Multiphysics Rainstorm Simulation Methodology Gm Brazil

Format:

Conference/Event

Author/Creator:

Lelis Alves, Lelis Alves, author.

Contributor:

Buscariolo, Filipe Fabian

Della Volpe, Leonardo José

Magazoni, Felipe

Maruyama, Flavio Koiti

Conference Name:

26th SAE BRASIL Inernational Congress and Display (2017-11-07 : Sao Paulo, Brazil)

Language:

English

Physical Description:

1 online resource

Place of Publication:

Warrendale, PA SAE International 2017

Summary:

AbstractThe advances in High Performance Computing-HPC and CPU's sizes and processing power, combined with new computational codes, which are capable of coupling different types of simulation, are the main contributors for the increasing number of the Multiphysics simulations inside the industry.Multiphysics are defined as simulations involving multiple physical models or multiple simultaneous physical phenomena. Among some examples, spray modeling is of great interest in several branches of the industry and, with the development on algorithms and codes, simulations presented reliable results, compared to experiments.This work aims to contribute to both Multiphysics and spray modelling by reproducing a rainstorm condition, focused on a vehicular application. This work objective is evaluate the possibility and feasibility to reproduce virtually rain storm condition on a highway checking water intrusion at Air Intake System - AIS and compare with physical test. Physical test has sprays pointed to the front of right and front tire, reproducing the water splash from the highway under rain stormThe results show that the proposed methodology is able to predict the water behavior and trajectory over the car.Considering the presented achievements, it is possible to apply this methodology at early stages of vehicle development in order to predict the system performance to avoid late project changes, which influence both time and costs

Notes:

Vendor supplied data

Publisher Number:

2017-36-0085

Access Restriction:

Restricted for use by site license