Document Type

Dissertation

Date of Award

7-2018

Keywords

Applied sciences, Advanced/additive manufacturing, Aerosol jet printing (AJP), Computational fluid dynamics (CFD), Flexible and hybrid electronics, Physics-based monitoring and control

Degree Name

Doctor of Philosophy (PhD)

Department

Systems Science and Industrial Engineering

First Advisor

Dr. Mark D. Poliks

Subject Heading(s)

Applied sciences; Advanced/additive manufacturing; Aerosol jet printing (AJP); Computational fluid dynamics (CFD); Flexible and hybrid electronics; Physics-based monitoring and control; Operations Research, Systems Engineering and Industrial Engineering

Abstract

Aerosol jet printing (AJP)—a direct-write, additive manufacturing technique—has emerged as the process of choice particularly for the fabrication of flexible and hybrid electronics. AJP has paved the way for high-resolution device fabrication with high placement accuracy, edge definition, and adhesion. In addition, AJP accommodates a broad range of ink viscosity, and allows for printing on non-planer surfaces. Despite the unique advantages and host of strategic applications, AJP is a highly unstable and complex process, prone to gradual drifts in machine behavior and deposited material. Hence, real-time monitoring and control of AJP process is a burgeoning need. In pursuit of this goal, the objectives of the work are, as follows: (i) In situ image acquisition from the traces/lines of printed electronic devices right after deposition. To realize this objective, the AJP experimental setup was instrumented with a high-resolution charge-coupled device (CCD) camera, mounted on a variable-magnification lens (in addition to the standard imaging system, already installed on the AJ printer). (ii) In situ image processing and quantification of the trace morphology. In this regard, several customized image processing algorithms were devised to quantify/extract various aspects of the trace morphology from online images. In addition, based on the concept of shape-from-shading (SfS), several other algorithms were introduced, allowing for not only reconstruction of the 3D profile of the AJ-printed electronic traces, but also quantification of 3D morphology traits, such as thickness, cross-sectional area, and surface roughness, among others. (iii) Development of a supervised multiple-input, single-output (MISO) machine learning model—based on sparse representation for classification (SRC)—with the aim to estimate the device functional properties (e.g., resistance) in near real-time with an accuracy of ≥ 90%. (iv) Forwarding a computational fluid dynamics (CFD) model to explain the underlying aerodynamic phenomena behind aerosol transport and deposition in AJP process, observed experimentally.

Overall, this doctoral dissertation paves the way for: (i) implementation of physics-based real-time monitoring and control of AJP process toward conformal material deposition and device fabrication; and (ii) optimal design of direct-write components, such as nozzles, deposition heads, virtual impactors, atomizers, etc.

Download

Share

COinS