Back to search

Georgia Institute of Technology

Deep Learning in Ultrasonic Wave Inversion for Thin Coatings

Abstract

dc:description.abstract

This research focuses on the use of machine and deep learning to non-destructively characterize the quality of a coating in a layered system in terms of thickness and uniformness. The coating's parameters (thickness and uniformness) are evaluated using dispersion curves. To obtain these dispersion curves in the first step, this work uses a computational finite element analysis (FEA) model to obtain dispersion data for coated specimen with different thicknesses. The simulated FEA data is then processed via a two-dimensional Fourier transform to obtain a frequency - wave number relation. This representation is a dispersion curve. Dispersion curves are characteristic and depend on the coating thickness and uniformness. Simulated dispersion curves for various coatings are obtained. They are then further processed and a feature representation for each dispersion curve is extracted. Those extracted features are then fed into machine learning classifiers which allow a thickness classification. The above-described machine learning procedure is limited to classifying the thickness of a uniform coating. To classify if a non-uniformess is present, deep learning steps in. To obtain information about if a coating is uniform, two convolutional neural network architectures are used. Both networks are evaluated and tested and recommendations on their use are given.

Degree

thesis:*
Level thesis:degree_level
Masters
Department dc:contributor.department
Civil and Environmental Engineering
Grantor dc:publisher
Georgia Institute of Technology
Year dc:date.issued
2022

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Schmitz, Maximilian
Advisor dc:contributor.advisor
  • Jacobs, Laurence J.
Committee members dc:contributor.committeemember
  • Valle, Christine
  • Kim, Jin-Yeon

Subjects

dc:subject × 8

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
http://hdl.handle.net/1853/66518

Chain of custody

source
Harvested from
Georgia Tech
Base URL
repository.gatech.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Schmitz, Maximilian. Deep Learning in Ultrasonic Wave Inversion for Thin Coatings. Masters thesis, Georgia Institute of Technology, 2022. http://hdl.handle.net/1853/66518