[yuig9488]

hi thx for read my thread.

i want fan curve
but that's too hard.
plz help me after see my condition

i tried first condition


inlet : total pressure 0 pa
outlet : static pressure 0 pa
fan omega : 100 rad/s
auto time scale: 0.02
i use sst model ,
diameter of fan domain = 0.40m
turbulence : first order
scheme : high resolution



after that i use parameter about outlet.
i increased static pressure of outlet .

When the static pressure was 0, the value for mass flow (airflow) converged well,
but even if I increased the static pressure by about 10 pa,
the mass flow value was too small and it took a long time to converge.

i tried second condition
outlet : mass flow rate : 0 kg/s
after that i use parameter about outlet.
i increased mass flow of outlet .

In the second situation, the pressure was 8 pa when the mass flow was 0 kg/s,
but it increased to 100pa when the mass flow was 0.3kg/s.


As far as I know, the positive pressure of the fan curve has a large value at the point where the mass flow is 0.
This deviation is too great.


I have read many threads but couldn't find the correct answer for me.
I hope you correct the wrong way to get the fan curve, or any other misleading information of mine.

thx for read

https://www.bilder-upload.eu/bild-87...39781.png.html

https://www.bilder-upload.eu/bild-4b...39835.png.html

https://www.bilder-upload.eu/bild-92...39993.png.html

https://www.bilder-upload.eu/bild-e8...40009.png.html

[urosgrivc]

Has your simulation converged at all?
Try to look at your results in the post, do these seem reasonable?
I mean the flow, does it flow all the way through the domain.

I have a filing that you should try with a larger timescale it might be that because of the rotating
domain the timescale is small and that's ok for the impeller part but the tube is very long and you
don't leave the simulation running long enough that the fluid would flush the whole domain to establish a constant flow.

If I were you I would set a monitor point at the beginning of the end of the tube and monitor the mass flow while it is solving,
this must converge to some steady-state for the simulation to have any meaning.

[yuig9488]

Originally Posted by urosgrivc

Has your simulation converged at all?
Try to look at your results in the post, do these seem reasonable?
I mean the flow, does it flow all the way through the domain.

I have a filing that you should try with a larger timescale it might be that because of the rotating
domain the timescale is small and that's ok for the impeller part but the tube is very long and you
don't leave the simulation running long enough that the fluid would flush the whole domain to establish a constant flow.

If I were you I would set a monitor point at the beginning of the end of the tube and monitor the mass flow while it is solving,
this must converge to some steady-state for the simulation to have any meaning.

Thanks for the answer. Is it okay how to get my timescale and fan curve?
As a result of observing the streamline, the air from the fan outlet goes back to the fan inlet. Is this a natural phenomenon?