INVESTIGATION OF TRANSPORT IN COMPLEX FLOWS (RENEWAL OF CTS090025)

Dimitrios Papavassiliou, University of Oklahoma

0000-0002-4583-0820

ACCESS Allocation Request CTS080042

Abstract:

This is a proposal to renew our allocation for the computational investigation of transport phenomena in complex flows. The focus will be on (a) the interfacial behavior of nanofluids in the presence of surface-active agents, and (b) the hydrodynamic dispersion in porous media. 1) Interfacial stability: Fluid/fluid interfaces (e.g., air/water and oil/water interfaces) are common in applications such as cosmetics, pharmaceuticals, and the food industry. Surface-active agents, such as nano-sized particles (NPs) or surfactant molecules, when present in these systems, act as stabilizers of the interface. We consider the behavior of oil drops migrating in a nanochannel with both surfactants and NPs, and the behavior of the interfaces under compression, since the stability of drops depends on the behavior of the stabilizers under flow. Understanding the synergistic effects when both stabilizers are present is significant for tuning interfacial behavior. The research questions for this project are 1) Can NPs mobilize oil in the subsurface for oil extraction? 2) How do the NP shape or the NP surface chemical anisotropies affect the interfacial stability of oil-water interfaces? 3) What is the rheological behavior of an oil-water-particle-surfactant system under shear and how can we model it? Methodology: Dissipative particle dynamics (DPD) simulations The simulation approach is based on dissipative particle dynamics (DPD), which is a coarse-grained method to study complex systems with large molecules like surfactants. The advantages of DPD are that longer length and time scales are obtained when compared to atomistic molecular dynamics simulations, while maintaining the correct hydrodynamic behavior of the system. Our DPD simulations utilize the open-source software package LAMMPS, developed and maintained by Sandia National Laboratories. 2) Dispersion in porous media: We have developed novel models for the prediction of the dispersion and agglomeration of nanoparticles in porous media that are based on the Lagrangian time and length scales of the particles in flow. We will extend these computations to investigate the transport of nanoparticles in a catalytic reactor (i.e., a porous medium made of spherical particles) to produce hydrogen from methane. We focus on catalyst-assisted methane pyrolysis, where methane decomposes directly into hydrogen and solid carbon (without carbon dioxide production). This technology requires modeling of the carbon particle byproducts in the hydrogen production reactor and the modeling of the transport of hydrogen. The Research Questions are 1) How do solid carbon particles move in the reactor and how can our Lagrangian modeling be used to predict this movement? 2) How can dimensionless numbers that characterize the flow and the particles, like the Peclet number and the Stokes number, be used to predict the flow patterns in the reactor and maximize hydrogen production? Methodology: Lattice Boltzmann methods (LBM) and Lagrangian particle tracking (LPT) We use lattice Boltzmann methods (LBM) to simulate flows in porous media and we combine it with the Lagrangian particle tracking (LPT) methodology developed in our lab. We developed in-house a Fortran 90 code that has been parallelized with MPI. The LBM code has been validated using cases with known analytical solutions for the fluid velocity field. We model mass transport using Lagrangian particle tracking (LPT), instead of using conventional Eulerian models. The LPT method traces the trajectories of transported quantities through the flow field. Particle motion is composed of a convective part (obtained using an interpolation of the velocity field from the LBM simulations) and a diffusion part (i.e., Brownian motion obtained from a Monte-Carlo approach). In LST, computational efficiency is accomplished by releasing large numbers s of different markers with different molecular diffusivity in the same hydrodynamic field.

---

Allocations:

2025	TACC Dell/Intel Sapphire Rapids, Ice Lake, Skylake (Stampede3)	23,479.0 Node Hours
The estimated value of these awarded resources is $4,695.80. The allocation of these resources represents a considerable investment by the NSF in advanced computing infrastructure for the U.S. The dollar value of the allocation is estimated from the NSF awards supporting the allocated resources.
2024	TACC Dell/Intel Sapphire Rapids, Ice Lake, Skylake (Stampede3)	25,826.0 Node Hours
2024	TACC Long-term tape Archival Storage (Ranch)	62.0 GB
The estimated value of these awarded resources is $5,168.30. The allocation of these resources represents a considerable investment by the NSF in advanced computing infrastructure for the U.S. The dollar value of the allocation is estimated from the NSF awards supporting the allocated resources.

Click to show/hide prior allocations »

---

Other Titles:

Click to show/hide prior titles »