Simulating Wind Flow over a building

[davidferguson92]

Hello

I was wondering if anyone in the autodesk community could help me with a problem I am having using the CFD 2014 software. I am trying to simulate the wind flow over a building in a real environment in order to determine the turbulent intensity, wind speed and to visualize wake regions over a building.

I am a student and new to the software so I am having difficulty setting up the simulation correcty and unsure how I should be setting it up. I was wondering if someone would be kind enough to explain the steps required to run an accurate simulation for what I am trying to do.

Also I am trying to simulate a velocity profile coming into the domain, I have read a post related to this and it seems it isn't possible to input a power law or log law profile however you could achieve the same result by splitting the domain into several faces and assigning an appropriate velocity to each. I was wondering if this could be explained further as well please?. Would you create several new solids around the building until you reached your required fluid domain height, assign a velocity to each input say 20 different velocity to represent the velocity profile and for each of the solid assign air and slip/symmetry around all sides and tops of the solids? Except the bottom as it represents the ground?

I hope someone can help me on this and provide step by step procedures if possible or even images of what to do.

I would really appreciate the help, and couldn't thank someone enough if they can set me in the right direction.

Kind Regards

David

[Jon.Wilde]

Hi David,

You could do this with a single cuboid over the model to represent the air. Everything else would be suppressed from the mesh.

Only the inlet surface needs to be broken into sections, each would have a different inlet velocity applied to it. Draw this in CAD.

As you say, slip/symm is typically used on the sides with a P=0 at the outlet to let the flow escape.

Your first run will likley have a mesh which is too coarse, maybe get it running first and then share an image or the mesh and results on a cut-plane. It is probably best to refine the mesh to a good level and then use mesh adaption for the final runs.

Kind regards,

Jon

[davidferguson92]

Thanks Jon

That really clears a lot up for me, regarding the splitting of the inlet surface how would I do this. Simply by adding a sketch plain to the inlet surface of the fluid domain and having various lines over the face?

Also should I specify a material for my building which I am trying to simulate or simply specify all parts as air? Also I am unsure if my simulation should be set as compressible or incompressible?

Sorry one more thing should I specify the velocity as fully developed?

Sorry for the harassment of questions I am just very inexperienced using the system and have found it very difficult to work the software for what I require it to do, as I haven't been able to find a video tutorial on the same application. I really appreciate the time you have spared to answer my questions, and the information you have giving so far has really helped.

Kind Regards

David

[Jon.Wilde]

Hi David,

That really clears a lot up for me, regarding the splitting of the inlet surface how would I do this. Simply by adding a sketch plain to the inlet surface of the fluid domain and having various lines over the face?

- You need to split the surface, lines will not suffice. The goal is multiple surfaces.

Also should I specify a material for my building which I am trying to simulate or simply specify all parts as air? Also I am unsure if my simulation should be set as compressible or incompressible?

- If you are suppressing the building the material does not matter.

- Compressible is for air speeds with compressible effects. These would be at mach 0.5 or so, which I doubt you will see here. Stick to incompressible unless you know someting I do not about the location of this structure 🙂

Sorry one more thing should I specify the velocity as fully developed?

- I would say no. It can be useful but as you have many inlets touching, there is no point in having the flow already developed (which means max v in the middle and 0v at the edge of the inlet surface). It would create an even stranger inlet pattern.

Sorry for the harassment of questions I am just very inexperienced using the system and have found it very difficult to work the software for what I require it to do, as I haven't been able to find a video tutorial on the same application. I really appreciate the time you have spared to answer my questions, and the information you have giving so far has really helped.

Don't worry, we are here to help. There should be lots of informaion on our Wiki site.

Have a nice weekend,

Jon