This page shares the codes for our simulation and deep learning studies. You can also find the research datasets and computational facilities in the SEWES lab.

Codes for simulations and deep learning

HyThermMix:

The simulator for Hydrogen's Thermodynamic properties in Mixtures with other gases (HyThermMix) is an open-source code that provides a thermodynamic modeling tool based on the GERG-2008 equation of state.  The HyThermMix simulator has been tailored for analyzing hydrogen mixtures in various underground hydrogen storage scenarios.  You can download the code at the following GitHub link.  

Citation:

The HyThermMix is an effort developed by Dr. Qingqi Zhao, a former Ph.D. student in the Chen's group. If you use this code in your work, please cite our paper:

Zhao, Q., Y. Wang, and C. Chen  (2024), Numerical simulation of the impact of different cushion gases on underground hydrogen storage in aquifers based on an experimentally-benchmarked equation-of-state, International Journal of Hydrogen Energy, 50, 495-511, https://doi.org/10.1016/j.ijhydene.2023.07.262.

GitHub Repository: View the Code

Research data

Time-series micro-CT data of 3D pore geometry evolution

You can download the 3D micro-CT reconstruction data of the porous media demonstrated in Figures 1 and 2 of our WRR paper (2008WR007252) from this Google drive.  These CT datasets show the time-series evolution of the 3D pore geometry resulting from colloidal zirconia (ZrO2) particle deposition on the surface of glass beads.  The CT pore structural data can be imported into pore-scale numerical simulators (e.g., a lattice Boltzmann model) as interior boundary conditions of fluid-flow modeling to simulate single-phase and multi-phase flows.  Please see details of the micro-CT datasets in the "read-me" file.

This work was supported by Dr. Aaron Packman's NSF award via grant EAR-0310657. The micro-CT scanning was performed at the Northwestern Synchrotron Research Center located at sector 5 of the Advanced Photon Source at the Argonne National Laboratory. 

Citation:

If you use these micro-CT datasets in your work, please cite our WRR paper:

Chen, C., B. L. T. Lau, J. Gaillard, and A. I. Packman (2009), Temporal evolution of pore geometry, fluid flow, and solute transport resulting from colloid deposition, Water Resources Research, 45, W06416, doi:10.1029/2008WR007252.

Computing equipment

The SEWES lab has two customized high-performance workstations equipped with multiple GPU cards for large-scale GPU parallel computing and machine learning training.  

Left: Exxact 4-GPU workstation equipped with four Nvidia A30 24 GB GPUs, two Intel Xeon Ice Lake Gold 5317 CPUs, 512 GB total system memory, and 24 TB hard drive.  

Right: Exxact 4-GPU workstation equipped with four Nvidia A30 24GB GPUs, two Intel Xeon Gold 5218R CPUs, 256 GB total system memory, and 20 TB hard drive.