Non-equilibrium thermodynamics of Quench and Partition (Q&P) steelsPublic Deposited
In support of a scientific foundation for the predictive design of composition and processing of quench and partition (Q&P) martensite/austenite TRIP steels, theory of coupled diffusional/displacive transformation is experimentally calibrated to control austenite carbon content and its associated mechanical stability. Under paraequilibrium constraint, the calibration quantifies an effective BCC stored energy that incorporates both stored and dissipated energy associated with displacive interfacial motion during the partitioning treatment. Consistent with prior study of bainitic transformations, the effective stored energy is found to decrease linearly with partitioning temperature, attributed to the effect of dislocation recovery on the forest hardening contribution to interfacial friction. The calibrations are based on highly accurate experimental measurements using electron microscopy, high-energy x-ray diffraction and 3D atom probe tomography to quantify the amount and carbon content of retained austenite as a function of Q&P treatment. Varying the initial quench temperature to vary the initial retained austenite amount, it is demonstrated that the effective BCC stored energy changes with the direction of motion of the interface in association with a sign change of the dissipation contribution, favoring greater C partitioning for BCC->FCC motion. The minimum time for completion of partitioning at the partitioning temperature is consistent with DICTRA paraequilibrium diffusion simulations incorporating the effective stored energy. Measuring retained austenite mechanical stability by the characteristic MsÏƒ temperature below which transformation controls yielding, martensite nucleation theory is calibrated using the forest hardening friction derived from the partitioning experiments to define the characteristic nucleation site potency in the retained austenite. Model predictions of austenite carbon content and stability were validated using a set of new designed alloys with individually optimized Q&P cycles. Comparison of the stress-strain curves of the Q&P martensite/austenite samples with fully martensitic material shows little influence of the austenite on yield strength, but a dramatic reduction of the ultimate tensile strength by transformation softening which greatly reduces initial strain hardening, retaining higher hardening to higher strains for greatly enhanced flow stability. An unusual correlation between MsÏƒ and the temperature of maximum ductility is attributed to a bimodal austenite stability associated with two morphologies of blocky vs thin-film in the martensite microstructure. A principal limitation of the Q&P martensite/austenite steels relative to their bainite/austenite counterparts is the more rapid tempering of martensite leading to a major fraction of the alloy carbon being lost to carbides. A preliminary parametric analysis correlating rate of carbide precipitation to paraequilibrium cementite driving force and coarsening rate constant predicts Cr as the most effective alloying element to retard carbide precipitation, with some support from literature data. Predictive design of carbide-free Q&P martensite/austenite steels could double the amount of optimal stability austenite for greatly enhanced TRIP to achieve useful ductility of significantly higher strength levels.