U.S. Naval Research Laboratory (NRL) research engineer, Dr. John Steuben, is awarded the prestigious American Society of Mechanical Engineers (ASME) 2015 Best Ph.D. Thesis of the Year Award for his dissertation entitled Massively Parallel Engineering Simulations On Graphics Processors: Parallelization, Synchronization, and Approximation.
Since 2013, the ASME Computers and Information in Engineering (CIE) Division has presented the award in recognition of promising young investigators who authored the best Ph.D. thesis of the year in the area of computers and information in engineering.
Receiving a master's in mechanical engineering from Colorado School of Mines in 2011, Steuben continued his doctoral studies at Mines under the guidance of Dr. Cameron Turner, Associate Professor of Mechanical Engineering. In 2012 he began participating in the
Naval Research Enterprise Internship Program (NREIP) under the supervision of Dr. John Michopoulos, head of the Computational Multiphysics Systems Laboratory (CMSL) at NRL.
During this tenure, Steuben contributed in the development of a fast characterization approach that reduced the computing time required to characterize material specimens (tested in NRL's six degree-of-freedom automated test frame) from 24 hours to approximately 30 seconds. This research, hosted through a NREIP internship by NRL at CMSL, continues through present day and is devoted to the development of surrogate model approaches for the characterization of composite materials.
Graduating from Mines in 2014 with a doctorate in engineering systems, Steuben's dissertation examined three computer-aided engineering (CAE) methods; boundary element method (BEM), discrete element method (DEM), and finite element method (FEM) for particular applications. These three methods are crucial elements of the core research upon which CIE was founded. Steuben's work led to a general-purpose computing on graphics processing unit (GPGPU) parallelized BEM solver, a GPGPU parallelized DEM model of icing effects of wind turbine blades incorporating frictional, thermal and phase change effects (requiring synchronization and parallelization), and a surrogate approximation enhanced FEM approach (parallelization, synchronization and approximation) to solving the inverse material characterization problem for non-isotropic composite materials including those being tested at NRL.
Dr. Steuben's research thesis was that parallelization must be considered in terms of both synchronization and approximation, said Turner. As his work concluded on his dissertation, he proceeded to formulate extensions of his research into characterizing the as-built parameters of additively manufactured components using a combination of his DEM simulations and his FEM experience with composite materials.
Since January 2015, Steuben has been conducting postdoctoral research at NRL with funding from the
National Academies of Sciences, Engineering, and Medicine. This work has focused on the development of computationally efficient multiphysics simulations of additive manufacturing processes and components. This research will allow the design and fabrication of novel, useful, and efficient structures and components across a wide range of engineering disciplines.