Atomic scale modeling of structural phase transformations in AlCrFeMnMo high-entropy alloys during thermal treatments

Ordering and phase separation in multi-principal-element metallic alloys: Contribution from mean-field atomic-scale modelling and simulation

“…This two-phase enclave growing exclusively within the LRO domain (while the A2 domain is only subjected to uniform shrinking parallel to the Fe-Co axis) is eventually found to reach the equiatomic composition at roughly 500 K. This is visible on Figure 3d, which also displays two constant- Co isopotentials delineating the PS domain around the equiatomic composition. Interestingly, the PS is found to be of mixed ordered / disordered A2 + B2 type, a feature in agreement with the conclusions of an earlier MC work [9,13] on the same system using the same energetics. One should however keep in mind that the isopotentials displayed on isothermal sections (Figure 3d) are not sufficient to infer the precise compositions of the phases occurring at PS, since in general, due to the constraints on xD and xE,  D and  E are not constant are discontinuous on either side of this domain along the isopotential (cf.…”

“…Following the same guidelines as in our previous MC study [13], we first focus on the general properties (i) and (ii), namely trends on long-range order (LRO) and phase separation (PS), as predicted from PMF, considering in turn 1) AlCoCrFeNi, 2) AlCrFeMnNi, 3) AlCrFeMnMo. As mentioned above, our approach conveniently provides, for decreasing temperatures, pseudo-ternary isothermal sections at fixed atomic fractions for a couple of chemical species (D and E in the notation of the Appendix) selected at convenience, respectively (Cr;Ni), (Mn;Ni) and (Cr;Mo) for the three systems listed previously.…”

“…Conversely, the LRO panorama becomes hardly tractable in higher-order systems such as quinary ones, owing to the drastically increasing number of possible combinations for the chemical species on the four sublattices. In this context, following previous MC simulation works [9,13], our investigations of LRO were simplified by restricting to a unique general B2-type kind of LRO, covering all these hardly classifiable possibilities, and considered to be the sole complementary case to the A2 solid solution. This approximate distinction between A2 disorder and B2-type LRO is sufficient for our purpose of characterizing overall LRO and PS trends, and these two MPEA states will be displayed respectively in red and black on all the phase diagrams presented below.…”

“…In this respect, the above frame for atomic-scale simulations may be useful to compare more rigorously candidate alloys, since it provides a quick route to investigate the expected effect on properties (i) and (ii) of elaboration choices such as Fe enrichment and Mo depletion. To bring information on these issues, a previous work [13] was carried out by Monte-Carlo (MC) simulations using the above mentioned 1NN pair energetics. While this previous MC simulation work was primarily concerned with the interpretation of the experimental behaviour of AlCrFeMnMo, the approach was made more comprehensive and illustrative by considering three bcc MPEA systems, in order to monitor the effect of the AlCoCrFeNi → AlCrFeMnNi → AlCrFeMnMo sequence for alloy selection (corresponding to substitution of Co with Mn, followed by substitution of Ni with Mo).…”

“…We focus on bcc alloys, being primarily concerned with the ill-known AlCrFeMnMo system, but we also include in our analysis the more widely documented AlCoCrFeNi system. To investigate the strengths and weaknesses of PMF in the context of MPEAs, we also make appropriate comparisons with a previous work [13] concerned with the same alloys using MC simulations, taking the latter as a thermodynamically exact, reference case.…”