October_AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | O C T O B E R 2 0 2 0 3 6 Nitinol materials [7] . This figure is quite revealing and demon- strates that not all Nitinol material is the same with respect to fatigue properties. A logistic statistical analysis was con- ducted to determine the most probable factors to explain these results. Among the factors considered were microscop- ic stress concentrators (inclusion length and location, and grain size), probabilistic factors (area fraction of inclusions), and macro-mechanics (transformation temperature, upper and lower plateau stress, strain amplitude, mean strain, test temperature). The model that most accurately represented the experimental data contained inclusion length, density of inclusions, and strain amplitude. Of these variables, because strain amplitude is a test condition control variable and not a material characteristic, the effects of inclusions are the most dominant factors to influence the prediction of fatigue behav- ior. Furthermore, the influence of inclusion length was four times greater than that of inclusion density to predict fatigue fracture. For example, in Fig. 1 the Nitinol material with the lowest fatigue strain limit has a device maximum inclusion length of 101 µm, whereas, the material with the greatest fa- tigue strain limit in this study has an inclusion length of only 40 µm. This investigation also demonstrated that the effects of inclusion length (measured on the finished device) were more pronounced under cyclic bending conditions compared with the uniaxial tension-tension cycling. This difference is likely due to the amount of deformed volumes (outer surfaces in bending and entire cross section in axial) and the probability of an inclusion located in the high-strain region. These results afford great insight into the potential ben- efits of even greater purity microstructures, such as those observed in electron beam remelted (EBR) Nitinol. A recent study [9] compared the fatigue response of EBR Nitinol with other high-purity Nitinol with diamond coupons, similar to those in Robertson et al. [7] . Diamonds were processed from 10 mm OD x 0.53 mm wall thickness tubing to a transforma- tion temperature of 20˚C. Testing was conducted at 37˚C with a crimp strain of 6%, mean strain of 5%, and a range of strain amplitudes to 10 million cycles. The results of these tests are shown in Fig. 2a and confirm that the two high-purity materi- als from the Robertson et al. paper have comparable fatigue behavior. The EBR Nitinol fatigue data illustrates an enhanced 10 million-cycle fatigue strain limit (~1.9%) based on testing from 13 diamond lots, two coupon manufacturing vendors, and from five ingots. Figure 2b-d shows representative micro- structures of the three high-purity Nitinol materials in Fig. 2a from the longitudinal direction from the diamond coupon test specimens. The darker gray particles are oxide inclusions that form in the melt process; some of the particles in Fig. 2b are FEATURE Fig. 2 — (a) Probability of Nitinol diamond fracture at 10 million cycles versus strain amplitude plots for Generation II VAR, Generation II VIM/ VAR and Generation III VAR/EBR. (b) Generation II VIM/VARmicrostructure showing a ~40 µm long oxide inclusion along with amixture of small- er oxide and carbide particles. (c) Generation II VARmicrostructure showing a ~40 µm long oxide inclusion “stringer” with voids. (d) Generation III VAR/EBR microstructure showing a dispersion of ≤ 5 µm oxide inclusions. After Pelton et al. [9] . (a) (b) (c) (d) 6