Wind hitting the ground


#1

Hi everyone. I am new to SimScale. I just learned to use the wind tunnel feature of SimScale. Now I would like to try a new simulation.

I would like to simulate wind hitting the ground. I would like to see how the wind’s velocity changes once it hits the ground and flows away. Wind hitting the ground is associated with weather events such as “downbursts” when denser air falls through less dense air and hits the ground.

If any of you are knowledgeable about how to do this, can you please explain to me or point me in the right direction? Thank you! I apologize if this topic is already covered in a tutorial somewhere. If it is, please kindly point me to it.


#2

Hi @derekfong!

One approach would be to define a big domain with a topology that is right above the bottom surface and you vary the angle of attack via variable inlet conditions. Another method would be to use a mesh like here: Flat Plate Validation Case and also adapt the boundary conditions - the slip condition at the top might be changed to an inlet as well or custom boundary condition. Let me know what you think!

Maybe our PowerUsers @Get_Barried , @vgon_alves and @Anware have some additional tips that help!

Cheers!

Jousef


#3

Hi @derekfong,

The setup of the simulation my not be simple for something as complex as a "downburst ". Replicating such a phenomena as a inlet can be very challenging. For starters, it is possible to create a top down inlet where you can insert a high velocity in an attempt to simulate that “downburst” and see how it performs. The simulation will probably need to be a transient one as I do not believe a steady-state simulation will give representative behavior.

Geometry wise, a simple box is likely sufficient. As mentioned replicating the complexity of the inlet will be the challenging part. Do try it out and let us know!

Cheers.

Regards,
Barry


#4

Hi everyone,

Well, reinforcing what Barry and Jousef said, it’s a big challenge to craft a domain that’s true to the downburst phenomenon.
Another initial proposal to your geometry may be to configure in your domain a sphere in motion with the properties of the downburst creating a primitive geometry here in SimScale. But for this to work, you must take into account what is said below:

Some considerations that can give you a North:

  1. A very fine mesh to capture the behavior of the phenomenon with fidelity;
  2. LES Approach to solve the large scales of turbulence that you must have in the model may be interesting;
  3. The timing of the phenomenon must also be a challenge in terms of processing, and [1] should certainly require many core hours.

If you have good source material about this phenomenon and you know the mass of air you want to simulate, you can try the multiphase flow by setting that density variation you cited for each fluid material, one for ambient air and one for the downbusrt . Then you can set a passive scalar transport to see the movement of downburst only with respect to flow properties, leaving alone the material differences.

Finally, it’s a very good idea!!The results will be really cool! If I had had this idea, I’d be obsessed to see it working! :smile:

Have a good time!

Vinícius