Dynamics Seminar by Lennaert van Veen (Ontario Tech) on August 21

On Wednesday the 21st of August at 4 pm, Lennaert van Veen (Ontario Tech) will give an Amsterdam Dynamics seminar.

The talk will take place in the Maryam seminar room (9A-46).

Title: The confirmation bias and statistics of solutions to the Kuramoto-Sivashinsky equation

Abstract: The formation of interfaces through deposition of particles is a common process in manufacturing, for instance of smart materials. One important quantity in this process is the surface roughness, computed as the second moment of the distribution of surface heights. A plethora of stochastic models of surface growth have been used to predict the scaling of roughness in time. Many of these models fall in the Kardar-Parisi-Zhang universality class, meaning that they exhibit the same scaling exponents after appropriate ensemble averaging, even though the dynamics of individual realizations differ. The KPZ exponents are found in the stochastically driven Burgers equation. A modification of the linear operator in this equation changes it into the Kuramoto-Sivashinsky equation. In the 1980s, Yakhok conjectured that, on large spatial scales, the KS equation behaves like the noisy Burgers equation, the noise being provided by the small-scale dynamics. This would place the KS equation in the small wave number limit in the KPZ universality class. In this work, I will review the pubished numerical work that purports to support the hypothesis and present more comprehensive and accurate data. An objective analysis points at a strong confirmation bias in earlier work and the absence of clear numerical evidence of Yakhot's conjecture.