Volume 6, Issue 5 (May 2009)
Modeling of Quenching and Tempering Induced Phase Transformations in Steels
Quenching and tempering are common processes used in the manufacture of steel components. The development of simulation tools for quenching is critical for improving process performance by minimizing component distortion and maximizing service life. While modeling of quenching has received considerable attention, there has not been much work on similar tools for tempering. This paper presents an efficient simulation tool to predict microstructure, temperature, and stress evolution, and is applicable to both quenching and tempering processes. This method takes into account the temperature dependence of material properties, transformation strains, latent heats of transformation, and transformation plasticity. Furthermore, three different micromechanical approaches are implemented and studied to simulate steel as a multiphase constitutive material: the average property model, the Voigt model, and the Reuss model. These models assume that the properties of a unit volume of material can be derived, respectively, by applying the linear rule of mixtures to the material properties of its constituent phases, by assuming that all constituent phases have the same strain field, and by assuming that all constituent phases have the same stress field. The model predicts the microstructure, stress, and distortion in the heat treated component. The simulation model is implemented within the framework of the ABAQUS finite element package by taking advantage of its advanced features to incorporate user defined material properties. Given that the material properties are strongly dependent on carbon content, the simulation method is tested using experiments with modified 4320 steel plates that were carburized on one side to amplify distortion when quenched, due to martensitic phase transformation. Distorted shapes are measured and compared to model predictions for both quenching and tempering. The detailed comparisons provide confidence in the model as well as suggestions for improvement.