Apache system setpoint influence

[danylore]

Dear all,
I'm making a CFD simulation for a museum space with air distribution from diffusers placed across the perimeter of the building. I have calculate sensible and latent load with apache and establish the correct supply condition/air flow for each diffuser. The simulation works perfectly but I have a big doubt: I have to import in Microflow a boundary from a simulation with cooling and heating setpoint set to off in Apache in order to understand if my diffusers grants the correct temperature inside the space?

Setpoint on (22°C) in Apache:
I have quite a good uniformity inside all the room around the desired temperature (22°C)


Setpoint set to off in Apache:
I have quite a disaster with the desired temperature only near the diffurers and a mean temperature across the room far away from desired.


What is the right way to make the CFD simulation and have some truthful results?

Thanks!

[harshadj]

If you have the set points turned off, the surface temperatures obtained in the room are going to be drastically different than what you obtain with the set points turned on.

Since the boundary conditions are different, the results are going to differ.

Can you please explain this line a bit more:

in order to understand if my diffusers grants the correct temperature inside the space

[danylore]

Yes, I think the same as you for boundary condition and for this reason I'm quite sure that the right way to simulate it is the first one with setpoint on.

My concern is because I have to verify that my diffusers, with that flow rate, with that temperature/humidity, are able to grant 22°C inside the room so my question was "if I set 22°C in apache it could be possible that that temperaure is automatically verified in microflow because I have forced it?". During boundary import phase from vista pro also temperature is forced inside the ambient? This is my question, maybe stupid, but with my client I want to be sure to make thinks right!

Thanks for support

[harshadj]

Yes, you are right that would be a difficult client to deal with.

But you have to remember, the only way the room is going to achieve 22C is going to be with those boundary conditions that you imported.

The boundary conditions in MicroFlo are fixed. e.g. lets say a wall was set to 60C and air is being supplied at 10C. The supply air temperature has no effect on the wall as it stays constant at 60C. So the simulation where the fabric is reacting to the supply air temperature cannot be done in MicroFlo.

That is why we have Apache.

What you can say from your results is that the further CFD modelling is not required in this case. The room is small enough that the results of DSM are directly emulated in the CFD model as well.

if the room size were very large in one direction e.g. a double or triple height atrium with displacement ventilation. Here you would see a strong temperature gradient across the height despite the DSM saying the room temperature is 22C. I hope you understand what I mean.