Degree Type

Dissertation

Date of Award

2020

Degree Name

Master of Science

Department

Industrial and Manufacturing Systems Engineering

Major

Industrial Engineering

First Advisor

Frank Peters

Abstract

The objective of this research is to investigate the impact of varying cast surface conditions on fatigue performance, in the presence of other casting indications such as gas and shrinkage porosity. Additionally, this research aims to draw connections between nondestructive evaluation (NDE) techniques and fatigue results of cast test specimens. A process of specimen manufacturing, processing, and inspection is presented in this research, along with fatigue testing results. It is known that poor surface condition can impact fatigue life, even when comparing surface finishes produced by different manufacturing processes. Cast surface roughness is thought to contribute to reduced fatigue life, which may lead to over-processed or over-designed parts. Little has been done to investigate the impact of different cast surface conditions on fatigue life to justify current industry practices. Fatigue specimen design, inspection techniques, and fatigue testing techniques were developed in this study to compare the impact of cast surface condition on fatigue in the presence of other indications. To investigate this impact, axial load-controlled high-cycle fatigue tests were conducted on large lab-scale specimens cut from cast plates. All specimens underwent radiographic inspection, wet magnetic particle inspection, laser scanning, and visual surface characterization. Cast surfaces were characterized utilizing ASTM A802 comparator plates and through digital methods. Fatigue results showed no difference in mean fatigue lives produced by different surface classifications. Additionally, no correlation was found between digital surface classification and fatigue life. These results indicate that cast surface texture is a not reliable indicator of fatigue life. Post-test measurements of fatigue crack initiation sites provided statistically significant results in a log-log regression with fatigue life. This shows that variation in fatigue performance for a given cast material can be explained by the size of casting indications.

DOI

https://doi.org/10.31274/etd-20210114-149

Copyright Owner

Jeffrey Alan Tscherter

Language

en

File Format

application/pdf

File Size

66 pages

Share

COinS