Direct Numerical Simulation of Diesel Jet Flame Stabilization at High Pressure

PI: Jackie Chen, Sandia National Laboratories
Code: S3D
Allocation: 50 million hours

As the supply of traditional fossil fuels such as gasoline and diesel continues to decline, new fuels and new engines must be developed. A team led by Jacqueline Chen of Sandia National Laboratories is using Oak Ridge National Laboratory’s petascale Jaguar supercomputer to explain how jet flames in diesel engines ignite and become stable at high pressure. By simulating the workings of these flames under the intense pressure encountered in diesel engines, the team will deepen our knowledge of this process and aid efforts to develop new, efficient, and less-polluting fuels. For this research, the increase in computational power to petascale allows for an increased grid size and more time steps in the simulation. The research team will use S3D, a massively parallel direct numerical simulation solver, to make fundamental observations of the fine-scale interactions between turbulence and chemistry in combustion. The team will perform direct simulations over several decades of scales to provide detailed scientific understanding and develop models that encapsulate the understanding to bridge the different ranges of scales. The objective is to provide quantitative physical understanding of the underlying issues that drive combustion under low-temperature, high-pressure environments for diverse fuels.

Direct numerical simulation of a lifted autoigniting ethylene-air jet flame. Volume rendering of formaldehyde (blue/green) and hydroxyl radical (red/white) denoting the location of preignition upstream of flame and lifted flame, respectively. Courtesy H. Yu of Sandia and K. L. Ma of SciDAC Institute for Ultrascale Visualization.
Direct numerical simulation of a lifted autoigniting ethylene-air jet flame. Volume rendering of formaldehyde (blue/green) and hydroxyl radical (red/white) denoting the location of preignition upstream of flame and lifted flame, respectively. Courtesy H. Yu of Sandia and K. L. Ma of SciDAC Institute for Ultrascale Visualization.