Project No: FE0006932
Performer: Princeton University
Robert Romanosky Crosscutting Research Technology Manager National Energy Technology Laboratory 3610 Collins Ferry Road P.O. Box 880 Morgantown, WV 26507-0880 304-285-4721 firstname.lastname@example.org Patricia Rawls Project Manager National Energy Technology Laboratory 626 Cochrans Mill Road Pittsburgh, PA 15236-0940 412-386-5882 email@example.com Sankaran Sundaresan Principal Investigator Princeton University Department of Chemical Engineering Princeton, NJ 08544 609-258-4583 firstname.lastname@example.org
DOE Share: $300,000.00
Performer Share: $120,366.00
Total Award Value: $420,366.00
Performer website: Princeton University - http://www.princeton.edu
This project will use a combination of continuum simulations for model validation and discrete particle simulations for model refinement. The first phase will be immediate implementation into Multiphase Flow with Interphase eXchanges (MFIX)—a general-purpose computer code developed at NETL—of a steady-shear, continuum rheological model for dense granular flows developed recently by the research group. The researchers will perform hopper, bin, and chute flow simulations using MFIX, with the results compared to literature data. While the continuum simulations of the first phase will employ a formulation of boundary conditions that already exists in MFIX, the second phase of this project will aim to develop and implement a more rigorous treatment of boundary effects. These will be investigated via simulations using discrete element method (DEM) implemented in the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) developed by Sandia National Laboratories. The simulations will focus on simple shear flows of identical mono-disperse spheres, and will be performed on both a traditional computer cluster as well as a specialized graphics machine, which offers significant increases in the size of the systems that can be simulated. During the final phase, the team will connect macroscopic and microscopic descriptions of the flow, thereby allowing for the construction of a more detailed constitutive model for the particle-phase stress. Each model refinement will be implemented into MFIX and validated by simulating various test cases that include chute and hopper flows as before, as well as simple and oscillatory shear in a Couette cell.
Program Background and Project Benefits
Dense granular flows are ubiquitous in both natural and industrial processes. They manifest three different flow regimes—commonly referred to as the quasi-static, inertial, and intermediate regimes—each of which exhibits its own dependencies on solids volume fraction, shear rate, and particle-level properties. The differences in these regimes can be attributed to microscale phenomena, with quasi-static flows being dominated by enduring frictional contacts between grains, inertial flows by grain collisions, and intermediate flows by a combination of the two. Existing constitutive models for the stress tend to focus on one or two regimes at a time. Recent research at Princeton University (Princeton) has centered on a rheological model for dense granular flows that captures stresses in all three regimes under steady-shear conditions and the transitions among them. Since the inception of the Department of Energy (DOE) National Energy Technology Laboratory (NETL) University Coal Research (UCR) Program in 1979, the primary objectives have been to (1) improve understanding of the chemical and physical processes involved in the conversion and utilization of coal in an environmentally acceptable manner; (2) maintain and upgrade the coal research capabilities and facilities of U.S. colleges and universities; and (3) support the education of students in the area of coal science. As part of the UCR Program, NETL has partnered with Princeton in a project that will continue and advance the development of dense granular flow simulations to enable better understanding of the chemical and physical processes involved in the conversion and utilization of coal. The nonlinear continuum model developed at Princeton predicts stresses in all three of the aforementioned dense granular flow regimes and is a promising candidate for predicting flow behaviors in industrially relevant flows, which will enhance the ability to interrogate the dense phase flow behavior in large-scale processes. This research in combination with other modeling efforts are targeting improved accuracy and predictive capability of system and flow behavior which will lead to better process design and reduced cost associated with scale up advanced technologies that include complex flow regimes. Goal and Objectives
The goal of this project is to implement and validate a new rheological model for dense granular phase in MFIX while continuing to improve it to capture more complex flow behavior. Specific objectives planned to accomplish this goal include (1) implementing in MFIX the steady-shear rheological model developed; performing MFIX simulations of various test problems such as hopper, bin, chute, and Couette flows using this rheological model in conjunction with the no-slip and partial slip boundary conditions already available in MFIX; and comparing the results against experimental and DEM simulation data; (2) developing improved wall boundary conditions for the particle phase that can be applied in all three regimes of flow and implementing them in MFIX; examining the effect of refined boundary conditions on flow characteristics in the test problems previously mentioned; and (3) developing the rheological model to allow for dynamic evolution of the stresses, making appropriate modifications to the MFIX implementation, and conducting appropriate validation tests.
Princeton University tested their steady shear rheological model on the problem of flow in an hourglass, a commonly-modeled type of hopper, using MFIX and compared it with MFIX’s default model. In both cases, the gas phase was considered, and no-slip boundary conditions were used for the solids. The new rheological model predicts ratholing in the upper chamber of the hourglass as well as the formation of a sandpile on the hourglass floor, both of which are consistent with experimental observations.