Wind simulation test - wall not "blocking" the wind

[mikeragna]

Hello!

I've been running a few tests with the wind simulation, regarding interferences of surrounding objects on the structure, but the results are kind of odd. See the model attached, with a series of 5x5m concrete walls, aligned at different distances. Notice that, with two walls distant 1m from each other, the results of the wind simulation are equal on both of them, like there is no interference whatsoever. The front wall starts to really "blocking" the wind for the back wall only at 4m of distance.

Somebody had the same problem before? Or is that a normal behavior? (Doesn't seem like though.)

That has maybe something to do with the size of mesh used for the wind tunnel? If so, how to adjust to get better results?

[mikeragna]

A few more tests.

"Wind test 2" is the same as the original "Wind test", except I've isolated two walls distant 1m from each other. The results are better then the first case, with probably means that the mesh of the wind tunnel is related to the size of the model, so a small model shows better localized results. An interesting thing to notice is that the generated load on the front wall is basically double the value of the wind pressure, for some odd reason.

"Wind test 3" is a tower with outter and inner walls. I've selected all walls when simulating the wind load, and expected the simulation to "see" that the wind could not affect directly the inner walls, since they are confined by the outter walls. But that's not the case by the results, we can see that the wind is aplied to the inner walls, like if the outter ones do not exist.

"Wind test 4" is the same tower, but only the first floor. The results are similar to test 3, with the wind still affecting the inner walls.

How can I configure the wind simulation to make sure that it recognises the interference of the objects in the model?

[okapawal]

We've been working recently on improvements to the Wind Simulation feature and fixed many issues.

If you would like to check out the recent Preview release you will find much better results.

Note: you can not open files saved in Preview release back in 2020 version.

You can sign-up and download installation from the feedback portal.
 < https://feedback.autodesk.com/key/RSA_Feedback>

[mikeragna]

Hello!

I've made the tests with the pre-release 2021 version of the program like you sugested, but the results where quite similar regarding the problem that the simulation is not considering the interferences of close objects. Files attached.

On tests 1 and 3 (large models), close walls receive the same wind pressure. On tests 2 and 4 (small models), front/outter walls are correctly blocking the wind for the back/inner walls. Apparently, large models often apply the same wind pressure to closer objects, while small models recognize the interference of objects blocking the wind.

Any sugestion on how to make the simulation recognize the interferences of close objects on large models like it does on small ones?

[mikeragna]

Any updates on this topic? Could you try to replicate my results and see if there is another alternative?

[okapawal · Accepted solution]

Yes, small elements with close localization may not reflect real wind load.  As you suspected, dimensions of the virtual wind tunnel depends on the model size.  But the tunnel box is divided with a constant number of elements along an edge.  So in case of bigger models some small elements may fall into a single wind tunnel voxel.  As  result they get the same pressure load.

[mikeragna]

Thanks for the reply and confirmation of my assumptions! Considering that the wind tunnel box is divided with a constant number of elements along an edge, it would be great if, in future updates, we could control the size of the element division of the tunnel, to reflect better results if needed, although it whould take a lot more time to process.