[Tobi]

Dear community,

It's a while ago that I asked a question here but today I do have some thoughts in my mind that I was not able to solve.

The topic: I do have an electronic board that has a fan which sucks in the air for cooling. Nothing special so far. However, the chasing in which the fan is mounted limits the cross section for the incoming airflow. Furthermore, the cutoffs from the housing (where the warm air can leave the closed pcb room) are blocked up to 90 %.

What I want to check is, how much air might the fan push into the housing.

What I know: The maximum pressure the fan can create (1 mmH2O which should be a pressure drop of around 9.8 Pa).

Possible setups...
1. Ignoring the intake cross section reduction in front of the fan and ignoring the fan and appy a total Pressure at the Fan backside having 9.8 Pa higher pressure compared to the fixed outlet pressure. However, this will not incorporate any influence on the flow rate and pressure drop od the Fan due to cross section reduction in front of the fan.

2. Adding the part before the fan and model the fan with an fanpressure Jump (?). However, I was not able to set up the cyclic BC in that case. Nevertheless I was hopeful that using this BC will be a better option. However, I am not sure anymore as if we would close the whole front of the fan, no Flux will pass the fan and hence we don't have a constant pressure drop.


Are there any people around who already did something like that or do have similar investigations already successfully done?


So summing up. I know the maximum pressure drop and the flow rate of the fan at maximum pressure drop. However, if we limit the cross section before the fan, we have to get a lower Flux, right? The same is valid if the fan would try to push air inside a housing that has almost no holes where the air can leave (to high back pressure and thus low flow rate).


Question. Is it possible to investigate into that topic using fanpressure or fanpressure Jump? Or maybe other conditions and models?


Any idea is warmly welcomed.
Pictures will follow

[Tobi]

Dear community,


probably it is already working as I want to have it. I am using a fanPressureJump for the cyclic patch types and indeed it gives different flow rates. A normal channel gives 8000 [units] and the other one 83 [units]. Hence it works as expected.

[Daniel_C]

Hi Tobi,
old thread meanwhile but would you mind telling us some more detail about your beforementioned case setup, especially the domain boundaries? Where is the fan, and how and where did you apply the BCs?

Greetings

Daniel

[Daniel_C]

Hi,

to start with something. Does the setup look like the pic attached?

Maybe anybody else has an idea?:

ciao

Daniel