The shift in computational paradigms driven by large-scale scientific computing problems has propelled the development of general-purpose graphics processing units (GPGPU). The emerging lattice Boltzmann method in computational fluid dynamics (CFD) demonstrates significant advantages in computational efficiency and parallel scalability when coupled with advanced physical models. This study is based on the lattice Boltzmann method, designs and completes large-scale parallel computing in heterogeneous systems, and achieves high-fidelity numerical simulations of turbulence/transition and heat transfer with the following specific tasks: (1) An algorithmic framework and data communication based on the lattice Boltzmann method were designed on a heterogeneous computing platform. (2) A parametric combined momentum-exchange method combined with unified single-node curved boundary conditions is proposed for accurately calculating aerodynamic lift. (3) High-fidelity numerical simulations of three-dimensional Taylor-Green vortex flow were conducted by the D3Q19 model. The parallel implementation method proposed in this study achieves linear speedup and good parallel scalability, demonstrating the potential for efficient simulation on Exascale supercomputing systems.

Lattice Boltzmann method,Turbulence/Transition,Heat transfer,Compressible flow,High performance computing