling the tool -workpiece interfa ce dynamics mechanical adaptability is a key factor and can be of advantage to the current situation T he feasibility of mechanical adaptability at the mesoscale and macro -scale for fixturing has been successfully demonstrated but has not been implemented For the work reported in this thesis, the premise of the active fixturing can be stated as follows. If f ixturing methods which are capable of providing stiffness adjustments are developed they will enable greater control of the response of the workpiece due to runout and variation in contact forces at the tool -workpiece -fixture interface Consequently, it is necessary to develop designs and methods of implementation for mechanically adaptable fixturing devices comprising of fixels (fixture elements). Suc h a fixturing system is illustrated in Figure s 1 1 and 1 2 consisting of four fixels and integrated with a mesoscale machine tool. In the system illustrated in Figure 1 1, the fixels are in a horizontal plane whereas the plane of application of the fixtur e designs shown in Figure 1 2 is in the vertical plane. In such a system, gravitational effects will also play a major role. Once fixel designs are developed, a method to compare them would be required along with the analysis of the designs to validate the m 1.3 Focus of Research T his thesis focuses on designs for fixels for passive/ active fixturing which can provide mechanically adjustable characteristics In this thesis four cand idate designs are presented for fixturing mesoscale workpieces. The first design consists of monolithic compliant four -bar mechanism type fixels. The other three fixturing designs consist of cantilever beam type fixels. These designs must be evaluated in their ability to control stiffness, both in range and direction, via a minima l set of mechanical adjustments. To quantify the fixel functionality and its stiffness

PAGE 15

15 range, this study presents the theoretical models for determining the stiffness characteristics at the point of fixel -workpiece contact These stiffness characteristics are expressed as functions of the mechanical variables of each of the fixel designs Through variable stiffness characteristics, the response of the workpiece due to runout and variation in contact forces at the tool -workpiece fixture interface can then be better controlled. In order to determine the capabilities of these fixturing methods the modeling and analysis of candidate fixel designs is done T he stiffness adjustability of each of such fixel s can be integrated in a coordinated manner with other (ad justable) fixels The fixturing system formed by combining these independent fixels with the workpiece takes the form of a parallel kinematic mechanism (PKM). In this PKM setup, the fixels act as the independent kinematic chains and the workpiece is the en d -effector, thus forming a closed loop system. Leveraging these similarities with the PKM, passive/active fixturing also has the potential of providing a dual -capability of fixturing as well as manipulating micro/mesoscale workpieces [8], [9]. Furthermore, via stiffness adjustments of the fixels, the fixturing device will have the ability to tune the kinematics and dynamics o f the mesoscale machine tool and control the tool -workpiece interface dynamics. These characteristics can be used to reduce the errors caused by tool runout to improve the accuracy of the process and reduce the effects on the tool properties. This stiffness relationship can then be applied towards development of efficient control algorithms implemented through actuated adjustments of the mechanical variables of the fixels. While this thesis focuses on the mMT application, it should be noted that the methodology presented herein is applicable to both macro and mesoscale manufacturing. Furthermore, mesoscale manufacturing is performed using various machine tool sizes that vary from small to medium sized desk top machine tools to the more traditional sized machine tool.

PAGE 16

16 1. 4 Literature Review As mentioned in the previous section, there has been an increase in the research on developing more e fficient and accurate mesoscale machine tools (mMTs). Some of the mMTs that have been developed or are being developed are less expensive and more portable than conve ntional precision machine tools This is achieved by developing a thorough study of the dy namic behavior of the mMT [1], developing different methods for their calibration [2] and evaluation of the performance of the developed system s in mesoscale processes There has been prior research done on better understanding the process parameters that have a large effect on a mMT system and estimating the optimal operation conditions for such systems. The parameters such as depth of cut, magnitude of cutting forces, feed rate and spindle rotation speed impact the performance of a mMT system. The spindl e speed is usually high for such systems and is in the order of 60000 rpm ( with certain processes reaching speeds up to 120000 rpm ) while t he cutting forces are usually in the order of 10 mN [3]. Any design of a mesoscale system or it s components should be developed and analyzed for these values of the parameters. In recent years, research is being carried out on using compliant mechanisms for developing positioners for mesoscale systems which assist part positioning b y reducing the alignment errors Culp epper and Kim [6] developed a sixaxis reconfigurable manipulator consisting of compliant elements and actuators for part positioning. The characteristics of compliant mechanisms can then be extended to not just positioners but also to fixturing systems. Since the feasibility of mechanical adaptability at the mesoscale [4] [5] and macro -scale [7] for fixturing has been already been successfully demonstrated (without implementation), research needs to be done to take advantage of these properties. The use of compliant mechanisms to achieve variable stiffness characteristics can therefore improve the performance of mMTs.

PAGE 17

17 The performance of a mMT is dep endent on many process factors such as the spindle rotation speed, the runout in the tool, feed rate, cuttin g force and the depth of cut. Tool runout is one of the important parameters which directly affects the accuracy of the mesoscale manufacturing process [3]. Runout, often measured as Total Indicated Reading (TIR), is defined as the difference between the m aximum and the minimum distance of the tool point from the axis of rotation measured over one revolution as illustrated in Figure 1 3 [12]. Radial runout is the result of a lateral (parallel) offset between the rotational axis of the tool and the central a xis of the collet/spindle system Runout can also occur due to variation in the length of the teeth of the cutter resulting in varying depth of cut. A longer tooth will have a deeper cut as compared to the teeth of the correct size, thus affecting the accu racy of the machining process. Th e runout from the spindle of the system (or from varying teeth dimensions) causes a n extraneous force on the workpiece and results in increased excitations [3]. Even though there exists a relationship between runout and the resulting contact force, researchers have yet to define the specifics of this relationship in terms of magnitude, frequency and amplitude as a function of the runout and tool workpiece material properties. Non -contact p recision c apacitance s ensors are ty pically used to dynamically measure the runout of a tool. These sensors use capacitance technology to measure the runout and provide accurate readings for frequencies up to 120,000 rpm There are many factors which affect the tool runout such as frequency of spindle rotation the duration of the operation and the tool -workpiece interactions. The runout value increases for higher frequencies and these values change depending on the precision of the setup. T he typical values of runouts for high precision syst ems are in the range of 0.001 mm to 0.006 mm and in the range of 0.03 mm to 0.05 mm for lower precision systems [12] Although considerable research is being carried out to achieve reduced runout it still remains a challenge. Hence,

PAGE 18

18 alternate approaches a re required to achieve control on the adverse affects of tool runout Using fixturing methods with variable stiffness characteristics is one such approach. 1. 5 Thesis Outline This thesis is organized in the following manner. After giv ing a brief introducti on to the growing trend towards micro/mesoscale manufacturing and the challenges faced in accuracy due to the adverse affects resulting from tool -workpiece interactions in Chapter 1, Chapter 2 summarizes the problem statement followed by a discuss ion of th e assumptions made in developing the fixel designs This is followed by Chapter 3 which provides a detailed description of the principles of each of the four fixel designs and the relationships between fixel stiffness and the specific design parameters. Ch apter 4 describes the metric obtained to compare the compliant beam fixel designs followed by the plots depicting the stiffness characteristics of the fixel designs. Chapter 5 deals with the implementation and analysis of one of the fixel designs in a meso scale manufacturing process under typical operation conditions Later Chapter 6 identifies the major conclusions and highlights areas where this work needs to be further developed.

PAGE 19

19 Figure 1 1. Conceptual Integration of Fixturing and Mesoscale Machine Tool (horizontal plane of application) (mMT stage courtesy of J. Ni at S.M. Wu Manufacturing Research Center, University of Michigan, Ann Arbor, MI)

PAGE 20

20 Figure 1 2. Conceptual Integration of Fixturing and Mesoscale Machine Tool (vertical plane of application) (mMT stage courtesy of Thomas N. Lindem at Atometric, Inc., Rockford, IL) Figure 1 3 Illustration of r unout error

PAGE 21

21 CHAPTER 2 THEORETICAL FIXEL MO DEL ASSUMPTIONS As stated in Chapter 1, the overall objective of this thesis is to character ize the functionality and capabilities of an active fixturing methodology for mesoscale manufacturing. This requires a clear understanding of the compliant nature of each fixel and the system of fixels as well as the dynamic impact of active fixturing on the tool -w orkpiece interface dynamics. Chapter 1 first provides a description of the active system and how it may be implemented on two different typical micro/mesoscale machine tools (vertical and horizontal planes of application). This is followed by a narrowing of the problem statement to fixel design development, modeling and analysis in terms of fixel stiffness and dynamics. In this thesis, the workpiece tool interface scenario used as a basis for analysis and demonstration is tool runout and its adv erse effects on feature dimensional accuracy. 2.1 Problem Statement The focus of the current study is to develop fixture designs which have variable stiffness characteristics and can be implemented on typical mesoscale machine tool systems After develo ping the designs, modeling and analysis is to be performed to validate the implementation of the design in a typical mMT setup. Four different fixel designs are considered in this study a compliant four -bar mechanism type fixel and three beam type fixel s with the beam type fixels being fundamentally similar 2.2 Assumptions While the four bar mechanism based fixel design consists of compliant components and will typically be fabricated as a monolithic compliant structure, the modeling of this design is do ne using its PseudoRigid Body Model (PRBM). Therefore, the assumption for the compliant four bar mechanism is a four bar mechanism consisting of rigid links with spring loaded joints.

PAGE 22

22 Comparatively, there are different methods for obtaining the compliant structure. One method is to locate all the compliance within joint centric locations of the device for which the lengths of the rigid segments are large relative to the lengths of the flexural members. The flexure members provide localized compliance an alog ous to a spring loaded kinematic joint. Another method is to have a structure with a kinematic topology that yields motion under loading analog ous to a four bar mechanism. Either of these methods can be modeled using the PRBM method yielding kinematic al ly and dynamically equivalent designs. In this thesis, the first method is assumed (details of which can be found in a later section 3.1.2). It should however be noted that the second method is anticipated the preferred method for yielding the designs to be fabricated and implemented in the commercial application. This will be the subject of future research and development. For the beam type fixels three different implementati on approaches are explored. The fundamentals of the beam type designs arrive from solid mechanics principles for beam deflections As known from these fundamentals, the beam deflections can be expressed as a function of a given load and support locations [10]. From these principles, the fixels stiffness (proportional to force o ver displacement at the point of fixel -workpiece contact) can be quantified for each fixel configuration. The Euler Bernoulli beam theory assumption is made in deriving the theoretical models, i.e., the length of the fixel beam element (denoted by L ) is c onstrained to be at least 10 times larger tha n its cross-sectional thickness ( denoted as width h herein ) as shown in Equation 2 1. /10 Lh (2 1) Th is ratio of L to h is an important parameter for the designs which is explained in t he later sections.

PAGE 23

23 CHAPTER 3 PRINCIPLES OF THE DE SIGNS The principles involved in each of the f our fixture designs proposed in this study and the mechanical parameters involved in each of the designs are described in this chapter. The relationship between these parameters and the fixel stiffness is then established followed by developing the variable stiffness characteristics of the designs. 3.1 Design 1 The first fixture design under consideration is one in which each fixel is a monolithi c compliant four -bar mechanism consisting of links and compliant flexure joints as illustrated in Figure 3 1 The length of the links and the length of the flexural joints are the characteristics of this design. To simplify the modeling, an equivalent pseudorigid -body mod el (PRBM) of this fixel is developed in this study The next section gives a brief introduction to compliant mechanisms and the development of equivalent PRBMs. 3. 1 .1 Introduction to Compliant Mechanisms Compliant mechanisms are mechanisms which transfer motion, an input force or energy from one point to another, using flexible members. Unlike rigid body mechanisms, where the mobility of the mechanism is achieved through the degree of freedom in joints (kinematic pairs) compliant mechanisms achieve mobili ty through the elastic deformation of the flexible members. An example of a compliant crimping mechanism similar to a vice grip is shown in Figure 3 2 This mechanism consists of three flexible members, unlike in a vice grip which consists of revolute join ts instead of these flexural pivots. The input force applied on the hand grips results in some energy being stored in the mechanism in the form of strain energy in the flexural pivots. Some examples of other common compliant devices are binder clips, backp ack latches, paper clips, etc.

PAGE 24

24 The two main advantages of compliant mechanisms are cost reduction and increased performance [11]. Since compliant mechanisms are mostly monolithic and do not require kinematic pair type joints or pins, this results in reduce d part count as compared to a rigid body mechanism. This directly results in reduction in assembly time improved manufacturing process reliability and the ability to be easily miniaturized. Since there is no relative sliding motion between the surfaces of the joints as seen in rigid body mechanisms, friction i s generally assumed negligible in compliant devices. This results in reduced wear and thus reduced maintenance in terms of lubrication. Such mechanisms also do not ha ve the inherent problem s of backla sh and/or joint tolerances which is common in jointed mechanisms. Thus, these mechanisms also exhibit increased precision and increased reliability. 3.1.2 Pseudo Rigid Body Model of a C ompliant M echanism Although there are numerous advantages of compliant mechanisms, there are some challenges faced with respect to analyzing and designing them. The most well known technique to design and analyze compliant systems that undergo large deflections is the use of Pseudo Rigid Body M odel (PRBM) In this concept, the deflection of flexible members is modeled using equivalent rigid body components. The flexural pivots are modeled as kinematic pair type joints/pins with t he stiffness of the flexural members r epresented by torsional springs within the joints For each P RBM, the most important consideration is the placement of the revolute/pin joints and the value of the spring constant to be assigned [11]. A monolithic compliant four bar mechanism consists of four rigid links connected by flexural pivots. A PRBM for such a mechanism is modeled by placing revolute joints at the center of the flexible segments and torsional springs at each of these joints. The placement of the joints is based on the assumption that the lengths of the flexural members are small relative to t he lengths of the rigid segments.

PAGE 25

25 3.1.3 PRBM of the F ixel Figure 3 3 presents the pseudo -rigid -body model of the four -bar mechanism type fixel developed as explained in the previous section. Each of these four bar fixels will provide directional variable stiffness characteristics to the workpiece fixturing setup within its plane of motion In this model, the joint flexur al pivots are modeled as revolute joints (2, 3, 4 and 6) as shown in the figure. The stiffness of the flexural pivots is expressed by addi ng torsional springs of spring constants 234,, kkk and 6, k respectively. The base link has a length of L1 while the rigid links have equivalent constant link lengths of L2, L3, L4, and L5, and the contact point of t he fixel with the workpiece occurs at the end of link L5 which acts as the interface element. The angles between the links L1 and L2 is labeled as 2 (called the base angle), the angle between L1 and L4 as 4, the angle between L2 and L3 as 3, the angle b etween L3 and L4 as 6 and the angle made by link L5 with the vertical is denoted as 5. Due to the geometry of the four bar model, t he values of the angles 3, 4, 5 and 6 are functions of 2 and the link lengths and hence, can be obtained for any confi guration of the model. These relationships are expressed using the Equations 3 1. 22 36 12122 1 3 1236 1222 3364336 3 33 3 1 4 36 4 3 6234 5 23(2cos) sin(sin/) cos(( )/2) sin(sin/) 2a b a b bdsqrtLLLL Ld LdLLd dL (3 1) As mentioned previously, the tip of the interface element (link L5) is in contact with the workpiece. The x and y coordinate s (xc, yc) o f this point contact with respect to the X, Y frame affixed to the base link are given by Equations 3 2. Hence, the change in yc in one configuration

PAGE 26

26 to another configuration represents the displacement of the point of contact which will be required to cal culate the (vertical) stiffness constant at the point of contact. 555 1223cos1/2 (1/2)cossincxLLLL 555 223sin(1/2)sincoscyLLL (3 2) For this design, t he stiff ness characteristics of each fixel as seen at the point of fixel workpiece contact is mechanically adjusted by extending the base length ( L1) and/or an input angle (2). Hence, L1 and 2 represent the mechanical parameters for this design with the stiffness constant being a function of these parameters. 3.1. 4 Derivation of S tiffness of the F ixel at C o ntact P oint ( A long y direction) To derive the stiffness values f or the PRBM, the link lengths were chosen such that L2= L4 and L3 is shorter tha n the other links. Also, all flexures (torsional springs) are assumed to have equal stiffness values. During the equilibrium configuration of the fixel, there is no net force acting on the contact point and hence, the joint flexures torsio nal loads are calculated using angles measured from th is unloaded equilibrium At this equilibrium configuration, all the link le ngths and the respective angles are at their initial values and the link L3 is horizontal. At this configuration, the length of base link is denoted as L1o and 2 is denoted as 2o as shown in Figure 3 4a (the remaining angles are also denoted with similar subscripts). Wh en one or both of the mechanical parameters of the design ( L1 and 2) are varied from their equilibrium values assuming contact is not lost, it results in a contact force ( Fc) acting at the workpiece-fixture interface and a corresponding c hange in the y-coordinate of the point of contact ( yc). When the base length and/or the base angle are again changed, it results in a corresponding change in Fc and yc (denoted by cF and cy ) as compared to t he previous configuration The ratio of th is change in contact force to the change in the position of the point of contact is defined as the

PAGE 27

27 stiffness of the fixel at the point of contact. Hence, to derive the stiffness value, the corresponding values of cF and cy are calculated. The procedure to calculate this stiffness is described next. The base link length and the base angle can each be varied within particular kinematic ranges with the initial value of th e base length operation range being L1o. The range of values for base length can be chosen according to the footprint of the setup. To calculate the range of values of the base angle, a particular value of base length is chosen from the range and the unloa ded equilibrium configuration for this value (as shown in Figure 3 4a) is first determined Each of these unloaded equilibrium configurations, result in a four bar mechanism and the two toggle positions (motion limits) for this four bar model are then determined These limits provide the c orresponding kinematic range of base angles that comply with the design at that particular base length This process is repeated for all possible values of the base length and the individual range of values of base angles is obtained for all the base lengths From these different sets of ranges of base angles, the values which are common to all the sets represent the final operation range for the input base angle 2. T o calculate stiffness, the base length is first changed to a particular value in the operation range (say L1) using a prismatic actuator shown in Figure 3 3 Due to the geometry of the four bar model, this change in the base length results in the base angle changing from 2o to 0' 2 as sh own in Figure 3 4b (with the remaining angles similarly changing to 3o, 4o and 6o). Then, the base angle is increased to a certain value (say 2i) within the operation range using the revolute joint actuator shown in Figure 33. This increment of will result in a change in the angles of the model ( i.e. change in 3i, 4i and 6i) along with a displacement of the link L5 and a contact force (say Fc i) acting on the link L5. The contact force occurs on the link L5 so that the

PAGE 28

28 model remains in static equilibrium. The configuration of the fixel at this instant is shown in Figure 3 4c. At this configuration a static force analysis is carried out on the model and the equation shown in Equation 3 3 is obtained. 1 32 4,34,4 4 32 5,5 5,15,25,35,4 5,6 5 34 6 6,56,6 34100000 0 101010 0 010101 0 00 00 0000xi yi xi yi xi yiC CF F F MM V F MMMMMM V F V MM F (3 3 ) where Flm is the f orce acting on link l by link m subscripts x and y indicate the corresponding ( x y ) components of the forces Flm and the subscript i denotes the current configuration. In ad dition the values of V4, V5, and V6 for the cor responding torsional springs are obtained using the value of the spring constant and the change in angle at the particular joint as shown in Equation 3 4. The elements of the inverse matrix in Equation 3 3 are functions of the link lengths and the angles a nd are given by Equation 3 5. 00 00 00'' 4222333 '' 5 333666 '' 6666444** ** **ii ii iiVkk Vkk Vkk (3 4 )

PAGE 29

29 4,322 4,422 55 55 5,1 5,2 33 55 5,3 5,4 33 5,5 5 5 5,6 6,544*sin() *cos() *cos() *sin() *sin() *cos() 22 *sin() *cos() 22 *sin() ii ii ii ii iMLML MLML LL MM LL MM ML 6,644 *cos()iML (3 5 ) Using the above equations, the value of the contact force at the current configuration ( yiCF ) is calculated. The value of y -coordina te at this instant ( iCy ) is then calculated using the expression for yc shown in Equation 3 2 and the current values of the angles. Keeping the base length fixed, the base angle is then incremented to a nother value (say i+12 within the kinematic range ) which results in a corresponding change in the remaining angles. The new configuration of the model at this instant is shown in Figure 3 4d. A static force analysis is again conducted on the model in this configuration and equations similar to Equation s 3 3 to 3 5 are obtained. Using these equations, the value of the contact force at the new configuration ( 1 iyCF ) and the value of y -coordinate at this new configuration ( 1 iCy ) are then calculated. The difference in the contact forces at the new configuration and the previous configuration is calculated and is denoted by 1()ciF followed by calculating the displacement of the point of contact, 1()ciy (difference in the y -coordinates of the two configurations). The ratio of 1()ciF to 1()ciy gives the numerical approximation of the stiffness at the poin t of fixel workpiece contact corresponding to the current configuration ( L1 and i+12 ). This procedure is then repeated by keeping the length of the base link fixed and varying the base angles through the kinematic range of values to obtain different stiffness values for the d ifferent configurations ( L1 and 2 s). Another set of stiffness values is then obtained by

PAGE 30

30 changing the base length to another value (say L2) from the known range and repeating the procedure for all possible base angles. These stiff ness values correspond to the new set of configurations ( L2 and 2 s). By repeating this procedure for all base lengths and all base angles within the allowable range, a final set of all the possible stiffness values (along y-direct ion) is obtained. Thus, using this fixel design, adjustable fixel stiffness can be achieved by varying the two mechanical parameters the base length and the base angle of the fixel. 3.1. 5 Derivation of S tiffness of the F ixel A long x -direction The stiffne ss of the fixel along the x-direction for all possible values for L1and2 in the kinematic range of values are calculated by following the procedure similar to the derivation of the stiffness along ydirection. The value of the contact force along the x -di rection ( xiCF ) is calculated using Equation 3 3. The value of x-coordinate of the point of contact ( iCx ) is calculated using the expression for xc shown in Equation 3 2. Repeating all the steps followed in sec tion 3.1.4, a final set of all the possible stiffness values (x -direction) for all base lengths and all base angles within the allowable range is obtained. Hence, the Design 1 provides directional adjustable stiffness characteristics in X and Y, and has tw o mechanical variables for adjustment. 3.2 Design 2 Fixel Design 2 is a cantilever beam fixel with unidirectional variable stiffness characteristics. This design is a basic cantilever beam (beam fixed at one end) with contact between the workpiece and the fixel occurring at its free end. For a cantilever beam, the minimum value of k or stiffness, occurs at the free end of the beam. For a given material, this minimum k is a function of the length L of the beam measured from the fixed end and its moment of i nertia I F or the designs shown herein, the fixels are assumed to have rectangular

PAGE 31

31 cross -sectional areas with breadth b and width h Thus, it is understood that k is a function of L and h and this dependency is advantageous in designing fixels. For the cantilever beam type fixel designs proposed in this section, the mechanical adaptability is achieved individually varying L or h Thus, for these designs, there is a single mechanical parameter (single actuation) unlike Design 1. 3.2.1 Design 2a) M echanica l V ariable L The cantilever beam type fixel for Design 2a is shown in Figure 3 5 The integration of the four fixel s with the workpiece setup is shown in Figure 3 6 which consists of the cantilever/compliant beam, the fixture base module and the interface element. To decouple residual motions from occurring when making changes in the mechanical variables, the fixel contact with the workpiece is acquired via the interface component. One end of the interface component has direct contact with the workpiece an d the other end is connected to a point on the cantilever beam (near the free end). The contact force FC between the workpiece and the fixel acts along the interface element and occurs at this point on the cant ilever beam. This contact force will vary depe nding on the loads involved in the machining process. The fixture module clamp s the cantilever beam at the base with the length of the beam L being defined as the distance between this fixed/ constrained end and the point of action of the contact force (ne ar the free end of the beam ). Using the cantilever beam equation for stiffness, the principle involved in fixel Design 2a is to adjust the stiffness k at the point of fixel-workpiece contact ( near the free end of the cantilever beam) by modifying the beam length L Mathematically, the stiffness can be expressed as shown in Equation 3 6 where for fixel Design 2a 334 EIEbh and is a constant, and the mechanical variable is L

PAGE 32

32 3 3333 4 EIEbh k LLL (3 6 ) In response to the fixturing ne eds, the length of the cantilever beam could be controlled by unclamping the beam from the fixture module base and then moving it by the required length ; thus changing the effective beam length L (a s illustrated in Figure 3 6) In this design approach, t he value of k given by Equation 3 6 can be adjusted passively during fixturing set up as well as actively during the process. For the Design 2as fixel, the minimum value of k would occur when the length is Lmax and maximum value of k corresponds to Lmin as shown in Equations 3 7 and 38. 3 min 33 maxmax4 Ebh k LL (3 7 ) 3 max 33 minmin4 Ebh k LL (3 8 ) For the purpose of comparisons between fixel designs, a new parameter r is introduced which is defined as the ratio of minimum to maximum values of the f ixels mechanical va riables. This ratio provides a dimensionless measure for quantifying the mechanical variables range of motion, facilitating the comparison of the different fixel design types. For Design 2a, r is the ratio of Lmin to Lmax and substitut ing r into Equation 3 7 results in the E quation 3 9 which provides a relationship between the design constants (material and geometric parameters), kmin, and the range of moti on of the mechanical variables r 333 min 33 minmin4 Ebhrr k LL (3 9 ) Hence, for D esign 2a, Lmin and r are the two unique parameters. The selection of Lmin and r determine the value of Lmax and b y varying L from Lmin to this calculated Lmax, a set of values of k

PAGE 33

33 that are achievable for each given fixel design are obtained For investiga ting the mechanical adaptability of the fixel, the design variations of Design 2a are explored for different sets of the design parameter r for a given Lmin and kmin. 3.2.2 Design 2b) M echanical V ariable h The cantilever beam type fixel for Design 2b is shown in Figure 3 5. The integration of the fixel with the workpiece setup is shown in Figure 3 7 which is similar to Design 2a (four fixels). In this design, the length of the cantilever is held constant and the adjustable fixel stiffness is achieved by v arying the width h of the beam as shown in Figure 3 7 This can physically be achieved by adding multiple strips of same length and breadth to the cantilever beam thus changing the effective width h The effective width can be calculated using the Equation 310 where n is the number of thin strips and h is the width of each thin strip. '() hhnh (3 10) Since h has to be increased discretely (unlike L in Design 2a), a very small value of h can be chosen such that the change in k is equal to the least count of the system (the least value of change in k required) The value of h (and thereby the number of thin strips) needed to obtain a required k can be calculated using Equation 3 10. The width could be changed from a minimum value of hmin to a maximum value of hmax which would correspond to a minimum and maximum value for fixel stiffness k given by the E quations 3 11 and 3 12 where for fixel Design 2b would be a constant defined as 34 EbL and the mechanical variable is h 3 3 min min min 34 Ebh kh L (3 11) 3 3 max max max 34 Ebh kh L (3 12)

PAGE 34

34 Similar to Design 2a, introducing r as the ratio of hmin to hmax and subs tituting r into Equation 3 11, results in Equation 3 13. 33 33 max min max 34 Ebrh k rh L (3 1 3 ) Here, hmax and r are the two unique parameters of Design 2b and fixed values for these parameters determine the value of hmin. By varying h from hmax to this newl y calculated hmin, one obtains the set of values for k that are achievable for each given design. Similar to the previous design, the variations of Design 2b are explored for different sets of the design parameter r for a given hmax and kmin. These results will be compared with the designs of other fixels to further delineate their effectiveness. 3.3 Design 3 The fixel for this design consists of a beam fixed at both ends as shown in Figure 3 8 The workpiece would be held via contact with one end of an in terface element (similar to Design 2a) where the interface element connects at some point along the beam. Figure 3 9 is an illustration for the implementation of Design 3 using four fixels to hold the workpiece. Each contact point would be at a distance x from a fixed end and the contact force FC acting along the interface element will be exerted on the beam at this point. By moving the beam along the direction of its axis and by allowing the end of the interface element to roll on the beam, the contact for ce point of application can be changed without inducing any motion of the workpiece. The new value of x obtained by this change would correspond to the value of k required at this instance. This value of x can be calculated using the relationship between k and x given by Equation 3 14 where L is a fixed length and is a constant defined as 334 EIEbh 3 3 3L k xLx (3 1 4 )

PAGE 35

35 Thus, i n this design, the stiffness k can be controlled by varying the point of application of the contact force ( FC), i.e., by adjusting the mechanical variable x The distance of the contact point can be varied from a minimum value of xmin to a maximum value of xmax where xmax will be half the fixed length L of the beam (due to the symmetry of the fixel system). The corresponding minimum and maximum values of k are given by the Equations 3 15 and 3 -16. 3 min 3 max2 Ebh k x (3 1 5 ) 33 3 max 33 33 min minmin min4 EbhL L k xLxxLx (3 1 6 ) Introducing r as the ratio of xmin to xmax and substituting r into Equation 3 16 results in the E quation 3 17 below. 3 max 3 33 max2 2 Ebh k rrx (3 17) Here, xmax and r are the two unique parameters of Design 3. Also, similar to the previous designs, fixed xmax and r values will determine a unique value of xmin. A gain, similar to the previous designs, for a fixed xmax and for different sets of the design parameter r different variations of Design 3 are obtained. Each design variation results in a set of k values that can be achieved from varying its mechanical var iable ( x ) from xmi n to xmax.

PAGE 36

3 6 Figure 3 1. Design 1 S imple monolithic compliant four -bar mechanism consisting of four compliant joints Figure 3 2 Example of a Compliant M echanism a Crimping Mechanism consisting of three flexible members ( courtesy o f Compliant Mechanisms by Larry Howell)

PAGE 37

37 Figure 3 3. Pseudo Rigid Body Model for the Four bar monolithic fixture

PAGE 38

38 A B C D Figure 3 4. The various configurations of D esign 1 during stiffness calculations a) Initial Configuration b) Symmetric Equilibrium with Base Length L1 c) Displacement Due to Contact Force with Base Length L1 d) Change in Displacement Due to a Change in Contact Force with Base Length L1 L 1 0' 2 L 1 i0' 22 iFC L 1 i+1i22 i+1ii+1F=F+ CCC

PAGE 39

39 Design 2a) Design 2b) Figure 3 5 Design 2 Principles Involved Figure 3 6 Implementation of Design 2a with Four Fixels

PAGE 40

40 Figure 3 7 Implementation of Design 2b with Four Fixels Figure 3 8 Design 3 P rinciple I nvolved

PAGE 41

41 Figure 3 9 Implementat ion of Design 3 with F our F ixels

PAGE 42

42 CHAPTER 4 COMPARISON OF THE DESIGNS After establishing the principles of the designs and the relationships between fixel stiffness and the mechanical parameters of the designs, the fixel stiffness at the fixel -workpiece co ntact are plotted for Design 1 This is followed by a comparison of the stiffness characteristics of the similar compliant beam designs 2a, 2b and 3. 4.1 Design 1 For the Design 1 proposed in section 3.1, results were generated for the fixel design with co nstant parameters L2= L4=45mm, L3=30mm and all torsional springs having equal spring constants. The link lengths were chosen to exhibit the same relative footprint as Designs 2 and 3. T he fixel stiffness as seen at the point of fixel -workpiece contact for t he four -bar type fixel is plotted as a function of the two mechanical variables 12(,) L by following the procedure mentioned in the section 3 1 The kinematic range of values for L1 is chosen from 45mm to 55 mm and the range of values f or 2 for these link lengths was obtained to be from 40 degrees to 60 degrees. This plot for fixel stiffness with respect to L1 and 2 is shown in Figure 4 1. In the Figure 4 1, a dynamic symmetry is observed for the mechanical variable 2 for each value of L1. This symmetry can be attributed to the design constraints of L2= L4, joint flexures of equal stiffness and to the method of calculati ng the range of 2 from one toggle position to another Beca use of the dynamic symmetry, it is observed that the tren d in the fixels stiffness is independent of the length of the base link. This is apparent in the Figure 4 1 by observing the different contours in 2 for different L1. H owever, the magnitude s of the base link length impact the actual instantaneous stiffne ss values and thus the dynamic range of stiffness achievable for a given range of the base angles For example, an L1=45mm fixel configuration exhibits higher range of stiffness values achievable as compared to L1=5 5 mm fixel configurations.

PAGE 43

43 It is also obs erved that for the L1=45mm fixel configuration, a local minimum about its unloaded equilibrium configuration ( 2 50 deg) exists and therefore deviations from this configuration will take less effort While for L1=5 5 mm fixel configuration at its unloaded equilibrium configuration, the fixel exhibits a peak stiffness value and will be resistant to changes from this configuration It is also observed that the range of stiffness values achievable is lesser for smaller base length values Due to the presence of these variations in the trends it is observed that Design 1 will provide for greater control diversity in tuning the fixture dynamics via actuating the base length ( L1) and input angle ( 2) either individually or together. 4.2 Comparison of Designs 2a, 2b and 3 Arbitrarily considering values for each fixels fixed and unique design parameters, would not yield a discernable comparison of the different types of fixel designs. Thus, to compa re Designs 2a, 2b and 3, a common range of values of k achievable for all three design s is determined. It is known from the previous section that each of the three beam type designs have a particular kmin value for given r and fixed design parameters. Henc e, for more equitable comparison between the Designs 2a, 2b and 3, this minimum value of k for each of the designs with same specified value of r were made equal. Equations 4 1 to 4 3, shown below, are obtained by equating kmin and r values of each of the designs. 33 33 3 max min 333 min max design 2b design 2a design 32 44 Ebrh Ebhr Ebh kLLx (4 1 ) 33 33 3 max min 333 min max2 44 rh k hr h EbLLx (4 2 ) minminmin maxmaxmaxLhx r Lhx (4 3 )

PAGE 44

44 By restricting that all designs have the same minimal fixel stiffness ( kmin), the value of kmax at r equals 1 is also found to be the same for all three designs. Note, when r equals 1, the range of motion of the mechanical variable is restricted to be zero, thus resulting in the loss of the variable stiffness feature of the fixel With this common window of constraints on the stiffness values f ormed by the kmin restriction, the values of kmax, (equal to kmax kmin) and the average slope of k ( by change in the varying parameter of the design) for the intermediate values of r the different beam type fixel designs can be compared more direc tly. For Equation 4 1 to yield a viable comparison, it was assumed that the breadths ( b ) of the beams of each design are equal and that the material of the beam is also the same. Furthermore, the constant beam width h of Design 2a was made to be equal to hmin of Design 2b, and the constant beam width h of Design 3 was made equal to r times hmin of Design 2b. Similarly, the constant length L of Design 2b was made equal to Lm in of Design 2a and xmax of Design 3 was made to be twice Lm in. This leads to the dif ferent beam type fixel designs having the same lower bound on the slenderness ratio. The slenderness ratio ( s ) of a beam is defined as the ratio of the minimum length of the beam to its maximum width. For the different beam type fixel designs, the followin g relationships shown in Equations 4 4 and 4 5 yield the lower bound on s min maxL s h for Designs 2a and 2b (4 4 ) max max2 x s h for Design 3 (4 5 ) where, from the assumptions of Euler Bernoulli beam theory, s shou ld be greater than 10. To determine Lmin, the size of the workpiece has to be considered. Assuming that the workpiece has a size of 10mmx10 mm x10 mm, a safe estimate for Lmin is given by Equation 4 6.

PAGE 45

45 min10 mm L (4 6 ) From specified values of Lmin, r and s the parameters of comparative fixel designs and the corresponding k values can be obtained by the following procedure: 4.2.1 Design 2a Equation 4 6 gives the value of Lmin Using the specified value of s the value of hmax can be obtained from Equation 4 4. Then using Equation 4 3 and the specified value for r hmin can be calculated to determine the constant beam width h as shown by Equation 47. minhh (4 7 ) Using Equation 4 3 with the same value of r, the value of Lmax can be obtained from the known value of Lmin. For each fixel design variation defined by specified values of s and r a range of values for L varying from Lmin to Lmax is obtained which correspond to a range of values of k 4.2.2 Design 2b Equation 4 6 gives the value of Lmin which would be equal to the constant beam length L of this design as shown by Equation 4 8 minLL (4 8 ) Values of hmax and hmin for specified values of s and r are the same as that obtained above in Design 2a A range of values for h varying from hmax to hmin is thus obtained which correspond to a range of values of k

PAGE 46

46 4.2.3 Design 3 As previously explained, xmax is twice the value of the Lmin obtained from Equation 46 which can be expressed as Equation 4 9 maxmin2 xL (4 9 ) Also, the constant width of the beam would be r times hmin. Using the known value of hmin from Design 2a calculations, the fixed width can be calculated using Equation 4 10 shown below minhrh (4 1 0 ) The value of xmin for specified values of r can then be determined by substituting the value of xmax into Equation 4 3. For specified values of s and r a range of values of x varying from xmax to xmin is obtained which correspond to a range of values of k 4.2.4 Plots for Design 2a, 2b and 3 For each of the above beam type fixel designs, kmin, kmax, and average slope of k were plotted as a function of r and s (r varies from 0.1 to 1 and s varies from 10 to 30). Figure s 4 2 through 4 9 are the normali zed plots of the stiffness quantities ( kmin, kmax, and average slope of k ) with respect to ( Eb ). These results provide the trends for each design independent of the fixels material properties. Figure 4 2 demonstrates that all the three designs have the same instantaneous kmin values, shown as a function of r and s Figures 4 3 and 4 -4 demonstrate the variation of kmax for various combinations of [ r s ]. It is found that all the designs have the same maximum possible kmax value (which occurs at r= 1 and s =10). However, the variation of kmax as a function of r and s is different for Design 3. Figures 4 5 and 4 6 represent the variation of (difference between kmax and kmin) for the three beam type designs. Figures 4 7 through 4 9 are

PAGE 47

47 the plots for the average slopes of k where the average slopes are calculated as the ratio of over the range of t he change in the mechanical variable cor responding to each of the designs. From these plots, it is observed that the range of k s achievable reaches a maximum at around 0.8 r for all slenderness ratios. Also, a smaller s (shorter or thicker beams) yields a greater range in k For Design 2b, a smaller value for s results in a larger hmax value for given Lmin which has the effect of increasing the range of the mechanical variable h and thus increasing the range of k This increase in hmax effects the h of Design 2a which in tu rn has a similar amplification affect on this designs resulting range of k The are also found to be higher for Designs 2a and 2b than that of Design 3 (by a factor of 7). For an overall measure of k variations between kmin and kmax, Design 2b trends w ere found to be a factor of 10 higher in average change in k as compared to Designs 2a and 3. This indicates that Design 2b has greater mechanical adaptability for a given range in variable (defined by r ).

PAGE 48

48 45 50 55 20 40 60 80 0.8 0.9 1 1.1 1.2 1.3 x 10-3 2(degrees) dFy /dY (N/mm) versus [2(degrees),L1(mm)] L1 (mm) dFy /dY (N/mm) Figure 4 1. Fixel Stiffness (L2=L4=45mm, L3=3 0mm, all ks are equal )

PAGE 49

49 0.5 1 10 15 20 25 30 0 1 2 3 x 10-4 r from 0 to 1 kmin for designs 2a), 2b) and 3 versus [r,s] s from 10 to 30 Minimum k / Eb Figure 4 2 (kmin / Eb ) for Designs 2a, 2b and 3 versus [ r s ]

PAGE 50

50 0.5 1 10 15 20 25 30 0 1 2 3 x 10-4 r from 0 to 1 kmax for designs 2a) and 2b) versus [r,s] s from 10 to 30 Maximum k / Eb Figure 4 3 (kmax / Eb ) for Designs 2a and 2b versus [ r s ]

PAGE 51

51 0.5 1 10 15 20 25 30 0 1 2 3 x 10-4 r from 0 to 1 kmax for design 3 versus [r,s] s from 10 to 30 Maximum k / Eb Figure 4 4. (kmax / Eb ) for Design 3 versus [ r s ]

PAGE 52

52 0 0.2 0.4 0.6 0.8 10 15 20 25 30 0 1 2 3 4 5 6 7 x 10-5 r from 0 to 1 delta k for designs 2a) and 2b) versus [r,s] s from 10 to 30 Difference in max and min k / Eb Figure 4 5 ( / Eb ) x 105 for Designs 2a and 2b versus [ r s ]

PAGE 53

53 0 0.2 0.4 0.6 0.8 10 15 20 25 30 0 0.2 0.4 0.6 0.8 1 x 10-5 r from 0 to 1 delta k for design 3 versus [r,s] s from 10 to 30 Difference in max and min k / Eb Figure 4 6 k / Eb ) x 105 for Design 3 versus [ r s ]

PAGE 54

54 0.2 0.4 0.6 0.8 1 10 15 20 25 30 0 1 2 3 4 5 6 x 10-4 r from 0 to 1 Average slope of k for designs 2a) versus [r,s] s from 10 to 30 1/Eb (dk/dL) Figure 4 7 (Average Slope of k / Eb ) x 104 for Design 2a versus [ r s ]

PAGE 55

55 0.2 0.4 0.6 0.8 1 10 15 20 25 30 0 1 2 3 4 5 6 7 x 10-3 r from 0 to 1 Average slope of k for designs 2b) versus [r,s] s from 10 to 30 1/Eb (dk/dh) Figure 4 8 (Average Slope of k / Eb ) x 103 for Design 2b versus [ r s ]

PAGE 56

56 0 0.2 0.4 0.6 0.8 10 15 20 25 30 0 0.5 1 1.5 2 x 10-5 r from 0 to 1 Average slope of k for design 3 versus [r,s] s from 10 to 30 1/Eb (dk/dx) Figure 4 9. (Average Slope of k / Eb ) x 105 for Design 3 versus [ r s ]

PAGE 57

57 CHAPTER 5 ANALYSIS OF THE FIXE L DESIGNS In this chapter, a fixel design i s incorporated into a machine tool setup and the performance of the design under operation conditions subject to runout induced errors is analyzed. The fixel design and the number of compliant fixels (of that design) used in a setup will depend on the type of manufacturing process. For example, in case of a meso scale drilling process at least four fixels of Desig n 1 are required. The (X, Y) directional stiffness characteristics of fixels of Design 1 integrated in a coordinated manner is best suited for achieving the preferred fixturing capabilities and accurate drilling. In the case of a milling process two fixel s of any of the four designs are theoretically required to achieve fixturing capabilities Because each of the designs have the same principle of utilizing adjustable stiffness characteristics, the following analysis is done without loss of generality for only one of the designs for a particular meso manufacturing process The results of this analysis can be extended to other designs and processes. For the purpose of this study, Design 2a is selected for analysis and is implemented in an end milling machine tool system which has runout in the direction perpendicular to feed Another example of such a system could be a burnishing setup where a mesoscale workpiece with micro features is being burnished for improved finishing. When the tool is burnishing the surface of the workpiece which is close to the feature, the runout in the tool (perpendicular to the feed) inadvertently affect s the dimensions of feature during the burnishing process. One more mesoscale manufacturing setup which can have runout in the per pendicular direction to the feed is a peripheral milling setup with runout error occurring due to variation in teeth lengths. Considering that in such a setup, one of the teeth of the cutter is longer than the other teeth, the excess length of the longer t eeth corresponds to the runout error. Such runout errors in a peripheral milling setup results in varying depth of cut and this error can be

PAGE 58

58 compensated by following the same procedure as that for the end milling setup which is explained in Section 5.1 bel ow. 5.1 Static Analysis of Design 2a The fixture workpiece setup for a n end milling operation with two fixels of Design 2a) and two rigid fixels is shown in Figure 5 1. The fixture -workpiece is fixed to the platform of the milling machine such that the rig id fixels are aligned parallel to the direction of motion of the platform. These rigid links counter the cutting force during the milling operation. Assuming for the current setup, that the tool on the milling machine has radial runout in the direction per pendicular to the cutting direction, the compliant fixels are aligned in this direction. Due to this runout, the thickness of the slot being machined into the workpiece is larger than the required value. The perpendicular runout of the tool in such a syste m result s in a force (extraneous to cutting force) acting on the workpiece which then acts along the interface element of the compliant fixel Since the interface element is connected to the variable length compliant beam at its free end, it leads to the b ending of the beam which then results in a corresponding displacement of the workpiece The rigid fixels slide freely in this direction without their contact with the workpiece slipping so that it does not introduce any resistance to the compliant fixture s corrective displacement of the workpiece The objective of the current approach is to improve the accuracy of dimensioning in spite of errors resulting from runout by constantly controlling the beam deflection such that it is always equal to the runout v alue. This eventually results in no relative motion between the tool and the workpiece from its theoretical nominal cutting trajectory thus mitigating the adverse affects on the dimensional accuracy of the finished workpiece. The value of beam deflection x will be a function of the contact force Fc (created because of the tool runout) and the stiffness of the beam k which in turn depends on the length of the beam L as shown in Equation 5 1.

PAGE 59

59 3 34cEbh Fx L (5 1) If the runout value (and the co rresponding non -proportional contact force) changes during the process, the length of the beam can be adjusted so that the deflection of the beam is now equivalent to this new runout value thus maintaining the desired thickness of the slot T he runout valu es can be dynamically measured using Non -contact Precision Capacitance Sensors To prove th is concept of achieving beam deflection equivalent to the runout, a ProE model of a single fixel and the workpiece along with appropriate constraints is modeled as s hown in Figure 5 2 The mesoscale size of the workpiece is chosen as 5 mm x 5 mm x 5 mm. A set of values of the required runout s are initially chosen (similar to values found in literature [12] ). To calculate the range of values of the contact forces a va lue of 8 mm is chosen for L and the expected runout is chosen (from known range of values) as 0.06 mm. Using Equation 51, the value of contact force is calculated to be in the order of 25 mN. Now, the value of the contact force is varied in the range of 2 0 mN to 30 mN and the values of L corresponding to a range of runouts are calculated using Equation 5 1. The contact force is not necessarily proportional to the runout and keeping this in mind, different combinations of contact force and runout values are chosen to calculate the L values. To validate these estimated lengths using the model, the length of the compliant beam in the model is set to one of the calculated values of L and the corresponding contact force is applied to the workpiece. A force analy sis is done on this model and the deflection of the beam is obtained as shown in Figure 5 3. The value of the beam deflection is then compared to the value obtained from the formula. Since, the formula is obtained after making certain assumptions, these va lues vary by a small amount. Similar values of L were obtained for different sets of runout and contact forces and are tabulated in Table 5 1 which shows the possibility of achieving different runout compensation by varying the length of

PAGE 60

60 the compliant beam This is again shown in the form of a graph in Figure 5 -4. These values of length and corresponding stiffness values may or may not be the same for the analysis under the influence of a harmonic force. 5.2 Analysis of Design 2a U nder Harmonic Force The next step is t o analyze the design under the operation condi tions of a meso scale milling setup. Assuming that there exists runout in the tool of such a system, it results in a harmonic force F (as shown in Equation 5 2) acting on the workpiece which in turn acts on the compliant beam. The frequency of this harmonic force will be equal to the frequency of spindle rotation. This setup is modeled in MSC Adams by approximating the cantilever beams as springs with attached fixel mass es which in turn are in contact with the workpiece as shown in Figure 5 5 The rigid links which are aligned parallel to the cutting direction are modeled as constraints in this model. The springs are pre -stressed such that they are always compressed thus constantly applying a force on the workpiece. This system is equivalent to an undamped system under harmonic force excitation and hence, the displacement of the workpiece will now be a function of its mass and the frequency of the harmonic force in addition to the stiffness of the fixel s (e.g. cantilever beam) and the maximum amplitude of the contact force The maximum amplitude X of the particular solution of such an undamped system under harmonic force is obtained from Equation 5 2 shown below, where F0 is the maximum amplitude of the harmonic force, k is the stiffness of the springs (in parallel) in the system (cantilever fixels), m is the mass of the workpiece and is the frequency of the harmonic force [13]. 0 0 2sin() FFt F X km (5 2)

PAGE 61

61 For an undampe d system under a harmonic force, the phase of the particular solution is same as the phase of the harmonic force only when the ratio of the frequency of the harmonic force to the natural frequency n of the system lies between 0 an d 1 as shown in Equation 53 [13]. 0/1 n (5 3) But the natural frequency of the current system is a function of the spring constant of the cantilever beam and the mass of the workpiece (ignoring fixel mass) Hence, to cancel the erro rs induced by the runout, the fixels must move the workpiece in phase with the runout. Therefore, the stiffness of the beam has to be chosen such that the natural frequency of such a system is greater than the frequency of spindle rotation. Now, to study the behavior of the workpiece -fixels system under a harmonic force, the Adams model is to be analyzed at different spindle speeds ranging from 60000 rpm to 90000 rpm which is the typical range of spindle speeds for a mesoscale system. The value of the run out in the tool is assumed to be in 0.0012 mm and a nominal value of 25 mN is chosen for the maximum amplitude of the harmonic contact force It should be noted that the value of runout is chosen as 0.0012 mm since it is the least runout value usually obse rved in high precision systems (found in the literature ) and to verify the model for this low value. The value of the harmonic force is chosen to be same as in the static analysis so that there is consistency with the previous analysis Using Equation 5 2, the value of the stiffness of the compliant beam required to achieve the chosen runout value at different spindle speeds is calculated and tabulated as shown in Table 5 2. It should be noted that the resulting design values of k per fixel depend on the nu mber o f fixels and their arrangements, yielding an equivalent k value for Equation 5 -2.

PAGE 62

62 To validate the performance of the design, the frequency of the contact force in the Adams model is initially set to 60000 rpm and the value of the spring constant is set to the value from Table 5 2 corresponding to the spindle speed and the model is run for 0.5 seconds. For this analysis, a plot for the displacement of the center of mass of the workpiece with respect to time is obtained as shown in Figure 56. It is ob served from this figure that the system has an initial transient state for a very small duration (order of 0.01 seconds) following which the maximum amplitude of the displacement remains constant. This magnitude of the maximum amplitude is measured from th e plot and is found to be 0.0012 mm which is equa l to the expected runout value. To better compare the expected runout value and the displacement of the workpiece, these two quantities were plotted together as shown in Figure 5 7. It can be clearly observed from this figure that the displacement of the workpiece is not equal to the runout value for a small duration (close to 0.01 seconds for this example set of conditions ) at the beginning of the process. But after this initial transient period, the displac ement of the workpiece is equal to the expected runout value along with always being in phase with one another. Thus, the displacement of the center of mass of the workpiece is made equal to the expected runout by varying the length of the compliant beam t o achieve the required stiffness values. To analyze the workpiece -fixel system at various frequencies, the frequencies of harmonic force were varied from 60000 rpm to 90000 rpm and the Adams model is run for these frequencies. During each of the runs, the spring constant is changed to a corresponding value as obtained from Table 5 2. From the Adams models, it is observed that the systems reach steady state quickly and the values of the maximum amplitude of the displacement are equal to the

PAGE 63

63 expected runout v alues in the steady state. The displacement and runout values are also in phase with another for each of the systems. To study the error between the displacement of the center of mass of the workpiece and the expected runout value (0.0012 mm in each case) in the transient state, these parameters are plotted for each of the frequencies as shown in Figure 5 8. As can be seen in this figure, the error between these two parameters varies with the frequency of the harmonic force. Also, the time the system takes to reach a constant value is also dependant on the frequency of the force For higher values of frequencies, the system reaches steady state more quickly as compared to lower frequencies. Also, the maximum value of this error for all the frequencies is measured to be approximately 0.001 mm. Hence, higher frequencies of spindle rotation are optimal for mMT systems. While not shown in Figure 5 8 due to time scale shown for only 0.02 seconds, all cases converge to a low value of error. To estimate the physica l size of the canti lever beams required to obtain the calcu lated values of spring constants, the stiffness values from Table 5 2 are used to calculate the corresponding length s of the beam considering the breadth to be 1.25 mm and the w idth of the beam to be 0.75 mm These values have been shown in Table 5 3 and it can be observed from this table that the required lengths are comparable to the footprint of the mMT setup. It can also be observed from this table that since t he minimum and maximum stiffness va lues tabulated in Table 5 2 are nothing but the kmin and kmax values respectively of this configuration of design 2a the values for length obtained in Table 5 3 are the corresponding Lmax and Lmin values. The ratio r of these Lm in to Lm ax is calculated to be of the order of 0.8 which from the observations of Section 4.2.4 is the optimum value of r for achieving maximum range of stiffness. It is also observed from the table that the slenderness ratio s (ratio of Lmin to h ) for this configuration is greater than

PAGE 64

64 10, thus satisfying the assumption of the design. It should be noted that different combinations of breadth and width of the beam and length of the beam can be used where the critical aspect is acquiring the desired fixel settings. These stiffness v alues can also be obtained for designs 2b and 3 for the appropriate values of the parameters of the designs. To estimate the values of these parameters for design 2b, the same stiffness values as before from Table 5 2 are used to calculate the correspondin g widths of the beam considering the fixed length of the beam to be 10 mm and the breadth of the beam to be 1.25 mm. The thickness values have been shown in Table 5 4 and are comparable to the mMT setup It is observed that, similar to design 2a, the minim um and maximum values of width obtained in Table 5 4 represent hmin and hmax of the current configuration design 2b. The ratio of these minimum and maximum widths is calculated to be in the order of 0.8 which again is the optimum value of r (from Section 4 .2.4). Thus the parameters of the current configuration are the optimum values for design 2b. The slenderness ratio s (ratio of L to hmax) in this case is also greater than 10. Similarly for design 3, the values of the distance of the point of application of force from the fixed ends ( x ) required to achieve the stiffness values from Table 5 2 are calculated and shown in Table 5 5 To calculate these x values, the fixed length of the compliant beam is considered to be 5 mm, the breadth to be 0.5 mm and the w idth to be 0.25 mm. It can be observed from Table 5 5 that the values of x obtained are achievable for the particular chosen configuration. Similar to the previous designs, the ratio r (ratio of xmin to xmax) is same as the optimum value of 0.8 and the sle nderness ratio s (ratio of L to h ) is also greater than 10. Hence, by performing dynamic analysis on the fixel -workpiece system it is established that runout

PAGE 65

65 compensation is achievable using variable stiffness fixels (of designs 2a, 2b or 3) for different frequencies of the spindle rotation. 5.3 Compensation for Transient Dynamics As observed in the previous section, under the influence of harmonic force, the maximum displacement of the workpiece varies initially before reaching a constant value. This trans ient state exists for a very short duration but its effect on accuracy depends on the feed rate of the process. For small feed rates, the length of the workpiece machined in this small duration will not create excess inaccuracies. But in the case of high f eed rates, these inaccuracies cannot be negligible. If the effects of the transient dynamics are not negligible, to counter this variation it is proposed that sacrificial material be used for the initial duration of the process. The sacrificial material is to be obtained by using a longer (in case of milling) or thicker (in case of drilling) workpiece and discarding the excess length/thickness after the completion of the operation. In the determination of the appropriate fixel stiffness, the workpiece mass value will need to include the mass of the sacrificial material in the calculation of the length of the cantilever beam corresponding to a runout value.

PAGE 66

66 Table 5 1. Overview of the runout and length values for various trials obtained from the FE Analysis Runout values in mm Runout values obtained from the ProE model Length of the cantilever beam, L in mm 0.1488 0.15255 9.999 0.1354 0.13903 9.962 0.1182 0.11989 9.742 0.0957 0.09795 9.307 0.0813 0.08245 9.053 0.0645 0.06642 8.62 8 0.0492 0.05142 8.137 0.0381 0.03647 7.999 Table 5 2. Calculation of the stiffness values corresponding to different spindle speeds Spindle rotation speed in rpm Maximum amplitude of the contact force in mN Expected runout value in mm Calculated stiffness values in N/mm 60000 25 0.0012 29.645 66000 25 0.0012 33.683 72000 25 0.0012 38.10 6 78000 25 0.0012 42.913 84000 25 0.0012 48.10 5 90000 25 0.0012 53.681 Table 5 3. Calculation of the lengths of the compliant beams corresponding to the beam stiffness values for De sign 2a Calculated stiffness values in N/mm B readth of the compliant beam in mm W idth of the compliant beam in mm Length of the compliant beam in mm 29.645 1.25 0.75 9.725 33.683 1.25 0.75 9.320 38.10 6 1.25 0.75 8.945 42.913 1.25 0.75 8.597 48.10 5 1.25 0.75 8.276 53.681 1.25 0.75 7.979

PAGE 67

67 Table 5 4. Calculation of the widths of the compliant beams corresponding to the beam stiffness values for Design 2b Calculated stiffness values in N/mm B readth of the compliant beam in mm Length of the com pliant beam in mm Width of the compliant beam in mm 29.645 1.25 10.00 0.77 1 33.683 1.25 10.00 0.80 5 38.10 6 1.25 10.00 0.838 42.913 1.25 10.00 0.872 48.10 5 1.25 10.00 0.90 6 53.681 1.25 10.00 0.93 4 Table 5 5. Calculation of distance of point of appli cation of force from the fixed end of the compliant beams corresponding to the beam stiffness values for Design 3 Calculated stiffness values in N/mm Length of the compliant beam in mm B readth of the compliant beam in mm Width of the compliant beam in mm Value of x in mm 29.645 5 .00 0.5 0. 25 1.38 7 33.683 5 .00 0.5 0. 25 1.31 9 38.10 6 5 .00 0.5 0. 25 1.25 7 42.913 5 .00 0.5 0. 25 1.200 48.10 5 5 .00 0.5 0. 25 1.149 53.681 5 .00 0.5 0. 25 1.101

PAGE 68

68 Figure 5 1. ProE model for fixel workpiece configuration for a milling operation with two compliant fixels and two rigid fixels Figure 5 2 ProE model for single fixel workpiece configuration

PAGE 69

69 Figure 5 3 FE Analysis to calculate the maximum displacement of the compliant beam Figure 5 4 Variation of length of the cantilever beam for corresponding runout values

PAGE 70

70 Figure 5 5 MSC Adams model for the two fixels -workpiece configuration Figure 5 6 Plot showing the Displacement of the center of mass of the workpiece with respect to time where the frequency of the harmonic force is 60000 rpm

PAGE 71

71 Figure 5 7 Comparison of the d isplacement of the center of mass of the workpiece and the expected runout values for a frequency of 60000 rpm

PAGE 72

72 1.50E 03 1.00E 03 5.00E 04 0.00E+00 5.00E 04 1.00E 03 1.50E 03 0.00E+00 2.00E 03 4.00E 03 6.00E 03 8.00E 03 1.00E 02 Error in mm Time in seconds Error between displacement of the center of mass of workpiece and the expected runout values in the transient stage 90,000 rpm 84,000 rpm 78,000 rpm 72,000 rpm 66,000 rpm 60,000 rpm Figure 5 8 Plot showing the error between the displacement of the center of mass of the workpiece and the expected runout values in the transient stage

PAGE 73

73 CHAPTER 6 DISCUSSION AND CONCLUSIONS 6.1 Conclusions The motivation for investigating the fixel designs presented in this thesis is driven by the con jecture that through the provision of mechanically adaptable stiffness characteristics in passive/ active fixturing one can potentially improve the dynamic response of the workpiece during manufacturing of micron features. This study has provided the theore tical modeling of four such fixel designs and a comparison of what stiffness trends one may extract from their design. It is observed that four bar mechanism type fixel offers greater versatility in that it exhibits two mechanical variables for making adju stments. But t his adds to the complexity of the design by requiring twice as many actuators for controlling these variables. However, the results indicate that by controlling one of the parameters, the fixel can still be tuned to control the dynamic behavi or. In addition, the trends observed in the stiffness as a function of the mechanical variables exhibit local minimums For the beam type fixels, Designs 2a and 3 offer continuous adaptability fixel stiffness while the Design 2b uses incremental adjustment s in changing its stiffness. For these three designs, it was found that similar trends occurred in terms of k versus available range of input variable. Additionally, the lower bound on the slenderness ratio contributed to an amplification of the rate of ch ange in k with respect to change in input variable, as well as on the range of k Design 2a and 2b yield the greatest of all values for range of k while Design 2b yields the greatest value for the average slope of k But since design 2b is potentially hamp ered by the discrete nature of its design its practical application will be difficult. Hence, considering both

PAGE 74

74 the stiffness characteristics and ease of application, Design 2a is the best among the fixel designs. To analyze the performance of the proposed designs, one of the designs (Design 2a) is analyzed by implementing it in an end milling mesoscale machine tool system where the tool has runout along the direction of the fixels. The fixel -workpiece system was initially modeled in ProE and a static analy sis was performed, from which it was established that runout compensation can be achieved by adjusting the length of the compliant beam to obtain beam deflection equal to the runout error. To simulate a mesoscale tool system, a MSC Adams model was develope d and analyzed under the influence of a harmonic force and other typical mesoscale system operation parameters. It was observed that runout compensation can be achieved in this case as well, but the length of the compliant beam was now a function of the fr equency of the spindle rotation and the mass of the workpiece in addition to the stiffness of the compliant beam. To conclude, four different fi xel designs have been presented that exhibit the mechanical adaptability of their stiffness characteristics and analysis has been performed on one of the designs to prove the implementation of the design concept for fixturing for mesoscale manufacturing A provisional patent for the designs proposed in this thesis has been obtained. 6 2 Directions for Fu ture Resear ch For further research these designs need to be tested experimentally to further quantify their ability to control the dynamic response of the tool -workpiece -fixture interface In addition, implementation issues in terms of actuators and sensors and the i r real time control and effects of friction in the fixture -workpiece platform (support surface) need further study It should be noted that the research in this thesis focuses on application of the fixel designs within the fixturing system by using two or four fixels. The determination of the specific number and optimal

PAGE 75

75 l ocations of the fixels required, which is work piece and process dependent, has been left for future work. Also, the determination of the optimal contact force has been left for future work. It has been observed from the analysis of the fixel designs that the workpiece -fixture system under the influence a harmonic force experiences a transient period before reaching the steady state. The effect of this transient nature of the displacement of the workpiece on the accuracy of the process will be dependent on the feed rate values. A methodology to reduce the duration of the transient state needs to be developed which will be done during further research. Also, the sensitivity of the results of th e analysis to variations in the mass of the workpiece has not been studied. This will be studied in future work to quantify its significance on part accuracy for such scenarios in which material is being removed.

PAGE 76

76 REFERENCES [1] S. W. Lee, R. Mayor and J. Ni, Dynamic analysis of a mesoscale machine tool, Journal of Manufacturing Science and Engineering, 128 (1), 194 203, 2006. [2] A.H. Honegger, S.G. Kapoor and R.E. DeVor, A hybrid methodology for kinematic calibration of micro/meso -scale machine tools (mMTs), ASME Journal of Manufacturing Science and Engineering, 128 (2 ), 13 18, 2006. [3] H. L i X L ai C L i Z L in J M iao J N i, Development of meso-scale milling machine tool and its performance analysis in the Chinese Journal of Mechanical Engineering, 42 (11) 162 167 2006. [4] J. Bronson, G. Wiens, J. Palmer and J.R. Meyer, Feasibility study of mechanically adaptive microfabrication process tooling and fixturing, 1st International Conference on Micr omanufacuring, ICOMM 2006, University of Illinois at Urbana Champaign IL 2006. [5] G.J. Wiens, R.M. Wagner, A. Valdes and J.R. Bronson, "Mechanicallyadaptive fixture element for directional control of contact interface forces," International Conference on Micro Manufacturing 2007, Clemson University, SC 2007. [6] M.L. Culpepper and S. Kim, Sixaxis reconfigurable nanomanipulator systems for positioning in nano -scale manufacturing and research, International Symposium on Nano Manufactu ring, ISNM 2004, Daejeon, Korea 2004. [7] U.A. Tol, G.J. Wiens and J. Schueller Mechanical adaptability of active fixturing system, submitted for refereed journal publication, 2009 [8] B.S. El -Khasawneh and P.M. Ferreira, Computation of stiffness and stiffness bounds for parallel link manipulators, International Journal of Machine Tools & Manufacture, 39 (1), 321 341, 1999. [9] G.J. Wiens and D.S. Hardage, Structural dynamics and system identification of parallel kinematic mechanisms, in Proceedings of the Inter national Design Engineering Technical Conferences, DETC2006, 2006, pp. 10. [10] W.F. Riley, L.D. Sturges and D.H. Morris, Mechanics o f m aterials, John Wiley & Sons New York, NY pp. 475 484, 1999. [11] L.L. Howell, Compliant m echanisms, John Wiley & Son s New York, NY 2001. [12] D.L.Martin and R.O.Wilcox, Spindle runout effects on postional accuracy, in the University of Wisconsin Symposium on Small Hole Techonolgy, Palm Springs, California 1990. [13] S.S.Rao, Mechanical m ibrat ions, Pearson Eduactio n, Inc. New Jersey NJ, pp 221 225, 2004.

PAGE 77

77 BIOGRAPHICAL SKETCH Koustubh Rao was born in India, in 1984 He received his b achelors degree in m echanical e ngineering from the Indian Institute of Technology Guwahati Guwahati India in May 2006. Mr. Rao wi ll have received his Master of Science degree in mechanical engineering from University of Florida in August 2009. At University of Florida, he worked in the Space Automation and Manufacturing Mechanisms Lab oratory in the Department of Mechanical and Aeros pace Engineering under the guidance of Dr. G. Wiens