Abstract:
Numerical modeling of three variations of the process of gradient thermomechanical processing of a steel plate is performed to understand the patterns of thickness distribution of eigenstrains and residual stresses depending on the parameters of the technological process. One of these processes consisted in heating and holding the workpiece at the austenization temperature, followed by unilateral accelerated cooling; in the other two, the workpiece was previously subjected to plastic bending of a certain value of a positive or negative sign, and after unilateral accelerated and complete cooling, to bending of the opposite sign to straightening the workpiece. The numerical solution of the problem was performed in the COMSOL Multiphysics package. At the first stage, the problem of thermal conductivity (accelerated one-sided, and then slow complete cooling of the workpiece) was solved, then the problem of thermoelastoplasticity was solved to simulate the evolution of elastic and plastic deformations and residual stresses when the temperature of the workpiece changes. In this task, the dependence of elastic and plastic constants on temperature was taken into account. The bending and the straightening of the plate was modeled by setting a suitable thickness distribution of its eigenstrains. To confirm the calculation results, experiments were performed for a sheet of A32 marine steel with a thickness of 14 mm, for which an accelerated unilateral cooling unit for hot workpieces created on the basis of the Kerch Shipyard was used. A method is proposed for determining the distribution of eigenstrains and residual stresses over the thickness of the sheet, developing a cutting method by which these fields were determined in three specimens subjected to gradient mechanothermal treatment and one control specimen. The experimental results confirm the observation obtained in the calculation that all the methods of gradient thermomechanical treatment used provide compressive residual stresses on both sides of the sheet, the magnitude of which exceeds the background values obtained on the control sample.