New Interface Reconstruction Method for Multi-Fluid Flows Developed as
Part of LANL ASC Strategic Research
Los Alamos researchers Rao Garimella, Sam Schofield, and Marianne Francois, along with Raphael Loubere from Universite Paul-Sabatier, France, have developed a new interface reconstruction method for multi-fluid flows that is completely material-order independent. This is in contrast to traditional interface reconstruction methods, which produce very different results depending on the order in which materials are processed, leading to errors in advection of the materials.
The new method reconstructs the interfaces in three steps. First, the relative locations or approximate centroids of the materials are determined using a linear reconstruction of the volume fraction function over multi-material cells. Next, the approximate centroids of the materials are used to construct a weighted Voronoi or Power diagram subdivision of the cell while exactly matching the input volume fractions. Finally, the interface segments are smoothed using a constrained optimization process in order to minimize the slope discontinuity with interfaces in the neighborhood.
The method shows marked improvement in the reconstruction of multi-material interfaces compared to order-dependent methods. It can reconstruct straight lines exactly, and initial experiments indicate that it is second-order accurate. Finally, it also shows improved accuracy in dynamic advection tests of multi-material shapes. A preliminary implementation of the method is in LANL ASC Code
Project A.
  
Diagonal advection of a four-material bubble on a 40x40 grid with a velocity of (1.1,1.1): (a) initial configuration, (b) results of
order-dependent method after 100 timesteps, and (c) results of new
order-independent method with smoothing.
|