Abstract: Homogenised and quenched Ni65Al35 samples were heated and studied in situ in a CM20 electron microscope up to 900 degrees C. The Ni5Al3 phase first forming around 550 degrees C in the quenched L1(0) microtwinned martensite starts to decompose around 800 degrees C yielding B2 precipitates in a twinned L1(2) matrix. The latter twinning is a remainder of the microtwinning in the original room temperature martensite. Also the crystallographic relations between precipitates and matrix can be traced back to the original formation of twinned martensite plates within the austenite. Some aspects of the dynamics of the process are discussed on the basis of snap shots and video recordings.

Abstract: For column structures, such as fee-based alloys viewed along the cube direction, the concept of electron channelling through the atom columns is more and more used to interpret the corresponding HREM images. In the case of(partially) disordered columns, the projected potential approach which is used in the channelling description must be questioned since the arrangement of the atoms along the beam direction might affect the exit wave of the electrons. In this paper, we critically inspect this top-bottom effect using multi-slice calculations. A modified channelling theory is introduced which turns out to be very appropriate for the interpretation of these results. For substitutionally disordered column structures, it is also discussed how to link the chemical composition of the material to statistical data of the HREM image. This results in a convenient tool to discern images taken at different thicknesses and focus values.