Faculty Mentor
Hossein Taheri
Location
Russell Union Ballroom
Type of Research
Proposed
Session Format
Poster Presentation
College
Allen E. Paulson College of Engineering & Computing
Department
Manufacturing Engineering
Abstract
Resonant ultrasonic spectroscopy (RUS) is a common non-destructive testing (NDT) technique that is used for quality assessment of the parts and components in various applications. RUS is performed by exciting samples by an external load and capturing its vibrational frequencies to be analyzed and find irregularities to classify parts. RUS has been investigated to test additively manufactured Inconel 718 samples for the feasibility of accurate quality assessment. Frequency and amplitude data was collected, transformed using z-score normalization and then analyzed using different classification algorithms. Machine learning algorithms has been deployed for defect classification of good and bad samples. These findings will help advance defect detection through non-destructive testing methods.
Program Description
.
DOI
10.20429/GS4.2026.004
Start Date
4-23-2026 10:00 AM
End Date
4-23-2026 12:00 PM
Recommended Citation
Richard, Abigail, "Machine Learning-Based Defect Identification using Resonant Ultrasonic Spectroscopy" (2026). GS4 Student Scholars Symposium. 58.
https://digitalcommons.georgiasouthern.edu/research_symposium/2026/2026/58
Machine Learning-Based Defect Identification using Resonant Ultrasonic Spectroscopy
Russell Union Ballroom
Resonant ultrasonic spectroscopy (RUS) is a common non-destructive testing (NDT) technique that is used for quality assessment of the parts and components in various applications. RUS is performed by exciting samples by an external load and capturing its vibrational frequencies to be analyzed and find irregularities to classify parts. RUS has been investigated to test additively manufactured Inconel 718 samples for the feasibility of accurate quality assessment. Frequency and amplitude data was collected, transformed using z-score normalization and then analyzed using different classification algorithms. Machine learning algorithms has been deployed for defect classification of good and bad samples. These findings will help advance defect detection through non-destructive testing methods.
