October 1, 2016
Carter Mast Pedro Arduino Peter Mackenzie-Helnwein Gregory Miller
A paper authored by an alum and three UW CEE faculty members has received a best paper prize from the international geoengineering research journal Acta Geotechnica. The paper was one of the most cited papers published in Acta Geotechnica in 2015.
Lead author for the paper “Simulating granular column collapse using the Material Point Method” is alum Carter Mast (Ph.D. ’13), who currently works at Amazon as a software development engineer. Co-authors are geotechnical faculty member Pedro Arduino and structural faculty members Peter Mackenzie-Helnwein and Gregory Miller.
“The paper is the result of excellent cooperation between structural and geotechnical faculty and students,” Arduino said.
The paper presents an analysis of sand column collapse phenomena using the Material Point Method (MPM), which is used to simulate the behavior of materials such as liquids and gases, together with a constitutive model to capture how the sand responds to various loadings. While the technique is not new, this is the first time it has been applied to granular materials, which are difficult to analyze using alternative methods. The researchers conducted a series of numerical simulations of two dimensional sand column collapse configurations and compared their results with other simulations and experimental results, proving the MPM analysis method produced more accurate results.
“Granular materials have historically been challenging to model for a few reasons,” Mast said. “Two seemingly similar samples can behave very differently in the lab under the same loading conditions.”
From a computational standpoint, consistently modeling and capturing the variability of granular materials is challenging. This is because the material parameters are highly variable and are influenced by many factors, such as moisture content. The constitutive model, which links the deformed material to the stress in the material, is also problematic in the context of granular materials as properly capturing variables requires extensive model calibration from lab samples.
“This particular research was a proving ground for the MPM’s ability to accurately model a wide range of granular materials,” Mast said. “The study showed that under the right conditions, MPM does a very good job at accurately capturing flow characteristics including flow dynamics and deposit profile.”
Using this research as a benchmark, future applications may entail utilizing the MPM to model the interaction between a flowing medium and structural support, Mast said, such as a flowing landslide and columns supporting a bridge.