Methodology Development to Accurately Predict Aerodynamic Drag and Lift for Passenger Vehicles Using CFD. Tata Technologies

Format:

Conference/Event

Author/Creator:

Palaskar, Palaskar, author.

Contributor:

Kumar, Vivek

Vaidya, Rohit

Conference Name:

SAE 2016 World Congress and Exhibition (2016-04-12 : Detroit, Michigan, United States)

Language:

English

Physical Description:

1 online resource

Place of Publication:

Warrendale, PA SAE International 2016

Summary:

AbstractImportant vehicle performance parameters such as, fuel economy and high speed stability are directly influenced by its aerodynamic drag and lift. Wind tunnel testing to asses these parameters requires heavy investment especially when test wind tunnel is not available in the country where vehicle development center is present. Hence to save cost and to compress development time, it is essential to asses and optimize parameters of a vehicle in very early stages of development. Using numerical flow simulations optimization runs can be carried out digitally.Industry demands prediction of aerodynamic drag and lift coefficients (CD,CL) within an accuracy of a few counts, consuming minimal HPC resources and in a short turnaround time. Different OEMs deploy different testing methods and different softwares for numerical simulations. Hence, correlation level for numerical prediction depends on factors viz; type of meshing, boundary conditions, turbulent models, CAD geometry and wind tunnel used for correlation. In order to improve factors like accuracy, resources consumed and turnaround time, optimization of above mentioned factors is required. This work helps in identifying such optimized parameters for commercial RANS software.In our present study ANSYS Fluent RANS solver is used to simulate and develop a methodology for predicting CD, CL values. Predicted CD, CL values are correlated with MIRA tunnel test values. Conventionally simulation tunnel is of the size having 1% blockage. This size can be useful in drag prediction but lift predictions are not accurate with this approach. So improvements are suggested for tunnel dimensions and also for parameters like computational grid, boundary conditions and ride height settings for vehicle. Drag prediction achieved is within 3% and lift prediction is consistently within 10 counts for different configurations of a real world small car tested in MIRA tunnel. Steady state simulations are deployed with turnaround time of 1.5 days on 64 CPU HPC

Notes:

Vendor supplied data

Publisher Number:

2016-01-1600

Access Restriction:

Restricted for use by site license