The martensite transformation in Ni-based coatings & bulk alloys

This thesis examines the formation of martensite in Ni-based super-alloys in both an alloy-coating system and in the bulk alloy, NiPtAl. The alloy-coating study was designed to test the hypothesis that an extra layer of Ni could favourably alter diffusion gradients within the alloy-coating system in order to preferentially promote martensite(ß ’) formation. Such martensite formation would have the advantage of delaying the development of y’-Ni3Al, and reducing precipitation of the brittle topologically close packed phases, on subsequent heat-treatment. The work was a success in that deposition of a layer of Ni on the alloy, prior to aluminising did alter the diffusion profile of the coating-alloy system. During the aluminising process, a layer of very Ni-rich y’ formed at the coating/alloy interface. As the diffusivity of elements through this phase is notably lower than that of elements diffusing through the Al-rich ß -NiAl phase, the coating produced was thinner and had a higher Al content. These effects resulted in a coating that retained a ß +ß ’ majority composition for a longer period of oxidation, than its un-plated equivalent. As the molar volume change of the ß - ß ’ transformation is less than that of ß - y’, this successful promotion of ß ’ over y’ may also aid in minimising phase transformations misfit, and any subsequent rumpling leading to delamination. Although the TCP needles were still formed on oxidation, the altered diffusion gradient permitted back diffusion of elements such as Ta, into the alloy, therefore enabling Ta to offer the strengthening properties for which it was added, and preventing it from succumbing to undesirable oxidation. The martensite investigated in the NiPtAl bulk alloys was shown to have a crystallographic structure that was tetragonal, with L1o 3R ordering. The alloys were very sensitive to Pt content, with a 5 at.% increase, from 5 to 10 at.% causing a 440 C rise in the Ms temperature, and more than doubling the required activation energy from 35 to 87kJ/mol. The level to which Pt could diffuse influenced whether thermal or coherent phonon activity would dominate for the transformation. Tempering of martensite was also sensitive to Pt content, specifically at 360 C which is suggestive that the nose of the martensite Time Temperature Transformation curve is close to this temperature.