Shape memory alloy transformation problem

[davidyanni10]

A friend of mine is working with shape memory alloys and he's got one that is behaving strangely.

At "low" temperatures a "fresh" solution heat treated sample will form martensite upon cooling, and austenite upon heating as expected. Heat and cool all you want and you get the transformation.

If heated to a "high" temperature his alloy will become "stuck" in the austenite phase and won't transform back to martensite upon cooling to room temperature. Heat and cool all you want and the sample will stay in the austenite phase.

The interesting thing is that a sample of his alloy that has been solution heat treated and then aged for 100 hours (not sure the aging temp,) then it behaves like it should even if it's heated to high temperatures. What I mean is the aged sample will transform to austenite when heated to "high" temp and then transform back to martensite upon cooling to room temp instead of getting "stuck" in the austenite phase.

Okay so my question is does anyone have any ideas about why this is happening, and any ideas on how to test any mechanisms you propose?

I am thinking that either:
1) the austenite→martensite transition is sensitive to composition and the aged sample has precipitated out some particles. Whatever components the precipitates are rich in, the matrix will subsequently be depleted in. This alteration of the matrix composition could enable the transformation to take place even after being heated to high temperatures (when really really tiny omnipresent precipitates would usually have been re-dissolved back into the matrix and upsetting the composition balance that allows the transformation)

or 2) The precipitates caused by aging act as stress concentrators or else have some residual stresses associated with them and so they act as "heterogenous nucleation sites" for the formation of martensite (I use the term nucleation even though the transformation isn't diffusion mediated so I'm not sure if it's appropriate. I figure the transformation has to start somewhere though).

So do you see a flaw in either of these arguments and/or any ideas on how to test them.

BTW, sorry I can't be more specific about his alloy's concentration and temperatures used etc. I'm not working with his data he just explained the problem qualitatively to me.

 

[nbryan5]

I think that the best way to test these ideas would be to perform a thermo-mechanical characterization of the alloy, in order to gain insight into the various transition temperatures and the microstructural changes that occur with aging. You could also do electron microscopy to look at the changes in the microstructure due to aging, as well as chemical analysis to see if there are any differences in composition between the aged and "fresh" samples. Finally, you could use differential scanning calorimetry (DSC) to measure the enthalpy change associated with the austenite→martensite transformation, in order to determine how the energetics of the transition is affected by aging.