Date of Award

5-2025

Document Type

Dissertation

Publisher

Santa Clara : Santa Clara University, 2025

Degree Name

Doctor of Philosophy (PhD)

Department

Bioengineering

First Advisor

I.E. Araci

Second Advisor

O.S. Pak

Abstract

This thesis presents the development and integration of capillaric strain sensors (CSSs) for human movement recognition. CSSs operate by detecting volume changes in closed microfluidic networks in response to linear strain, offering tunable directionality and sensitivity. By adopting strategies from electrical digital circuits, CSSs were configured in OR and AND logic gate arrangements, enabling simplified movement recognition without extensive computational resources. Digital image correlation techniques were employed to map strain fields, facilitating accurate predictions of sensor responses. Skin-mounted CSS patches as small as 3 × 3 mm² demonstrated effective human movement detection, highlighting their potential in wearable technology applications.

Building on the demonstrated programmability and sensitivity of CSS‐based logic arrays for movement recognition, we next sought to simplify the readout and enhance robustness by introducing a true digital switching mechanism. Consequently, we developed a digital strain sensor (DigSS) that leverages the geometry-dependent corner flow in capillaric channels to produce a binary ON/OFF response once a predefined strain threshold is reached. This behavior is achieved through the channel geometry dependence of corner flow in CSSs, resulting in an electrofluidic switch. The DigSS operates robustly for hundreds of cycles with a strain limit of detection of 0.0026. To facilitate integration, a linear optimization-based computer-aided design tool (CAD) for the integrated DigSS (iDigSS) was created. Benchtop experiments showed that the CAD based iDigSS is capable of distinguishing a target strain-field profile from 35 of the 36 theoretically distinguishable profiles without requiring signal processing. Human subject trials demonstrated the system's ability to differentiate a specific shoulder movement from five others and to wirelessly record wrist extension counts and durations.

To address the challenge of uneven strain-induced bubble generation in adjacent reservoirs, overlapping reservoir designs were implemented. This architectural modification ensured synchronized meniscus formation across reservoirs under nonuniform and dynamic strain conditions, enhancing the consistency and reliability of sensor outputs.

Together, these studies advance the understanding of how mechanical, chemical, and geometric design principles can be combined to create integrated skin patches for movement recognition and assessment. These findings enable the development of skin‐strain sensors tailored for next-generation human movement detection.

Available for download on Thursday, July 16, 2026

Share

COinS