Using a DNS framework to test a splashed mass sub-model for lagrangian spray simulations

Document Type

Technical Report

Publication Date



Department of Mechanical Engineering-Engineering Mechanics


Numerical modeling of fuel injection in internal combustion engines in a Lagrangian framework requires the use of a spray-wall interaction sub-model to correctly assess the effects associated with spray impinge-ment. The spray impingement dynamics may influence the air-fuel mixing and result in increased hydrocarbon and particulate matter emissions. One component of a spray-wall interaction model is the splashed mass fraction, i.e. the amount of mass that is ejected upon impingement. Many existing models are based on relatively large droplets (mm size), while diesel and gasoline sprays are expected to be of micron size before splashing under high pressure conditions. It is challenging to experimentally distinguish pre- from post-impinged spray droplets, leading to difficulty in model valida-tion. In this study one commonly used splashed mass model, proposed by O’Rourke and Amsden, was studied through Direct Numerical Simulations (DNS) using an in-house 3D multiphase flow solver, which has been validated in previous studies. To test the splashed mass model in scenarios relevant to fuel impingement, input was generated from Lagrangian-Eulerian (LE) diesel injection simulations. A representative combination of droplet diameter and velocity before impact was selected from the LE output. The representative droplet was then used to initialize the DNS calculations, allowing the splashed mass model to be tested for micron-sized droplets. The splashed mass ratio was calculated from the DNS output and was compared to the model of O’Rourke and Amsden. This new LE/DNS framework allows Lagrangian sub-models to be tested in a way that is difficult to examine experimentally and provides unique feedback that will be used to enhance the accuracy of LE spray simulations.

Publisher's Statement

© 2018 SAE International; Argonne National Laboratory. Publisher’s version of record:

Publication Title

SAE International