GPU Accelerated Simulation of Conjugate Heat Transfer Problems

Abstract

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.Conjugate heat transfer plays a crucial role in numerous engineering applications, such as thermal management of electronic devices, aerospace heat shields, and energy systems. This paper explores the utilization of Graphics Processing Units (GPUs) and the CUDA-C programming language for solving conjugate heat transfer problems. Three distinct heat transfer scenarios are investigated: the Lid-Driven cavity, squared cavity with natural convection, and squared cavity with a solid square embedded in the center. These problems are solved using numerical methods and parallelized using GPU computing techniques to enhance computational efficiency and reduce simulation time. The Lid-Driven cavity problem involves the flow of a fluid within a square enclosure, where one side is subjected to a constant velocity boundary condition. In the squared cavity with natural convection, the study focuses on heat transfer phenomena occurring due to density-driven fluid motion. The buoyancy effects induce convective currents within the cavity, influencing the temperature distribution. The squared cavity with a solid square in the center represents a more complex conjugate heat transfer problem. The presence of the solid square influences the flow patterns and temperature distribution within the cavity. Through the utilization of GPUs and CUDA-C programming, the computational efficiency of solving conjugate heat transfer problems is greatly enhanced. The parallel processing capabilities of GPUs enable accelerated simulations, reducing the time required for solving these complex problems. The study demonstrated that the proposed algorithm achieved up to 99.7% reduction in simulation time for the laminar lid-driven cavity problem. The results obtained from the simulations provide valuable insights into the heat transfer characteristics, facilitating the optimization of thermal management systems and the design of more efficient heat exchangers. Overall, this study demonstrates the effectiveness of GPU computing in tackling conjugate heat transfer problems and its potential for advancing the field of thermal sciences.

Keywords