Please login to view abstract download link
We developed a parallel MPS code for large-scale jetting spray problems. As the free surface and large deformation take the most important part in this problem, Finite volume with level-set, VOF, and front-track methods are often used for interface capture to solve the problem, but these methods have problems such as the inability to distinguish small structures and momentum dissipation. Using the particle method for solving can not only take advantage of Lagrange characteristics, to resolve small-scale droplets and reduce the computation cost by avoiding discretization of a large space but also avoid the dissipation caused by the convective term, thereby improving the simulation accuracy. In order to improve the computational resolution to meet industrial needs, we developed MPI parallel code based on dynamic district decomposition, which enables large-scale particle parallel solving to accurately calculate the droplet spraying phenomenon. Also, an advanced adaptive variable size particle method is applied to provide high-resolution results and ideal efficiency for large-scale simulations. By comparing with the experiment, the simulation accurately reproduces the droplet spraying phenomenon after the jet collision. Based on our statistical distribution of different droplet sizes through post-processing methods, it is reasonable to assume that the results of the numerical experiment can predict the spraying phenomenon in actual applications.