Access to master’s theses is restricted to current ERAU students, faculty, and staff.

Date of Award

Fall 1996

Document Type

Thesis - Open Access

Degree Name

Master of Science in Aerospace Engineering

Department

Aerospace Engineering

Committee Chair

Dr. Eric v. K. Hill

Committee Member

Dr. Frank J. Radosta

Committee Member

Mr. Christopher J. Raczkowski

Abstract

Acoustic emission (AE) nondestructive testing can detect fatigue cracks as they occur in complex structures. One use for AE has been in-flight detection of fatigue cracks in aircraft. The KC-135 aircraft were successfully monitored as early as 1979. The main problem with this and subsequent applications was an unfavorable signal to noise ratio, the key being to separate the small amplitude crack signals from the large amplitude ambient noise. This was accomplished here through the use of a Kohonen self-organizing map (SOM) neural network.

In order to simulate a fuselage undergoing fatigue, a pressure vessel was constructed from a 0.040 inch thick 2024-T3 aluminum cylinder. The vessel contained a rivet line, a round hole with a notch filed in it to provide a stress concentration, and a repair patch and was cyclically pressurized from 20 to 80 psi in order to fatigue the aluminum and generate typical in-flight signals. During the test, AE sensors, powered by a data acquisition system, collected the AE parameter data from metal rubbing at the patch, rivet fretting at the rivet line, and fatigue crack propagation at the stress concentration. The SOM successfully separated the crack signals from the rivet and rubbing signals. A prototype system is currently being built by Martingale Research Corporation to provide aircraft with a real time in-flight fatigue crack growth monitoring capability.

Share

COinS