CAARC wind tunnel validation

Project Support
1

Hi,

So I ran some simulations to validate my model for the wind tunnel tests done to CAARC building model. Here is the result I got. Can anyone help me understand why my results deviated like this. Red line is the wind tunnel test results btw.

Screenshot (415)
Screenshot (415)732×546 29.8 KB

2

Hi @clab128,

Could you please paste your project link here?

3

Hi @goncalves ,

CA01 NPL by clab128 | SimScale here is the link

Thank you for any assistance you can provide

4

Hi again @clab128,

Are you using any Pressure plot on your simulation? Under Result Controls tab there is no surface or point data for quantity chart. I just see the convergence plots on the link that you sent above, so if you are using them to compare the accuracy of your model, just keep in mind that only the last value has a physical meaning (as it is a steady-state one).
Please tell us more on how you are getting those values.

best,

5

Hi @goncalves,

I m using Paraview to get the results and i took the last value. If you are familiar with the CAARC wind tunnel test we just have to get the pressure values at 2/3 rd of height and calculate the pressure coefficient. using those coefficient values the graph is plotted and compared with the wind tunnel pressure coefficient values. I cant seem to get correct results for one side and the leeward side of the building. is that because of the solver options i have selected or something else? can you please help me finding out why it is.

Thank you.

6

Hello @clab128 , one thing that we could do to improve the solution is extending the flow domain. It seems like the building is too close to outer boundaries; inlet, outlet and top wind tunnel face. In case the boundaries are not extended enough, unexpected virtual accelerations may occur near the building in exchange affecting your pressure distribution.

In this documentation page, you will find a few recommendations on these dimensions for a PWC simulation type, however they should be applicable to Incompressible simulation type as well.

When you extend the domain, the computational demand will increase as well. In order to optimize the mesh size, please try using a coarser global fineness and use region refinements near the building.

I hope this will be helpful,
Thanks!