CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Ahmed Body (http://www.cfd-online.com/Forums/main/1605-ahmed-body.html)

Chris Robinson December 2, 1999 13:05

Ahmed Body
 
I would be interested in hearing from anybody who has completed calculations on the Ahmed body. (Bluff, car-like model principally used for studying wake, drag and lift effects for motor vehicles). If you know of any decent journal references I'd be grateful if you would pass them on. Many thanks Chris

andy December 3, 1999 08:14

Re: Ahmed Body
 
We predicted it with RST and k-e models a few years ago. Conclusions as I recall:

- Standard k-e was a waste of time. The turbulent boundary layer on the floor fed into the stagnation region in front of the car generating a spuriously high level of turbulence which consequently knackered the boundary layers over the vehicle. This is a well known problem. If closing the turbulence equations at this level (i.e. approximating the Reynolds stresses with an eddy viscosity model) then add a kludge to reduce this sort of thing. Reworking the production term for the turbulence kinetic energy is popular.

- Various RST models did not suffer this problem or not to the same degree.

- It is vital to resolve the trailing edge well. The flow modes are very strongly influenced by this region. We could move the point (region strictly since it depends on whether the flow is attached or detached initially) where the flow mode switched by well over 10 degrees of slant angle by modifying the grid arrangement in this region (i.e. same grid density but different grid skewness). Far from grid independent solutions! As an estimate one is going to need 10 million grid points or so to bring down the effects of grid dependence to a reasonable level. At the time we did not have the resources to do this.

You do not say how you are going to simulate the flow. It seems a good test case for LES or Monte Carlo pdf models although the measurements lack details of the turbulent stresses. As I recall (but I could be wrong) a Japanese group performed a crude LES simulation (they ignored the turbulence model, used high order differencing for the convection terms but without controlling the energy transport) but with a high grid resolution and got good looking answers.

Sung-Eun Kim December 6, 1999 17:38

Re: Ahmed Body
 
Dr. Han at GM has published a paper early 80'.

Han, T., Computational Analysis of Three-Dimensional Turbulent Flow Around a Bluff Body in Ground Proximity, AIAA J., Vol. 27, No. 9, pp. 1213-1219, 1988.

Fluent Inc. has lots of experience in this flow. We have some results summarized in our forthcoming paper;

Francis T. Makowski, and Sung-Eun Kim, Advances in External-aero Simulation of Ground Vehicles Using the Steady RANS Equations, to be presented at SAE 2000

In a nutshell, a nice mesh combined with advanced turbulence models can predict the major flow features and forces with reasonable accuracy. In our calculation, we used a hybrid (prism near the wall and tet mesh elsewehre) and Reynolds stress transport model (RSM).


All times are GMT -4. The time now is 02:51.