There seems to be no wind here when simulating a large-scale environment

[lvzhoujing]

10x7x3: 20x20x3: 50x50x3: 150x150x3:

The process of modeling and CFD simulation is the same, and I found that no matter how much wind speed is set, it doesn't work. And I also want to know what the red particles represent?Is it because the particles are expired? There is a green wind arrow in the 10x7x3 room, but it has disappeared in other rooms(solved!!! wind_point_inv.scale.x = std::max(0.1,2*sqrt(pow(u,2)+pow(v,2)));). And when I put in an anemometer, the wind speed I get is very small (<1m/s). I don't know why there are spherical particles around the gas source, and they appear again at intervals.It seems like I fixed the issue by setting the End Time to 0.5s all the time during the CFD simulation(solved by changing the Write interval). but the wind speed measured by the sensor is still wrong.

Looking forward to your reply!

[lvzhoujing]

Changing the write control to Runtime makes it better.But wind values still strange.

[PepeOjeda]

Hello. I cannot really say for sure based on the images alone, but it seems like it might be a problem with the resolution at which the wind data was generated. Make sure that you are using the same resolution in gaden as you used in whatever CFD tool you used to create the wind, as otherwise there will be areas of the environment for which no wind vector exists.

Alternatively, if all you need is simple, homogeneous wind you can set the uniformWind parameter in the preprocessing node to true and create a file that contains a single vector value for the wind (for example: 1,0,0 would be all the file contents for a wind of 1 m/s in the positive x direction)

[lvzhoujing]

Hello. I cannot really say for sure based on the images alone, but it seems like it might be a problem with the resolution at which the wind data was generated. Make sure that you are using the same resolution in gaden as you used in whatever CFD tool you used to create the wind, as otherwise there will be areas of the environment for which no wind vector exists.

Alternatively, if all you need is simple, homogeneous wind you can set the uniformWind parameter in the preprocessing node to true and create a file that contains a single vector value for the wind (for example: 1,0,0 would be all the file contents for a wind of 1 m/s in the positive x direction)

I set the algorithm for generating the mesh to standard, and then the magnitude of the wind speed is normal. Thank you very much for your reply.

[lvzhoujing]

The gas profile is incorrect because the parameters time_step the GADEN.launch file are too large. I found that the anemometer reading at the (1,25,0.7) position has been small (<1m/s), then I added the wind speed value of this precise point in the wind.csv file, and the anemometer reading was normal.