Unexplained behaviors of duct airflows in transient simulations with WTH file

[Rodrigo Mora]

Hello,

I laid out a duct system with branches terminals and returns. I set up a constant volume flow rate (scfm) for the supply fans. I balanced the duct terminals. When running the project in steady-state the flows air maintained as designed. But when I run transient simulations, with a WTH file, the flow in the branches changes, and is not consistent with the constant flow of the fan. I am puzzled about the reason for this behavior (see the image attached). 

Even when I turn off the fans, the flow in the ducts is not zero in transient simulations. It goes from 40 scfm to about 10 scfm at the same hours every day. Furthermore, if I run the simulations with the fans off for several months I get the message "Airflow simulations failed to coverge". I am testing constant flow fans for now. But if I make this model work, I plan to run DCV airflow controls, with the fan flow ranging from 0% to 100%. 

Thanks,

-Rodrigo

[William Dols]

Rodrigo,

It is nearly impossible to address these issues without the input files.

You can send them to me directly if you are averse to sharing them with the group.

- Stuart

[Rodrigo Mora]

Thanks a lot Stuart,

I think I found the problem, but it still puzzles me. I realized that I had left a window in the ground floor of the building fully open every day on a schedule, which brings 500 cfm to the building due to stack effect depending on the outdoor temperature (from the WTH file). I suppose the amount of air entering through the window every day changes the pressure balances in the building, and thus the internal flow rates. However, I am still puzzled about the air distribution system. The system consists of a constant volume 350 cfm fan that supplies air to the rooms, which recirculates back to the fan. Shouldn't the supplied air, from the constant volume fan, not be affected by the interior pressures? The terminals have been balanced, and the ducts are modelled with zero leakage. I understand that the duct temperatures use the same temperature of the space where it is located. However, the interior temperatures remain fixed, not fluctuating with the exterior temperatures.  

I tried to explain this by following the conservation of mass, not conservation of volume, principle. Given that the outdoor air is cold, it will change the density of the indoor air and thus all the mass flow rates, even through the ducts. But, again, the temperatures in the spaces remain constant, i.e. are not affected by the outdoor air temperature, therefore the temperatures/densities in the ducts also remain constant. Furthermore, density variations should not have such a big impact on duct airflows.

My conclusion is that even if a constant volume fan is blowing air through the ducts, the actual amount of air flowing through the ducts is a function of the pressure differentials between the air terminals, right?

My file is attached just in case you want to look at it, the open window is named "openw12" in the bathroom of the ground floor.

Thanks!

-Rodrigo

[Dols, William Stuart (Fed)]

Rodrigo,

 

Ducts are simply flow paths that can provide connectivity between non-adjacent zones.
Even though constant fan flows will be met, zone pressures are accounted for with respect to relative duct/terminal flows when there are multiple branches/segments.
When fans are off, the ducts still provide leakage paths through which air can flow depening on driving pressures and relative leakages of the entire airflow network.

Junction temperatures are user-defined/scheduled.
Duct junctions do not use “interior” temperatures when running CONTAM uncoupled with TRNSYS.

I see that you’ve greatly reduced the Absolute Convergence Factor for airflow from 1e-6 to 1e-10.
This is quite strict and resetting back to the default seems to alleviate convergence issues .

I also see that you are not accounting for wind pressure in your model.

 

- Stuart