NCSI: An Open-Source High-Performance Computational Framework for Thermal-Hydrological-Mechanical-Chemical Modeling and Data Analytics

Status	Completed
Mentor Name	Anjali Sandip
Mentor's XSEDE Affiliation	Research Allocation
Mentor Has Been in XSEDE Community	Less than 1 year
Project Title	An Open-Source High-Performance Computational Framework for Thermal-Hydrological-Mechanical-Chemical Modeling and Data Analytics
Summary	In the last two decades, thermal-hydrological-mechanical - chemical (THMC) codes have been widely used in simulating subsurface engineering problems but they lack tools for uncertainty quantification and optimization (i.e. data analytics). The objective of this project is to develop an open-source computational framework that would integrate these tools with THMC modeling. Open-source software programs, OpenGeoSys - for THMC modeling - and Sandia Dakota - for optimization and uncertainty quantification â will be coupled. Furthermore, the coupled software would be successfully applied to benchmarks. The computational framework developed can be applied to solve a wide range of subsurface problems, including, carbon dioxide sequestration, geothermal heat extraction, geological disposal of nuclear waste, etc.
Job Description	- Familiarize with Sandia Dakota program (i.e. data analytics toolbox) - Apply programming skills (C++/Fortran/Python/Shell) to develop an algorithm that will integrate Sandia Dakota into OpenGeoSys - Apply data analytics (optimization, uncertainty quantification, sensitivity analysis, etc.) to benchmark multi-physics/phase problems - Data visualization using Matplotlib - Verify and validate interface developed
Computational Resources	- Dakota (available on Stampede 2) will be interfaced with OpenGeoSys (OGS). - Python, intel and gateway-usage-reporting will be used for developing the algorithm - The open-source framework developed in this study would be applied to benchmarks. These benchmark simulations would be computationally intensive as it involves applying data analytics (iterative analysis) to multi-physics/phase problems. To complete these simulations in a reasonable time, a multi-core processor with high performance computing capabilities would be required. Stampede 2 will be requested for this study. *OpenGeoSys will be installed in the directory space allotted.
Contribution to Community
Position Type	Apprentice
Training Plan	Skills and experience necessary to contribute to the project - 1. Programming (C++/Fortran/Python/Shell) 2. Basic understanding of optimization/uncertainty quantification/sensitivity analyses 3. Familiarity with Linux operating system *I would require assistance developing a training plan.
Student Prerequisites/Conditions/Qualifications
Duration	Summer
Start Date	05/25/2020
End Date	07/31/2020