P/M PROCESSING AND CHARACTERIZATION OF CONTROLLED TRANSFORMATION TEMPERATURE NiTi

Abstract
Double vacuum arc remelted NiTi ingots were vacuum induction remelted and subsequently atomized using a high-pressure gas stream. Powders were collected, loaded into cans, and hot isostatically pressed (HIP'ed). Prior to HIP'ing, powders of various known transformation temperatures (A_s), were precisely blended to achieve a desired intermediate transformation temperature (A_s). This intermediate A_s follows an empirical relationship which approximates a rule of mixtures based on weight fraction. A_s were measured using differential scanning calorimetry (DSC). Results for HIP'ed and high temperature annealed specimens indicate that this technique was accurate and reproducible for measuring the transformation temperature of smooth, nearly symmetrical DSC curves. DSC/DTA thermograms in the literature typically show a double peak exothermic-martensitic reaction which was different than the smooth peaks observed for the HIP'ed condition. The specific thermomechanical history performed on the P/M material resulted in the double peak exothermic reaction. To interpret the peaks, thermal arrest experiments were conducted on both exothermic peaks, the results of which clearly support the existence of a premartensitic reaction. Thermal arrest experiments were also performed to analyze incomplete thermal cycles on both heating and cooling. Thermal arrest of the martensite reaction resulted in reduced energy absorbed for completion of the austenite reaction. Conversely, martensite energy was reduced by an arrest of the austenite reaction. However, reheating the sample revealed that the arrest split the austenite reaction into two distinct peaks, the split occurring at the temperature of the prior thermal arrest. These results and results of additional DSC experiments serve to emphasize the influence of thermomechanical history on the kinetics and energetics of the SME transformation. The P/M blending process technique was found to result in unprecedented control of the A_s temperature. Scientific/engineering considerations justify P/M processing and mechanical blending as opposed to conventional cast/wrought processing to achieve accurate A_s temperatures.