High performance computing enabling predictive engine simulations for engine manufacturers

Award Year 

Environmental and natural resource pressures are driving the design of more efficient vehicles and internal combustion engines powered by alternative fuels. The large number of design parameters, combined with new combustion chemistries, is so complex that much more engine development must be done via computer simulation (modeling). In addition, engine combustion concepts and fuels are evolving rapidly, which creates a demand for highly predictive engine-fuel simulation software for both research and industrial design. (more)

Industry seeks the most effective, least costly, and most timely means to conduct research, engineering, and product development. Experimental research and prototyping have become expensive and often time-consuming; hence, there is strong incentive for industry to substitute less expensive modeling and simulation where possible.

Argonne National Laboratory transferred a two-part technology involving vehicle engine cylinder computer simulations: a unique engine fuel spray software model, and advanced computer load-balancing algorithm software technology and best practices that enabled high performance computing (HPC) simulations of combustion occurring in internal combustion engines. The technology enables engine modelers to perform more predictive engine simulations to determine what occurs when fuel enters the cylinder of an engine and is burned.

The technology was transferred to Convergent Science, Inc.; Cummins, Inc.; and Caterpillar, Inc. The fuel spray software model was integrated into Convergent Science?s CONVERGE computational fluid dynamics (CFD) software, a commercial product highly accepted by industry and used for engine modeling. The integration was tested by Cummins, using its products in the model. Argonne, Convergent Science, and Caterpillar used the CFD software plus fuel spray model to conduct simulations of the Caterpillar C15 engine on high performance computing platforms at Argonne.

The technology transfer effort has benefitted all parties with improved engine simulations and best practices for performing these simulations in industry. In particular, the technology has allowed Caterpillar to shrink its development timescales, resulting in significant cost savings. Caterpillar engineers predict that these engine simulation developments will reduce the number of multi-cylinder test programs by at least a factor of two, resulting in hundreds of thousands of dollars in cost savings per year. A further benefit is that insights gained by these advanced simulations will be extended to inform important engine design considerations such as fuel consumption, heat release rate, and emissions (nitrogen oxides and soot), impacting the design of advanced transportation technologies worldwide.