Improving Solution
 

Hello All, I am doing external flow on an object to compare it wind tunnel tests, but I am off on my drag, lift and moment coefficients, any idea how I might improve mu solutions ? Thanks

Re: Improving Solution
 

Hi,

Well where do we start...

Here's a few pointers, but there are so many things to consider!

1) Which component of drag is out the most pressure (form) drag or skin friction (viscous) drag?

2) Is the mesh fine enough in the near wall region to capture the boundary layer affects and satisfy the turbulence model Y+ requirements?

3) A fine mesh in the near wall region will also help to give more accurate predictions of pressure drag and also lift.

4) Is the mesh in the wake fine enough to capture the wake coming off the body?

5) Are the far field boundary conditions placed far enough away from the body being modelled?

6) Are you applying the correct boundary conditions upstream and downstream?

7) Are you setting the correct values at the boundary conditions i.e. the same inlet flow speed and turbulence level as measured in the wind tunnel.

8) Are you modelling exactly the same flow conditions/geometry as per the wind tunnel tests i.e. are you modelling the tunnel walls correctly as no-slip boundaries? Blockage effects might also affect the predicted forces.

9) Has the simulation been allowed to run long enough i.e. down to a low residual (1.0e-06 max or machine round off)

10) Is the simulation transient thus requiring a small time step to produce good convergence?

11) Have you set the fluid properties correct i.e. the correct density, viscosity (be careful not to get dynamic and kinematic viscosity mixed up!) & fluid temperature?

12) Are you using a high order discretisation scheme?

There are many more things which could affect the accuracy of your simulation not least the accuracy of the wind tunnel measurements themselves!

James