question drag coef
hi
i am learning to use fluent, and i am trying to study the flow around a rectangle (2D model). i am trying to see the effect of the geometry on the drag coeficient: so i made different models: one with a rectangle, another rectangle with round edges and one model of a rectangle with a system of vanes on the end of the rectangle. i put the value of the surface area (here in 2D, i enter a distance) and velocity in report, and plot the drag coeficient. i obtained the drag coeficient (for velocity 16.6m/s): for rectangle: 3.491 for rectangle with round edges: 2.4697 for rectangle with round edges and vanes system: 2.222 there is a decrease of the coeficient with the change of geometry but i wander if the value of the coeficient is not too high. i was expecting something smaller (cause for a car for exemple, the coef would be around 0.3 or 0.4). can someone tell me if the coef seems to be good? or how can i be sure of the result? i have another question about the turbulence model: i would like to know the difference between the k epsi model standard and the realizable. and what is the most accurate model (closest to the real case) to represent the turbulence: k epsi, Reynolds... thanks yann 
Re: question drag coef
Hi Yann,
If u fell the coefficients are high, I think you shd have a look at the Units while importing the mesh into Fluent, u can do this by Grid>Scale, u can scale r unscale the grid as u want it to be, with regard to the Turbulence model, it is gud that u have some sort of Experimental values and then use various Turbulence models to match the same, so thaat u can set a bench mark depending on ur problem, with which u can further go ahead with ur computations, for further info u may refer the Fluent manual for reference. krishna. 
Re: question drag coef
hi,
thanks for your reply. i have check the scale with fluent, everything seems to be good in meter and at good size, but i am not sure if my drag coeficient is correct (3.491 for a k epsi turbulence model). i have try to make a 3D model and the drag coef is much smaller (around 0.7, which is more what i expected). is there nothing special to do for the calculation of drag coeficient for a 2d model?? for the choice of the turbulence model, i would like to find some experimental value so i can compare to the computation! where can i find such experimental value? thanks yann 
Re: question drag coef
Hi Yann,
What u say is 100% correct, 3d model predict the phenomena more realistically than the 2d models, while 2d problem considers a Infinite domain in the 3rd axis, which may be the problem, this has also been experienced by me. krishna. 
Re: question drag coef
Two dimensional simulation overpredicts the drag coefficient for flow across bluff bodies because it does not capture the large scale turbulence. I have similar problems to you.

Re: question drag coef
Hi Mr. Li,
I have a question. Based on what you said about large scale turbulence not being captured in 2D analysis, will the Cd also be overpredicted when it comes to 2D airfoil analysis? 
Re: question drag coef
It depends on the shape of your airfoil. I have only done simulation for incompressible flow over circular cylinder by solving the 2d RANS. The 2D RANS has been modified to solve flow with moderate separation but not with very large adverse pressure gradient. That is why it does not work for circular cylinder. For airfoil, which will not have a large wake area, it might work. I think there are some publications on 2D simulations of airfoil. Good luck!

All times are GMT 4. The time now is 01:36. 