Abstract
Reliable ignition-assistant technologies are essential for combustion engines in modern aviation to optimize engine performance using flexible fuels for unmanned aerial vehicles (UAVs). Computational modeling can help characterize the operation of such ignition-assistant devices. In recent years, smoothed particle hydrodynamics (SPH) has proven to be a useful tool for investigating various thermo-fluidic phenomena. SPH discretizes the domain into numerical particles, each representing physical attributes such as mass, density, and temperature. While SPH offers several advantages, its computational demands are significant. Thus, a GPU-accelerated SPH solver using the CUDA framework is developed in this study. The present focus is on studying the temperature distribution and thermal stresses in a hot surface ignition device. The study pursues two main goals. Initially, it involves validating the developed solver across fundamental heat transfer problems with known analytical solutions. These tests involve various boundary conditions in two- and three-dimensional geometries to assess accuracy and effectiveness. Subsequently, a realistic glow plug geometry was modeled under engine-relevant conditions. The predicted temperature history and distribution on the plug surface agree well with the experimental data. Following that, the temperature and thermal stress responses of the plug for different materials were parametrically analyzed. The simulation results show that alumina exhibits a high thermal stress due to its high Young's modulus and thermal expansion coefficient. Meanwhile, β-sialon and silicon nitride exhibit similar behaviors.