Dr. Anthony Wachs: Combining fluid dynamics and high-performance computing

Combining fluid dynamics and high-performance computing

Anthony is no stranger to high-performance computing; he was introduced to it through work in France in the early 2000s. Today, his work is entirely computational. His lab uses high-performance computers to run advanced fluid dynamics calculations. 

As part of Anthony’s work, he also belongs to a research group called the Complex Fluid Lab. The group conducts interdisciplinary research involving fluid mechanics, often with strong industrial or applied motivations. Anthony’s group uses high-performance computing to solve partial differential equations on a grid that are representative of fluid flows. The work is distributed among many processors trying to solve equations as fast and accurately as possible. High-performance computing has made it possible to solve complex fluid dynamics problems in a relatively short amount of time when previously, these computations were either not possible or could take up to decades to compute. One can think of high-performance computing as being like a car and fluid dynamics as its passenger: high-performance computing enables the journey toward solving equations. 

Today, roughly one third of Anthony’s computing allocations are on UBC Advanced Research Computing’s Sockeye platform. He receives additional computational support through allocations at Simon Fraser University’s Cedar, the University of Toronto’s Niagara, and Compute Canada. 

Anthony research interests are in multi-phase flows and flows of non-Newtonian materials. Most of his multi-phase flow research focuses mostly on particle-laden flows. These flows can be thought of as a continuous fluid seeded with particles. Most of Anthony’s students are using supercomputing to solve basic conservation equations on a grid. Doing so requires a lot of special resolution and many degrees of freedom. The net result is published research papers and projects supported by industrial partners.

One way that Anthony uses high-performance computing is to understand what happens to the dynamics of particle-laden flows if particles are not spherical, which is not typically expected. It is a critical area of study because the use of these particle-seeded fluids (called suspension) is ubiquitous, and in real life particles are seldom perfectly spherical. This has practical applications, for example, in the case of understanding how sediments are transported from a river to an estuary or of determining the efficiency of chemical engineering processes. 

“We can speculate that a particle looks like a sphere and we can model it as a sphere. But can we compute these complex flows of non-spherical particles and what can we infer in terms of new insights, and how can we translate that into a more efficient industrial process?”