Eventos Anais de eventos
ENCIT 2022
19th Brazilian Congress of Thermal Sciences and Engineering
DETERMINATION OF THERMAL PROPERTIES AT HIGH TEMPERATURES BY THE APPLICATION OF AN OPTIMIZED INVERSE PROBLEM METHODOLOGY
Submission Author:
Ernandes José Gonçalves do Nascimento , SP
Co-Authors:
Ernandes José Gonçalves do Nascimento, Arthur Mendonça de Azevedo, Elisan dos Santos Magalhães
Presenter: Ernandes José Gonçalves do Nascimento
doi://10.26678/ABCM.ENCIT2022.CIT22-0349
Abstract
The evolution of engineering design techniques created a demand for determining material parameters under severe working conditions. However, the application of direct experimentation methods in measuring thermal properties usually does not provide accurate results for high-temperature appliances. The present work demonstrates the use of an inverse methodology numerical tool that makes possible the estimation of the thermal properties of a material at high-temperature values. The presented technique is an enhanced version of the Quadrilateral Optimization Method (QOM). The applied approach consists of a multivariable model developed to estimate the specific heat function’s parameters. A laser beam welding problem (LBW) was simulated in three dimensions to exemplify the efficiency of the proposed methodology. The partial differential heat conduction equation was converted into an algebraic form employing the Finite Volume Method (FVM). The enthalpy method was applied to account for the phase change in the weld bead. The Time-Traveling Regularization (TTR) scheme was applied in the objective function to regularize the results. The welding torch was modeled as a 3D gaussian volumetric heat source. Temperature-dependent thermal properties were applied to the welded specimen material. A parallelized in-house CUDA-C (Compute Unified Device Architecture) language code was developed to simulate the autogenous welding problem. The software computed the unsteady state simulation by running in a Graphical Processing Unit (GPU) instead of the standard Central Processing Unit (CPU) methodology. The Nvidia RTX3060 video card was used as simulating hardware device. Finally, a simulated welding experiment was performed to validate the proposed approach and estimate the real experimental errors. The proposed inverse problem tool demonstrated a very low specific heat average error. The enhanced methodology can reduce costs and increase accuracy compared to traditional advanced experimental apparatus applied in the measurement of thermal properties at high temperatures.
Keywords
Inverse problems, Numerical phase change, Finite volume method, GPU processing, CUDA-C language
