Non-Zero Mean Fatigue Tests of Nitinol Wire

The effect of mean strain on fatigue of superelastic Nitinol has long been known to be nonlinear [1, 2, 3].  Pelton, Gong and Duerig [4] were the first to systematically study the mean strain effect of superelastic Nitinol using diamond shaped fatigue specimen that simulated the processing of laser cut self expanding stent.  However, non-zero mean fatigue test of superelastic Nitinol wire was difficult.  The most knowledge on superelastic wire fatigue characteristics are learned from rotary bending tests [5].

This article uses a C-shaped specimen to evaluate the fatigue performance of superelastic wire under low, yet non-zero mean strains.  The goal was to create a 10 million fatigue limit at non-zero mean so that a fatigue safety line could be drawn for design of superelastic Nitinol wire based medical implantable devices such as stent.  The results are compared with Pelton et al [4] and the results from rotary wire fatigue tests [5].  It is found that in the low mean strain range that we have performed tests, the tread of the fatigue limit is close to the Soderberg approach.  This indicates that Pelton’s data at very low mean strain is likely caused by a processing related lower fatigue limit inside of the diamond nodes.