The simulation of beam energy transfer and medium dynamics is essential in nuclear studies and designs involving the beam loading on targets. Although the existing methods are relatively mature in simulating the thermohydrodynamics of a solid or liquid target coupled with an energetic beam, they may not be entirely applicable to a complex porous medium such as granular materials with massive discrete elements. In this study, a GPU (Graphics Processing Unit) based discrete energy deposition simulation method is proposed for the thermohydrodynamical simulation of granular flow targets with beam-grain coupling. The results of physical validations show that a significant improvement in accuracy was achieved by our method compared to the equivalent homogenization method which just simplified the granular medium to a whole block. This method provides a more accurate alternative solution to the problem of calculating energy deposition on grains. Although the method is computationally intensive, it can be synchronously executed with the granular dynamics simulation owing to the powerful parallel computing capability of GPUs. Thus, the studies on the dynamical and thermal behaviors of beam-grain coupling can be conducted in a relatively fast and precise way.

束能量转移和介质动力学的模拟在涉及目标上的束加载的核研究和设计中是必不可少的。尽管现有方法在模拟固体或液体目标与高能束耦合的热流体力学方面相对成熟，但它们可能并不完全适用于复杂的多孔介质，例如具有大量离散元素的颗粒材料。在这项研究中，提出了一种基于 GPU（图形处理单元）的离散能量沉积模拟方法，用于具有束-粒耦合的颗粒流目标的热流体动力学模拟。物理验证的结果表明，与等效的均质化相比，我们的方法实现了准确性的显着提高只是将颗粒介质简化为整个块的方法。该方法为计算晶粒上的能量沉积问题提供了更准确的替代解决方案。尽管该方法计算量大，但由于GPU强大的并行计算能力，它可以与颗粒动力学模拟同步执行。因此，可以以相对快速和精确的方式进行对光束-晶粒耦合的动力学和热行为的研究。