**Azadeh Keshtgar \*, Christine M. Sauerbrunn and Mohammad Modarres**

Center for Risk and Reliability, University of Maryland, College Park, MD 20742, USA; csauerbr@terpmail.umd.edu (C.M.S.); modarres@umd.edu (M.M.)

**\*** Correspondence: keshtgar@umd.edu; Tel.: +1-301-789-3106

Received: 29 May 2018; Accepted: 23 July 2018; Published: 25 July 2018

**Abstract:** In this paper, AE signals collected during fatigue crack-growth of aluminum and titanium alloys (Al7075-T6 and Ti-6Al-4V) were analyzed and compared. Both the aluminum and titanium alloys used in this study are prevalent materials in aerospace structures, which prompted this current investigation. The effect of different loading conditions and loading frequencies on a proposed AE-based crack-growth model were studied. The results suggest that the linear model used to relate AE and crack growth is independent of the loading condition and loading frequency. Also, the model initially developed for the aluminum alloy proves to hold true for the titanium alloy while, as expected, the model parameters are material dependent. The model parameters and their distributions were estimated using a Bayesian regression technique. The proposed model was developed and validated based on post processing and Bayesian analysis of experimental data.

**Keywords:** reliability; acoustic emission; structural integrity; crack growth; fatigue life prediction; uncertainty analysis; nondestructive testing
