From Particles to Landforms: Integrating Theory, Computation, Experiments and Field Data to Overcome Empiricisms
Abstract
Particulate systems are ubiquitous in many applications of practical relevance, from geomechanics to additive manufacturing and on to terradynamics. To address these applications, today s solution approaches embrace either a fully resolved representation of the system, which leads to highly accurate but prohibitively expensive solutions, or empirical simplifications that lead to expeditious solutions that have a limited and not clearly understood domain of validity. This research effort will investigate modeling techniques, numerical solution methods, and software designs in order to get the best of both worlds; i.e., develop highly accurate methods that produce results fast. The overarching technical aim of this project is essentially two-fold: (1) to develop a rigorous modeling framework for dry and wet systems that is general enough such that boundary conditions and particle-level information is everything that one needs in order to predict system-state evolution in problems from the particle scale up to the largest scales, and (2) to develop innovative and expeditious computational tools that extend the limits of the state-of-the art in terms of system size and phenomenology, from the discrete and continuum perspectives. These aims are intricately connected; fast particle-by-particle simulations in (2) provide data to hone and test the upscaled modeling ideas in aim (1), and models that emerge from (1) will be migrated into numerical implementations in (2) using innovative computational and software approaches. In pursuit of predictive models, we will study how particle details such as shape, internal breakage mechanics, and surface friction translate into continuum models for yielding, dilation, fabric anisotropy, rate effects, and pore fluid dynamics. At the small (particle) scale, we will develop methodologies for fully resolved, discrete-element method simulations accounting for grain size distribution, shape, and breakage as well as pore fluid. Through high-performance-computing developments, we will push the envelope on the size of these simulations, up to billions of interacting bodies, allowing our simulations to reflect conditions of experiments. An iterative model-simulate-validate loop will be carried out using smaller scale simulations to test upscaled continuum-level hypotheses. Moreover, through simultaneous verification of models, we will merge computational models at disparate scales in an effort to maximize efficiency in predicting certain processes, such as size-segregation in highly polydisperse systems. The outcome of these efforts will be a comprehensive, predictive model for dry and fully-saturated particulate materials that can connect particle-scale physics to large-scale flow phenomena. The modeling vision outlined above will advance the state of the art in particulate system simulation technology from several perspectives: computational science (large data sets, high-performance computing platforms, use of hardware accelerators), fidelity (particle morphology, breakage modeling), multi-scale algorithms (hybridized discrete/continuum approaches for fine-particles), and numerical methods (improvements to the material-point method, differential variational inequality approaches, fluid-solid interaction). Through its goal of promoting a physics-based understanding of emergent behavior in particulate systems, this project stands to improve the community s ability to predict, interpret, and control behavior in scenarios ranging from lab-scale to field-scale. Given that particulate systems are ubiquitous, understanding their dynamics is poised to be relevant in a range of practical applications such as additive manufacturing, nanoparticle self-assembly, composite materials, pyroclastic flows, formation of asteroids and planets, meteorite cratering; and also in industries such as pharmaceuticals, chemical and biological engineering, food processing, farming, manufacturing, construction and mining.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 19, 2019
- Source ID
- W911NF1910431
Entities
People
- Dan Negrut
Organizations
- Army Contracting Command
- United States Army
- University of Wisconsin–Madison