Featured Project 02

PhD Research · MEMS Design & Characterization

Low-Power SU-8 Thermal Actuator

Designed, fabricated, and experimentally validated a low-power SU-8/Cu bimorph actuator achieving 63 µm out-of-plane displacement at 36 mW. Characterized in air, it was developed for potential use in fluidic and underwater environments.

63 µm

Maximum measured displacement

36 mW

Input power

≈200 ms

Simulated response

Fabricated SU-8 thermal actuator and simulated displacement profile

I led the multiphysics modeling, mask design, microfabrication, and experimental characterization of the actuator.

01 — The Engineering Challenge

Large motion. Minimal power.

Thermal microactuators can generate useful force, but power consumption and heat accumulation often constrain compact systems. The goal was to translate controlled Joule heating into substantial out-of-plane motion using an SU-8/Cu bimorph.

Power efficiency

Generate useful displacement at milliwatt-level input power.

Out-of-plane motion

Convert in-plane Joule heating into predictable vertical bending.

Thermal operating window

Maximize travel without crossing the device's damage boundary.

Experimental Scope

Characterized in air. Fluidic and underwater operation remained a proposed future application.

02 — Coupled Multiphysics Model

Three physics. One predictive model.

A coupled COMSOL model linked electrical current, Joule heating, heat transfer, and structural deformation—allowing displacement to be predicted before fabrication.

Electrical

Current density · Electric field · Potential

Thermal

Joule heating · Conduction · Convection

Structural

Thermal expansion · Out-of-plane displacement

Diagram of a U-shaped SU-8/Cu cantilever bending out of plane from a fixed wall, labeled 'U-shaped SU-8/Cu cantilever (out-of-plane bending)'

Electrical input

Temperature field

Mechanical deflection

Boundary Conditions

Terminal and ground · Heat transfer to ambient air · Fixed anchor

Model Confidence

Mesh-independence analysis verified solution stability before parameter sweeps.

03 — Geometry & Performance Optimization

Pattern the structure. Shape the motion.

Alternative geometries were evaluated in COMSOL to understand how patterning changes stiffness, thermal–stress distribution, and directional deflection.

Simulation Design Space

Type 0

Type 1

Type 2

Type 3

Simulated Tip Response

Simulation Study

Line chart of simulated tip displacement in micrometers versus input current in milliamps for four geometry types. Type 3 produces the largest displacement and Type 0 the smallest across the 50 to 150 milliamp range.

Patterning increased bending potential—but also changed lateral motion and stress localization.

Stiffness

Slots and openings redistributed structural compliance.

Directional Control

Geometry altered x-, y-, and z-direction response.

Design Decision

The standard U-shaped configuration was carried into experimental validation.

04 — From Mask Design to Fabricated Device

Two masks. One released bimorph.

An aligned metal-and-polymer process translated the simulated geometry into a suspended SU-8/Cu actuator.

Fabrication Sequence

Sacrificial Layer

PECVD SiO₂ on Si

Heater Lithography

Pattern the U-shaped active layer

Cr/Cu Deposition

Thermal evaporation

Lift-off

Retain heater and contact pads

SU-8 Patterning

Form the passive cantilever and release holes

SiO₂

Cr/Cu

SU-8

Fabrication Evidence

Active Layer

Optical micrograph of the active heater layer, a U-shaped metal trace on the substrate, with a 100 micrometer scale bar

Passive Layer

Optical micrograph of the passive SU-8 layer showing the release holes and thin-film interference fringes near the trace boundary, with a 100 micrometer scale bar

Released Device

Scanning electron micrograph of the released SU-8/Cu thermal actuator

05 — Experimental Characterization

Measuring out-of-plane motion.

The fabricated actuator was contacted on a probe station and driven through a controlled DC power sweep. Tip displacement was tracked optically while the device was characterized in air.

DC Input

Probe-Station Contact

Optical Readout

Input Control

Current and electrical power sweep

Measured Output

Out-of-plane tip displacement

Test Environment

Characterized experimentally in air

Underwater operation was proposed as a future application and was not experimentally demonstrated.

06 — Model-to-Hardware Validation

The model predicted the hardware.

Steady-state measurements followed the simulated trend across the tested power range, providing experimental confidence in the coupled multiphysics model.

Tip displacement vs. input power

Simulation–experiment comparison

Simulation

Experiment

63 µm

Maximum measured displacement

at 36 mW input power

Measured steady-state points

PowerDisplacement

2 mW2 µm

8 mW10 µm

18 mW27 µm

25 mW40 µm

36 mW63 µm

Strong agreement between simulation and experiment validated the design approach; remaining differences were attributed to environmental and measurement effects.

07 — Operating Boundary & Engineering Lessons

Performance has a thermal ceiling.

High-power testing exposed a geometry-dependent boundary between useful deflection and irreversible damage.

Device-specific operating boundary

Long Actuator

010203040 mW

63 µm measured at 36 mW

Short Actuator

010203040 mW

Burning observed above 30 mW

Presentation evidence also documents a damaged-device example around 40 mW.

Observed Failure Evidence

Optical micrograph of a thermal actuator glowing from Joule heating well above its safe operating power, showing irreversible high-power damage

High-power damage observation · experimental evidence

Irreversible Damage

Geometry Matters

Longer actuators produced greater displacement.

Power Limits Are Device-Specific

The safe operating margin changes with actuator length.

Next Iteration

Pattern the active layer and apply geometry-specific current limiting.

These observations define an engineering boundary—not a universal rated limit.