Optimal dynamic filtration approach for inverse heat conduction problems with moving body

Abstract

A Kalman filter method in conjunction with the nonuniform control volume methodology is adopted as a numerical tool for solving two-dimensional inverse heat conduction problems for a moving body. This work focuses on estimation of the transient temperature in the solid region of the workpiece during welding process that involves a moving heat source. The following parameters having major influence on the numerical solution are considered: the welding speed, thermal properties, the steps size (such as δx, δy, time step), the number of the sensors and the location of the sensors. The noise in measurement data is also considered to validate the method of solution. The accuracy of the numerical solution for inverse problem is examined by comparing the results with direct solution of the problem. The results show that this method can be the powerful tool for solving inverse heat conduction problems, with or without moving heat sources.