Fatigue and crack-growth behavior in a titanium alloy under constant-amplitude and spectrum loading

Abstract

A titanium alloy (Ti-6Al-4V STOA) plate material was provided by the University of Dayton Research Institute from a previous U.S. Air Force high-cycle fatigue study. Fatigue-crack-growth tests on compact, C(T), specimens have been previously performed at Mississippi State University on the same material over a wide range in rates from threshold to near fracture for several load ratios (R = Pmin/Pmax). These tests used the compression pre-cracking method to generate fatigue-crack-growth-rate data in the near-threshold regime. Current load-reduction procedures were found to give elevated thresholds compared to compression pre-cracking methods. A crack-closure model was then used to determine crack-front constraint and a plasticity-corrected effective stress-intensity-factor-range relation over a wide range in rates and load ratios. Some engineering estimates were made for extremely slow rates (small-crack behavior), below the commonly defined threshold rate. Newly-developed single-edge-notch-bend, SEN(B), fatigue specimens were machined from titanium alloy plates. These specimens were fatigue tested at two constant-amplitude load ratios (R = 0.1 and 0.5) and a modified Cold-Turbistan engine spectrum. All of the tests were conducted under laboratory air and room temperature conditions. Calculated fatigue lives from FASTRAN, a fatigue-life-prediction code, using small-crack theory with an equivalent-initial-flaw-size (semi-circular surface flaw) of 9-μm in radius at the center of the semi-circular edge notch fit the constant-amplitude test data fairly well, but under predicted the spectrum loading results by about a factor of 2–3. The reasons for the large under prediction were discussed. Life predictions made with linear-cumulative damage (LCD) calculations agreed fairly well with the spectrum tests.