CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   Proper BCs for internal fan (https://www.cfd-online.com/Forums/cfx/77538-proper-bcs-internal-fan.html)

 serezhkin June 25, 2010 21:18

Proper BCs for internal fan

Hi everyone,

I am trying to setup and internal fan in CFX and I would like to get your opinion.

I am specifying an CFX inlet (fan outlet) with static pressure BC, where
P(fan_outlet)=P(fan_inlet)+dP,
and where dP is a function of mass flow rate at fan inlet according to fan curve.

The BC on fan inlet is CFX outlet BC with mass flow rate condition, where mass flow rate (fan inlet) is equal to mass flow rate (fan outlet).

It seems like these conditions work, but I would like to get your opinion if this is a good way to setup an internal fan.

Any suggestions and comments are appreciated.

 ghorrocks June 26, 2010 07:31

You can also do this where you get the pressure on either side of the fan and apply the mass flow rate from the fan curve. Same fan curve, just looking at it a different way - but this way will probably be more stable.

 serezhkin June 28, 2010 13:48

I think you are right. The pressure condition does seem less stable. I will try the mass flow rate BC instead. Thank you!

Update: I tried setting Mass flow rate (MFR) on both sides of the fan using the Fan curve and pressure difference across the fan, and it turns out this BC is unstable and leads to a divergence within 3-5 iterations.

Update 2: Setting MFR_in=MFR_out on fan outlet and pressure BC (according to fan curve and MFR) on fan inlet resulted in oscillation of residuals (no convergence), so even though it is more stable, it does not converge.

Update 3: The poor convergence (oscillations) was resolved by weakening the average pressure condition on fan inlet side: P_weak(average)=0.5*P_current+0.5*(P_fan_outlet-dP(fan curve))
The fan outlet condition is still mass flow rate: MFR_outlet=MFR_inlet.
I will experiment with coefficients in the weak pressure BC, but as long as there no oscillations of residuals, I believe the BCs work fine.

 nagarjunae July 28, 2010 10:04

a...................................

 All times are GMT -4. The time now is 22:47.