Eventos Anais de eventos
ENCIT 2018
Brazilian Congress of Thermal Sciences and Engineering
A NOVEL OPTIMIZATION METHODOLOGY APPLIED TO PARAMETERS OF HEAT SOURCES IN WELDING NUMERICAL SIMULATIONS
Submission Author:
Rodrigo Martins Farias , SP , Brazil
Co-Authors:
Rodrigo Martins Farias, Paulo Teixeira, Louriel Vilarinho
Presenter: Rodrigo Martins Farias
doi://10.26678/ABCM.ENCIT2018.CIT18-0057
Abstract
Welding processes on plates involve complex physical and chemical phenomena, which make mathematical modeling difficult. Several types of heat sources are available to be used in numerical simulations of welding processes. However, to get accuracy, some heat sources have many geometry parameters to be set and it demands much computational time, even using optimization methodologies. In this paper, a welding numerical analysis of single pass butt joint weld, with square groove, of AISI 1020 carbon steel, using the conventional TIG process is shown. The plates are 3.2 mm thick, 100 mm long and 50 mm wide. The numerical simulations were performed by ANSYS® Multiphysics software, considering three types of moving heat sources distributions (Gaussian, Conical and Double Ellipsoid), convection and radiation heat transfers on the surfaces and temperature-dependent material properties. The fusion zone shapes of the plates obtained from the experiments developed at the laboratory of research in welding engineering (LAPROSOLDA – UFU), located in Uberlandia, MG, Brazil, are compared with numerical results. This data, obtained by macrography, were used by a new optimization methodology to find the optimal parameters of these heat sources, by using a Genetic Algorithm approach. Analysis of the computational time and the accuracy in relation to experiments was carried out. In general, thermal results obtained by numerical simulation are in good agreement with experimental ones. Besides, the use of the proposed methodology decreases strongly the computational time.
Keywords
optimization methodology, heat source, Welding Processes, Numerical simulation, reduced geometry
