基于SPH方法和DPD方法的聚合物充模问题的宏-介观多尺度并行算法研究

The filling process is the most complex and most important stage of the injection molding. And the mesoscopic characters are the direct factor that affects the flow and physical property of the fluid in the injection molding. So it is important and has profound guiding significance for the polymer processing to study the macroscopic flow and the mesoscopic characters in the filling process. However, the macro-micro methods based on the grid-based methods ignore the effect of the mesoscopic information on the macroscopic flow, and limit the scale of the simulation. Therefore, this project focuses on studying a macro-mesoscopic multi-scale algorithm based on the smoothed particle hydrodynamics (SPH) method and the Dissipative particle dynamics (DPD), thus the macroscopic flow and the mesoscale information can be simultaneously obtained using by the proposed multi-scale algorithm. A corrected SPH method which can accurately solve the pressure without the pressure Poisson equation is first developed. Then the way to transfer the information between the macroscopic SPH particle and the mesoscopic DPD particle is investigated according to the commonalities of the SPH method and the DPD method. At last, based on a quickly particle searching-technique, the parallel computational technique is investigated to improve the computational efficiency, the macro-mesoscopic multi-scale parallel algorithm of the filling process for the polymer melt is set up, and the macroscopic flow and the mesoscopic characters in the filling process are simulated and predicted. This project will solve the problem that is difficult to simulate using by the for the mulstiscale method given by the commercial software or the related references, and lays the foundation for the integration researches of the polymer processing and the prediction of the properties of products.

充模过程是注塑成型中最复杂、最重要的阶段，而介观信息是影响充模过程的复杂流动性能及物理特性的直接因素。因此，研究充模过程中的宏观流动和介观特性对聚合物材料加工具有重要的指导意义。然而以往基于网格类方法的宏-微观多尺度方法不仅限制了模拟的尺度，而且忽略了介观信息对宏观流动的影响。因此，本项目拟基于宏观SPH方法和介观DPD方法研究聚合物充模问题的宏-介观多尺度并行算法。首先发展一种不需要求解压力Poisson方程且能准确求解压力的改进SPH方法。其次利用SPH方法和DPD方法的共性， 研究宏观SPH粒子和介观DPD粒子之间信息传递的途径。最后给出一种快速粒子搜索技术，建立聚合物充模问题的宏-介观多尺度并行算法，模拟和预测聚合物充模过程中复杂的宏观流动和介观分子演化过程。本项目的研究成果将解决目前商用软件及文献中多尺度算法难以模拟的问题，为聚合物动态成型过程与制品性能预测的一体化研究奠定基础。

注塑成型是塑料制品生产的主要方法，充模过程是决定塑料制品质量的重要阶段。 在充模过程中，介观分子演化过程决定了塑料制品的最终物理特性。因此，充模过程中宏-介观流动对注塑成型过程有重大影响。该项目主要研究聚合物充模过程中基于SPH方法和DPD方法的宏介观多尺度算法、模拟及应用。首先，将SPH方法和FPM思想相融合，得到了不需要求解压力Poisson方程，且具有较高数值精度和较好数值稳定性的改进SPH算法。其次利用SPH方法和DPD方法的共性，借助握手区内粒子位置相同的特点和商业软件Lammps，给出了SPH粒子与DPD粒子信息交换的方法。最后，建立聚合物充模问题的宏-介观多尺度算法，模拟和预测关注区域内复杂的宏观流动和介观分子演化过程。本项目的研究成果为聚合物动态成型过程与制品性能预测研究提供了理论支持。