利用并行八叉树划分技术开展复杂构型航天器放气污染分析研究

Parallel and octa-tree-based acceleration of outgassing contamination simulation for spacecraft of complex geometries

放气污染对航天器造成的有害影响需要定量评估。放气产物分子的输运过程可以采用Monte-Carlo方法模拟,并用射线表示放气分子输运轨迹。如此一来,即转换为射线追踪问题。由于航天器构型复杂,射线追踪的计算很耗时,所以有必要研究其加速算法。文章将复杂构型航天器表面用三角形非结构网格表示,利用八叉树划分技术开展分子运动轨迹与表面相交的加速模拟计算。同时,在基于共享内存多核计算平台,利用OpenMPAPI软件实现了计算的并行化。该方法在对空间站的实例计算中显著提高了运算速度,并对空间站的污染情况进行了初步评价分析。

The organic materials used in spacecraft may outgas under the space environment, including, mainly, the vacuum, the high & low temperature, and the influence of the ionization irradiation. The outgassing product has harmful effects on the spacecraft itself. So a quantitative assessment of the contamination effects is necessary. The molecular transport process is modeled by the Monte-Carlo method, with the rays representing the molecule trajectories, tosimulate the motion process of a large number of molecules. Thus the transport process is transformed into a ray tracingproblem. Due to the complexity of the spacecraft configurations, the ray tracing computation might be very timeconsuming, therefore, it is extremely necessary to have an accelerating algorithm. In this paper, the complex geometryof the spacecraft surface is modeled with triangular unstructured grids, the octa-tree space subdivision technology isused to accelerate the molecular motion and surface intersection calculations. At the same time, based on the sharedmemory multi-core computing platform, the OpenMP API is used to implement parallel computing for the simulation.The test case shows a significant acceleration in the calculation, and the developed method is applied to the contamination analysis of a space station.