MicroSystems Engineering Team Louisiana State University 1/78 Thesis Defense – April 3, 2001 Design and Fabrication of a Thermomechanical Microactuator Proyag Datta Department of Mechanical Engineering Louisiana State University April 3, 2001 Thesis Defense
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
MicroSystems Engineering Team
Louisiana State University
1/78
Thesis Defense – April 3, 2001
Design and Fabrication of a
Thermomechanical Microactuator
Proyag Datta Department of Mechanical Engineering
Louisiana State University
April 3, 2001
Thesis Defense
MicroSystems Engineering Team
Louisiana State University
2/78
Thesis Defense – April 3, 2001
Acknowlegdements
Project was funded by AFOSR
MicroSystems Engineering Team
Louisiana State University
3/78
Thesis Defense – April 3, 2001
PRESENTATION OUTLINE
• Introduction • Design and Modeling • Fabrication Process Developement • Conclusion
MicroSystems Engineering Team
Louisiana State University
4/78
Thesis Defense – April 3, 2001
INTRODUCTION Trapped Vortex(TV) Combustors
• Continuous interest towards improving the performance of aircraft propulsion systems
• Improved fuel efficiency, better specific energy release, extended life, extended lean flammability limit and reduced emission of environmental pollutants
• A Trapped Vortex combustor is a means to implement a stabilized combustion process in an engine
MicroSystems Engineering Team
Louisiana State University
5/78
Thesis Defense – April 3, 2001
INTRODUCTION Concept of ‘Breathing Wall’
• TV-combustors experience thermo-acoustic instabilities and ‘hot spots’, which lead to lowered efficiency in the combustor
• Hot spots can be controlled by injecting cooler air through dilution holes on the combustor walls
• Distributed air injection would – control local stoichiometry – lead to uniform temperature distribution – minimize wall temperature – minimize NOx formation
MicroSystems Engineering Team
Louisiana State University
6/78
Thesis Defense – April 3, 2001
INTRODUCTION Schematic of TV Combustor
MicroSystems Engineering Team
Louisiana State University
7/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Microvalves
• Properties of an ideal valve – Low leakage – Low power consumption – Low dead volume – Large differential pressure capability – Insensitivity to particulate contamination – Low response time – Potential for linear operation – Ability to handle fluids of any density/viscosity/chemistry
• Valves are designed for specific conditions of operation
MicroSystems Engineering Team
Louisiana State University
8/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Microvalves
• Valves are classified as ‘passive’ or ‘active’ • Passive Valves
– No external power or control – Usually one-way or check-valves
• Active Valves – Powered actuation mechanism – Driving Mechanisms
• Survival at elevated temperatures • Actuation distance (~500 µm) • Force Required
• Compactness of design • Integrable into combustor walls • Frequency response (>100Hz) • Rugged design for operation in harsh
environment
MicroSystems Engineering Team
Louisiana State University
11/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Design Considerations
• Most methods of active actuation fail due to high temperature (e.g. piezoelectric, magnetic)
• Passive actuation chosen • Temperature gradient as energy source to drive
the actuator • Thermal expansion as method of actuation • Array structure chosen
– Resistant to particulates – Tailored to meet force and deflection requirements
MicroSystems Engineering Team
Louisiana State University
12/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Recurve Architecture
• Direct thermal expansion produces insufficient deflection
• Deflection of a single bimetallic element is insufficient for the amount of deflection reqd.
• Bimetallic elements cannot be stacked as tip rotation nullifies deflection
• Recurve architecture suggested by Ervin and Brei (1998) chosen
MicroSystems Engineering Team
Louisiana State University
13/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ‘Recurve’ Schematic
MicroSystems Engineering Team
Louisiana State University
14/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Recurve Architecture
• Basic building block - composite beam made of two materials with different coefficients of thermal expansion
• Produces a parallel displacement of the endpoint relative to the base
• Can be combined into arrays to obtain greater net deflections or forces
• By reversing positions of high and low CTE materials, pull type actuators can be fabricated.
MicroSystems Engineering Team
Louisiana State University
15/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Behavior of a Recurve Element
3-D solid model of a recurve element shown in undeformed(Left) and deformed(Right) state
MicroSystems Engineering Team
Louisiana State University
16/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Recurve Array
MicroSystems Engineering Team
Louisiana State University
17/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Quasi-static Modeling
• Strain energy based analytical derivation using Castigliano’s second theorem
• Equations derived for – Displacement – Force
MicroSystems Engineering Team
Louisiana State University
18/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Quasi-static Modeling
• Equation for Recurve
• Moment in bimetallic strip
DLM
DL
mF
nez
412.
23
+=Δ
( )IE
hTMeΔ−
+=
).(.1224.2 21 αα
ϕ
MicroSystems Engineering Team
Louisiana State University
19/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Quasi-static Modeling
Force vs Height
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0 100 200 300 400 500 600
Height of Recurve(Micrometers)
Blo
ckin
g F
orc
e(N
)
Deflection vs Height
0
10
20
30
40
50
60
0 100 200 300 400 500 600
Height of Recurve(Micrometers)
Def
lect
ion
(Mic
rom
eter
s)
MicroSystems Engineering Team
Louisiana State University
20/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Quasi-static Modeling
Deflection vs Thickness
050100150200250
0 100 200 300Thickness (Micrometers)
Def
lect
ion
(M
icro
met
ers)
Force vs Thickness
0.00
0.20
0.400.60
0.80
1.00
0 100 200 300Thickness (Micrometers)
Blo
ckin
g F
orc
e(N
)
MicroSystems Engineering Team
Louisiana State University
21/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Quasi-static Modeling
Deflection vs Length
0
20
40
60
80
100
0 5000 10000 15000Length (Micrometers)
Def
lect
ion
(M
icro
met
ers)
Force vs Length
0.000.501.00
1.502.002.50
0 5000 10000 15000Length (Micrometers)
Blo
ckin
g F
orc
e(N
)
MicroSystems Engineering Team
Louisiana State University
22/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ANSYS Modeling
• 3-D ANSYS model created • Steady state analysis carried out • Alternate configurations simulated • Coupled field analysis carried out • Sequential Method of analysis used
MicroSystems Engineering Team
Louisiana State University
23/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ANSYS Modeling
• Meshed with Solid87 3-D, 10-Node Tetrahedral elements for thermal analysis
• Uniform steady state temperature attained • Nodal results read in for structural
analysis • Elements changed to Solid92, a 10-node
tetrahedral structural solid
MicroSystems Engineering Team
Louisiana State University
24/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ANSYS Modeling – Model I
Meshed ANSYS Model - I
MicroSystems Engineering Team
Louisiana State University
25/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ANSYS Modeling – Model I
Deflection of Recurve Model- I
MicroSystems Engineering Team
Louisiana State University
26/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Comparison of Results
Comparison of Deflections Predicted by Analytical Model and ANSYS (Model I)
0
50
100
150
200
250
0 100 200 300 400 500Temperature (C)
Def
lect
ion
(mic
rom
eter
s)
AnalyticANSYSError
MicroSystems Engineering Team
Louisiana State University
27/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ANSYS Modeling – Model II
Meshed ANSYS Model - II
MicroSystems Engineering Team
Louisiana State University
28/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ANSYS Modeling – Model II
Deflection of Recurve Model- II
MicroSystems Engineering Team
Louisiana State University
29/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Comparison of Results
Comparison of Deflections Predicted by Analytical Model and ANSYS(Model II)
0
100
200
300
400
500
0 100 200 300 400 500Temperature (C)
Def
lect
ion
(mic
rom
eter
s)
AnalyticANSYSError
MicroSystems Engineering Team
Louisiana State University
30/78
Thesis Defense – April 3, 2001
DESIGN and MODELING ANSYS Modeling-Max Stress
Max Stress predicted by analytical model =1.482E-5 kgf/sq µm
MicroSystems Engineering Team
Louisiana State University
31/78
Thesis Defense – April 3, 2001
DESIGN and MODELING Dynamic Modeling
• Assess the order of dynamic response of the passive actuator
• Graphical system-modeling tool • Uniform treatment of various energy domains • Lumped parameter pseudo bondgraph model of
heat transfer in the recurve elements developed • Coupled with the mechanical system bond graph
Ohmic corrected polarization curve for nickel-iron bath
MicroSystems Engineering Team
Louisiana State University
46/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Ni-Fe Electroplating - Issues
• Stress generation – cracks, brittleness • Passivation – required hard is hard to
obtain, plating stops/slows down for no apparent reason
• Composition varies from top to bottom • Rusting
MicroSystems Engineering Team
Louisiana State University
47/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Mask Fabrication – Optical Mask
• Autocad drawings – Multilayered
• Optical Mask – Autocad file conversion – 5x5 inch commercial wafer with Chrome &
Positive resist – Exposure on MANN 3600 pattern generator – Development – Chrome etch – Resist removal
MicroSystems Engineering Team
Louisiana State University
48/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Mask Fabrication – Autocad Drawings
MicroSystems Engineering Team
Louisiana State University
49/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Mask Fabrication – X-Ray Mask
• X-Ray Mask Fabrication – Glass ring cut by waterjet – DFP3 graphite cleaned and stuck to glass ring using UV-
cured glue – 50 A of Chrome and 300 A of gold E-beam deposited – SU-8 spun on wafer and baked – Wafer exposed using optical mask
Glass Ring
Evaporated Chrome & Gold
SU-8
DFP-3 Graphite
Glass UV Chromium Mask
UV Exposure
MicroSystems Engineering Team
Louisiana State University
50/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Mask Fabrication
– Post-bake and developed – Gold and chrome etched from around alignment marks – Plasma ashing to clean wafer – 20 µm of gold electrodeposited in SU-8 mold – Mask mounted on standard NIST ring – Process was used to manufacture two X-Ray masks
Gold and Chrome etched from around alignment mark
Gold Plated into pattern
Alignment Mark
Developed Pattern
MicroSystems Engineering Team
Louisiana State University
51/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Mask on Glass Ring
Gold on Graphite X-Ray mask mounted on glass ring
MicroSystems Engineering Team
Louisiana State University
52/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Close up on mask
SU-8 structures with gold plated around them
MicroSystems Engineering Team
Louisiana State University
53/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Alignment Marks on Mask
Complementary alignment marks on mask
MicroSystems Engineering Team
Louisiana State University
54/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Mask Fabrication - Issues
• Glass surface should be clean & blemish-free • Alignment marks need not be complementary –
two crosshairs work better • Distance of alignment marks from structures is
critical • SU-8 layer sinks into graphite, depending on
graphite density • SU-8 removal still a problem
MicroSystems Engineering Team
Louisiana State University
55/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Substrate Preparation
• 4 inch Titanium plate – Clean with HF for 1 min – Rinse in DI
• Oxidation – Sodium Hydroxide and Hydrogen Peroxide – 65°C for 20 min
• Copper Plating – Copper Sulphate based bath – 20mA/sqcm for 30 min
• Hand polished to improve surface
Titanium
Titanium
Titanium Oxide
Copper
MicroSystems Engineering Team
Louisiana State University
56/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Photolithography
• Spin coat photoresist – SJR-5740 positive photoresist – 2000 rpm for 30 sec to give 10 µm thick coat
– Bake at 95 °C for 8 min • Exposure
– G-line UV-exposure station at CAMD
– 400 mJ/sqcm – Only alignment marks
exposed • Development
– Microposit 354 developer for 8-12 min
Copper
Titanium Oxide
Photoresist
UV Exposed Photoresist
MicroSystems Engineering Team
Louisiana State University
57/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Photolithography
• Nickel plating – Activation using C-12 – Sulphamate Bath – 20 mA/sqcm for 20 min
• Strip photoresist – Acetone
• Oxidation – Better visibility & adhesion
Copper
Nickel Alignment Marks
Copper
Copper Oxide
MicroSystems Engineering Team
Louisiana State University
58/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Visible Alignment Marks
Wafer after oxidation – alignment marks visible
MicroSystems Engineering Team
Louisiana State University
59/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Photolithography - Issues
• Contact printing process – optical mask has to be cleaned regularly
• Perfect contact essential for good exposures • Optical mask should have as much clear field as
possible • Sacrificial electrode essential for controlling
plating into small areas
MicroSystems Engineering Team
Louisiana State University
60/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS LIGA-Process Overview
MicroSystems Engineering Team
Louisiana State University
61/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS LIGA
• Bond PMMA – 500 µm thick stock PMMA sheet – MMA based glue – 20 psi bonding pressure
• Alignment – X,Y displacement and rotation adjustments
• Exposure – X-ray exposure on CAMD beamline XRLM3
Copper
PMMA
MicroSystems Engineering Team
Louisiana State University
62/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS LIGA
X, Y - Displacement Setscrews
Alignment Mark on Mask
Alignment Mark on Substrate
Glass
Rotational Displacement Setscrew
Optical Microscope
Schematic of Alignment Process Alignment Jig
MicroSystems Engineering Team
Louisiana State University
63/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS LIGA
• Development – GG developer – 20 min in Developer, 40 min in Rinse – 1 cycle for every 100 µm of PMMA – Rinse in DI
• Etch Copper Oxide – Vacuum wafer under etch solution
Exposed PMMA
Copper oxide etched to expose copper
MicroSystems Engineering Team
Louisiana State University
64/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS LIGA
• Nickel-Iron Electroplating • Polishing • Bond 500 µm thick PMMA sheet • Flycut down to 100 µm above previous layer
Exposed PMMA
Electroplated nickel-iron, polished down to level
Second layer of PMMA flycut down to 100 µm
MicroSystems Engineering Team
Louisiana State University
65/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Post 1st Electroplating
Wafer after electroplating for 1st layer
MicroSystems Engineering Team
Louisiana State University
66/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image - Post 1st Electroplating
Part of structure after Nickel-Iron electroplating and polishing
MicroSystems Engineering Team
Louisiana State University
67/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS LIGA
• Alignment and 2nd exposure • Development • Copper oxide etch • Nickel electroplating
– Nickel Sulfamate bath – Current density of 20 mA/sqcm Second Exposure of PMMA
Nickel plated into exposed PMMA mold
Copper oxide etched to expose copper
MicroSystems Engineering Team
Louisiana State University
68/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS LIGA
• Polish Nickel • Strip PMMA
– Acetone – Heat & Stir
• Etch Copper – 50% NH4OH and 50% H202
• Etch Titanium – HF
Nickel polished down to level with nickel-iron
PMMA removed using Acetone
Copper oxide and copper etched to release structures
MicroSystems Engineering Team
Louisiana State University
69/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Post 2nd Exposure
MicroSystems Engineering Team
Louisiana State University
70/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Alignment Error
MicroSystems Engineering Team
Louisiana State University
71/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Post 2nd Electroplating
MicroSystems Engineering Team
Louisiana State University
72/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Image – Post final polish
MicroSystems Engineering Team
Louisiana State University
73/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Issues
• Accurate alignment is difficult • Unpredictable X-ray exposure results
– Mask setting faulty • Electroplated Ni-Fe has poor mechanical
properties • Bond strength between Ni & Ni-Fe suspect • Adhesion on titanium is poor
MicroSystems Engineering Team
Louisiana State University
74/78
Thesis Defense – April 3, 2001
FABRICATION PROCESS Future Work
• Alignment Issues – Reduce alignment steps by fabricating alignment marks
with first PMMA layer – Use better alignment marks – Use better alignment system
• Ni-Fe plating – Additives – Varied pulse times at lower currents – Better understanding of material properties of
electroplated Ni & Ni-Fe alloy
MicroSystems Engineering Team
Louisiana State University
75/78
Thesis Defense – April 3, 2001
ACKNOWLEDGEMENTS
Dr. Michael Murphy • Committee Members:
– Dr. Elizabeth J. Podlaha – Dr. Sumanta Acharya – Dr. Wajun Wang
• CAMD Staff – Yohannes Desta – Zhong Geng Ling – Kun Lian – Jost Gottering – Harish Manohara
MicroSystems Engineering Team
Louisiana State University
76/78
Thesis Defense – April 3, 2001
ACKNOWLEDGEMENTS
• Kevin Zanca • Abhinav Bhushan • Kabseog Kim • John Fuller • Tracy Morris • Summer Dann-Johnson • Dawit Yemane • Jason Sevin