NETL’s Historically Black Colleges and Universities and Other Minority Institutions (HBCU-OMI) program has enabled more than 40 groundbreaking energy research projects since 2010. Two such projects, which were selected under the most recent University Training and Research funding opportunity announcement, have the potential to bolster NETL’s world-renowned Multiphase Flow with Interphase eXchanges (MFiX) software suite through the development of machine learning (ML) and artificial intelligence (AI) techniques for computational fluid dynamics code (CFD).
“MFiX is the world’s leading open-source design software for multiphase flow systems,” said Sydni Credle, NETL’s technology manager for University Training and Research. “The software is continuously being updated, and emerging ML and AI techniques hold the promise of enabling more accurate simulations and faster development of clean energy technologies.”
Teams of professors and students at the University of Texas at San Antonio (UTSA) and Florida International University (FIU) have begun work that aims to improve the predictive capabilities and reduce the computational uncertainty of multiphase numerical codes such as the MFiX code. To this end, the teams are working to provide general and accurate methods for the determination of the drag coefficient of assemblies of non-spherical particles, which better represent the shape of real-world particles.
Teams from both universities plan to conduct numerical computations with a validated CFD code and use AI and ML methods to develop artificial neural networks. The models will be implemented in Google’s TensorFlow and linked to the MFiX code.
“Accurate prediction of air-particle multiphase flow is very important for developing more efficient and cleaner fossil energy technologies,” said Charlie Lin, Ph.D., associate professor at FIU and principal investigator on the project. “We are very excited to have the opportunity to make a contribution in multiphase flow modeling by developing a general drag model using emerging machine learning techniques.”
In addition to improving NETL’s MFiX code, the projects will help to educate and train several graduate and undergraduate students in the science of multiphase flow and the use of in-house CFD codes, the MFiX code and TensorFlow.
“The DOE funding of HBCU-OMI projects has afforded me the opportunity to refine and define my particular focus in the wide world of mechanical engineering,” said Daniel Hinojosa, a UTSA graduate student. “While uniquely honing my research and AI development skills, our project will improve lives in our local, national and global community.”
The UTSA and FIU projects are not the first time a university partnership has been leveraged to improve NETL’s CFD software suite. An earlier version of MFiX was released in September of last year that included an improved modeling capability that allows for more accurate descriptions of real particle-size distributions, offering an important new tool for designing next-generation energy systems to power the nation. The new capability was developed as part of a collaboration with Arizona State University — another minority serving institution.
“HBCU-OMI collaborations are valuable aspects of our research and development planning,” Credle said. “It’s a real win-win situation because the federal funding these institutions receive is enabling segments of the college population, not typically tapped for educational research in these subject areas, to advance their technical skills and academic training. At the same time, the Lab benefits from the results of these projects, such as a more robust CFD code, which, in turn, will benefit the nation through improved design of clean energy technologies.”
The U.S. Department of Energy’s National Energy Technology Laboratory develops and commercializes advanced technologies that provide clean energy while safeguarding the environment. NETL’s work supports DOE’s mission to ensure America’s security and prosperity by addressing its energy and environmental challenges through transformative science and technology solutions.
Posted with permission from netl.doe.gov.
