Comments on the interpretation of differential scanning calorimetry results for thermoelastic martensitic transformations: Athermal versus thermally activated kinetics

摘要

In the previous article (1) Van Humbeeck and Planes have made a number of criticisms of our recent paper (2) conceming.the interpretation of the results obtained by Differential Scanning Calorimetry (DSC) from the Martensjfie Transformation of Cu-Al-Ni-Mn-B alloys. Although the martensitic transformation df these shape memory alloys is generally classified as athermal, it has beencbnfirmed that the capacity of the alloys to Undergo a more Complete thermoelastic transformation (i.e. better reversibility of the transformation)' increased with the Mn content (2, 3). This behaviour has been explained byihterpreting the DSG results-obtained during thermal cycling in terms of a thermally activated mechanism controlling the direct andbfceverso transformations (2). In particular, the DSC curves characteristic of heat absorption orreleascduring the reverse or direct transformation were not identical when the heating/cooling rates varied (as shown in figure 2 of (2)). When the healing rate increases during the reverse transformation the DSC curves.shift towards higher temperatures while they shift towards the lower temperatures when the cooling rate was increased during the direct transformation. Since the starting transformation temperatures (As, Ms) donotshifti Van Humbeeck and Planes (I) state that there is no real peak shift and assume that our DCS experiments werecarried out without taking into account the thermal lag effect between sample and cell. On tha following line theydeduce a time constant, t, of 60 seconds because the peak maximum shifts. In factlhe assumption made by Van Humbeeck and Planes is false. Gur experiments were carried out aftencbntrollJiig that the instrumental shift was neglignble by melting indium at different heating rates" varying between l-SQ'C/min, The melting point of indium is 156.4°C, which is within the range of transformation temperatures of our Ml 5 alloy (sec figure 2 in (2)). Thus, the values of the time constant measured, fofthe&ting rates between 10-30°C/min. vary between 6.2 and 6.8 seconds. The shift of the peak dueto iajstrutnentartimo'lag deduced from the expression AT = x(dT/dt) varied between 0.3-1 °C for the heating ra{e