Term of Award
Fall 2025
Degree Name
Master of Science in Applied Engineering (M.S.A.E.)
Document Type and Release Option
Thesis (restricted to Georgia Southern)
Copyright Statement / License for Reuse
This work is licensed under a Creative Commons Attribution 4.0 License.
Department
Department of Manufacturing Engineering
Committee Chair
Hossein Taheri
Committee Member 1
Bishal Silwal
Committee Member 2
Jana Abou Ziki
Abstract
This research develops and validates a Model-Assisted Probability of Detection (MAPOD) framework to quantify the reliability of Phased Array Ultrasonic Testing (PAUT) and the Total Focusing Method (TFM) for detecting AWS D1.5-classified weld defects. Conventional Probability of Detection (POD) studies demand extensive libraries of flawed specimens, which are costly and impractical for structural welds. MAPOD addresses this limitation by combining experimental POD with finite element simulations of ultrasonic wave–defect interactions, producing statistically robust and physics-informed POD results. Experimental studies were performed on AWS D1.5-compliant specimens containing planar flaws (toe cracks, root cracks, lack of fusion) and volumetric flaws (porosity, slag inclusions). Both Hit/Miss and signal-response POD methods were conducted to extract a90/a95 values and assess detection sensitivity. At the same time finite element models in COMSOL were used to simulate flaw responses at multiple ultrasonic frequencies (1.5, 2.0, and 2.5 MHz) to verify the outcome of experimental POD analysis. Experimental POD analysis results verified by simulation response feedback (MAPOD) with planar flaws showed that TFM achieved superior detection and sizing planar flaws especially shallow cracks, due to its pixel-based focusing and diffraction capability. Similarly, PAUT demonstrated greater consistency in identifying volumetric flaws, aided by its broad sectorial coverage. MAPOD frequency analysis also indicated that higher frequencies (2.5 MHz) improve sensitivity to small and shallow flaws, on the other hand lower frequencies (1.5–2.0 MHz) offer better penetration for larger and deeper flaws. This study confirms that MAPOD not only validates experimental POD results but also predicts inspection reliability under conditions that are difficult to reproduce experimentally. Beyond weld inspection, MAPOD offers a generalizable decision-support framework for selecting inspection methods and parameters in advanced industries such as additive manufacturing.
Recommended Citation
Irtiza, Chowdhury MD, "Model Assisted Probability of Detection Analysis of Phased Array Ultrasonic Testing and Total Focusing Method for AWS D 1.5-Classified Weld Defects" (2025). College of Graduate Studies: Theses & Dissertations. 3030.
https://digitalcommons.georgiasouthern.edu/etd/3030
Research Data and Supplementary Material
No
