By Theresa Gabrielli
December 21, 2022
Third-year Ph.D. student Caleb Schoenholz was awarded third place in the Society for the Advancement of Material and Process Engineering (SAMPE) University Research Symposium (URS) competition for his paper, “An automated evaluation method of tool surface condition in composites manufacturing using machine learning and sparse sensing.”
“I applied to the URS competition to share my passion for materials science and engineering,” said Schoenholz. “[I wanted to] showcase some of the innovative research being conducted in the UW Composites Group, and learn from other current and upcoming leaders in the SAMPE community.”
The URS is a national, year-long contest for both graduate and undergraduate students in MSE-related research fields, to present the results of their technical research at the Composites and Advanced Materials Expo. The process to compete is very involved, beginning with an initial application and abstract submission, after which semi-finalists are invited to send in their full papers. Schoenholz was among 11 Ph.D. student finalists selected and invited to present their papers to a panel of judges at the conference in Anaheim, CA.
The research coming out of the UW Composites Group, led by Assistant Professor Navid Zobeiry, encompasses the development of new composite materials, improving how existing composite materials are manufactured, and determining ideal methods of recycling or degrading composite materials that have reached the end of their lifecycle.
Schoenholz’s paper addressed a common issue faced in aerospace manufacturing. The molds for manufacturing large composite parts are often treated with a semi-permanent “release coating” that gives the mold a sort of non-stick surface, allowing the part to be easily removed. Unfortunately, these coatings are only about 1-3 microns thick and degrade with each use. Imperfections in the coating can lead to deformed or contaminated parts, and therefore the coating must be regularly inspected, cleaned, or reapplied by production workers. These workers’ only real method of deciding what a coating needs is their personal instincts and know-how, which is not always reliable.
To address this, Schoenholz and his co-authors developed an automated inspection tool involving a camera, a microscope and a Fourier-transform infrared spectrometer, all equipped with machine learning models. This tool can be integrated into an existing manufacturing platform to automatically inspect a mold’s release coating and provide more precise data about it, allowing production workers to better determine whether the coating needs maintenance, which improves production efficiency. This data could also be used to better understand how release coatings age, leading to further developments in how manufacturing tools are prepared.
“This work included both fundamental scientific research and industry-focused engineering,” said Schoenholz. “I thought this project’s findings would appeal to both academics and industry professionals within SAMPE.”
Clearly, he was right. After a monthlong deliberation, the judges awarded him third place, which included a $400 cash prize.
“I thought all the student finalist presentations were top-notch, so I was honored to be selected as one of the top three,” he said.
Schoenholz will give another presentation on this research topic at the national SAMPE conference in April. This research was funded by the Joint Center for Aerospace Technology Innovation (JCATI) and Toray Industries, Inc. The full paper will be published in the January 2023 issue of SAMPE Journal’s Tooling Technology Advancement/Applications.