We study Saffman-Taylor flow in the presence of intermediate noise numerically by using both a boundary-integral approach as well as the Kadanoff-Liang modified Diffusion-Limited Aggregation model that incorporates surface tension and reduced noise. For little to no noise, both models result reproduce the well-known Saffman-Taylor finger. We compare both models in the region of intermediate noise where we get occasional tip-splitting events, focusing on the ensemble-average. We show that as the noise in the system is increased, the mean behavior in both models approaches the $cos^2(pi y/W)$ transverse density profile far behind the leading front. We also investigate how the noise scales and affects both models.
Download