Development of High-fidelity and Reduced-order Models for Thermal Runaway
PI: Ashish Kumar (Energy and Mineral Engineering)
Funding for the remainder of the junior researcher’s salary will come from Dr. Kumar’s endowment and teaching assistantship within the Department of Energy and Mineral Engineering.
Dr. Ashish R. Kumar, assistant professor of energy and mineral engineering, [webpage] will be the project’s Principal Investigator (PI). Dr. Satadru Dey, assistant professor of mechanical engineering [webpage] will also contribute to the project.
The level of effort appropriate for the proposed project
Funding for two semesters at 50% RA is requested. This will enable the development of reduced order mathematical models and their validation with high-fidelity data.
Plan for funding tuition (for graduate students) or the rest of the researcher’s salary (for postdocs or research faculty)
Funding for the remainder of the junior researcher’s salary will come from my endowment and teaching assistantship within the Department of Energy and Mineral Engineering.
A brief (up to 1 page) description of the proposed project
There is an immediate need to enhance the production of critical minerals due to the lack of their substitutes in modern devices such as electric vehicles, photovoltaic units, items of national security, and many others [1; 2]. These minerals will be mined from surface and underground mining operations. Underground mines traditionally use diesel-operated equipment that releases particulate matter into the mine atmosphere. Mines have therefore started exploring the option of employing battery-powered vehicles (BEVs). Although considered safe under nominal working conditions, large-format batteries used in BEVs can fail through a violent combustion process called thermal runaway (TR). TR leads to a rapid release of combustion products, mostly flammable and toxic, in the confined underground environment.
Creating high-quality multiphysics computational fluid dynamics (CFD) models and reducedorder models (ROMs) will facilitate a deeper understanding of how combustion products spread [3; 4]. Open source and commercial CFD codes such as OpenFOAM, ANSYS Fluent, and Fire Dynamics Simulator (FDS) are used to develop CFD models [5]. Software such as MATLAB and Python could be used to develop ROMs [6]. While CFD models are computationally expensive and require large computational resources (days to weeks of run time on power workstations/ ICDS facilities), results from ROMs can be obtained fairly quickly (within 30.0 mins. on a standard workstation). Due to the inherent design of ROMs being optimized for speed, they incorporate various approximations. Therefore, they need to be compared to the high-fidelity CFD models for their accuracy.
To obtain precise results from these models and make critical decisions when applying them to underground mining equipment, it is essential to ensure they correspond with the outcomes of CFD simulations. Fig. 1 shown here compares the results from numerical models of a combustion event of a large-format lithium-ion battery in a tunnel. The CFD models and ROM show unique characteristics. Frequently, they could differ significantly in terms of spatio-temporal resolution. Variations in temperature, smoke, and other species forecasts from these models may have considerable consequences for the safety of miners operating in the confined areas of these mines. Therefore, the development of calibrated numerical models that can be validated with experiments wherever possible will be crucial to enhancing the safety of underground mining operations and alleviating the exposure of miners to hazardous combustion outcomes.
A list of specific computational and/or data science skills that the current team is particularly interested in recruiting to support the project
Since the research involves modeling transient-state combustion, the team is interested in recruiting a researcher with experience in machine learning methods applied to time series data.
Any other requirements or expectations of potential ICDS Junior Researchers
Any researcher with experience running CFD simulations on parallel cores on ICDS or similar facilities will be preferred. Knowledge of OpenFOAM or any open-source CFD software is a strong point.
A list of specific objectives for work supported by this call
(i.) I plan to start preparing for an NSF-CAREER grant next year. This call will enable me to generate preliminary data for the proposal.
(ii.) The scientific findings obtained from this proposal will be submitted to journals.
At least one medium to long-term goal
I intend to use this call to generate preliminary data for large proposals such as the NSF-CAREER and others. This will also enable me to strengthen my research laboratory.
A short statement explaining the connection of the project to ICDS’s mission
Penn State ICDS’s mission is to catalyze multidisciplinary and interdisciplinary research for society. The proposal draws faculty members from two colleges to research thermal runaway in large-format batteries. The results of this study will make a substantial contribution to the existing knowledge regarding the improvement of large-format battery safety as their applications continue to grow.
A paragraph summarizing team member’s recent and/or planned engagement with ICDS
I used ICDS to develop CFD models for a diesel-pool fire event. The research findings were published in a journal. Currently, my Ph.D. student uses COMSOL Multiphysics software for her combustion research. I also co-advise an MS student who uses ICDS to develop CFD models, including RANS, DES, and LES, to mimic the operations of a powered respirator in a tunnel. I plan to continue to use ICDS facilities for my research on large-format lithium-ion battery safety.