Stack effect and localized leakage

[Marcellus T]

Hi,

I am rather new in using FDS (Pyrosim). 

I want to simulate the pressurization of a stair case in a tall building.

1. For leakage trough the cracks of a closed door between rooms I want to use the localized leakage approach. From the tutorial local_leakage_temperature it appears every sealed room needs a pressure zone (?). Could anyone explain why pressure zones are needed in this case? I would say that the rooms are connected through the leakage path and therefore 'feel' each other's pressure? 

2. The stack effect has to be included. I have the FDS example stack_effect. In this example HVAC elements are used for this. But is it also possible to use the localized leakage approach?
In this approach do I need a pressure zone in every room where there is leakage to the outside world?

3. Does using pressure zones increase computational costs?

Thanks, Marcellus

[dr_jfloyd]

1. The bulk leakage model (i.e. using ZONE LEAK_AREA and SURF LEAK_PATH) only uses the background pressure. If everything is in the same zone, then everything will have the same background pressure.  The localized leakage model will use the local pressure so having individual ZONEs is not absolutely necessary; however, it will help with velocity errors at walls. It would be good practice to use ZONEs.

2. You can use the localized leakage approach.  The stack_effect case was created before we added the localized leakage inputs.  

3. There is no significant increase in computational cost or memory usage by having multiple pressure zones. 

[Marcellus T]

Thanks for your reply! 

1. Am I right: the local pressure =  zone pressure + a flow induced perturbation?

2. My idea was to model a certain part of the environment of the building and define a certain ambient temperature there. Then inside the building I initialize everything with a different temperature. This should induce the stack effect in the building.
Is it also possible to do this with pressure zone leakage (in stead of localized leakage)? In that case I define different pressure zones at all floors. And at every floor I define ZONE LEAK_AREA and SURF LEAK_PATH, modeling leakage through windows etc. at that floor). Is this going well with the difference in hydrostatic pressure inside and outside the building?

 

Is there also a possibilty to model the stack effect without modeling the environment of the building?

-> What would be the best approach for modeling the stack effect (HVAC/pressure zone leakage/localized leakage/...)?

3. Glad to hear

Op dinsdag 4 september 2018 15:12:21 UTC+2 schreef dr_jfloyd:

[dr_jfloyd]

1. Yes

2. See this thread for some discussion: https://groups.google.com/forum/?fromgroups=&nomobile=true#!topic/fds-smv/YKl2hhUSRNs

[Marcellus T]

2. I've already read that, unfortunately I can't find the answer.

Op dinsdag 4 september 2018 16:03:28 UTC+2 schreef dr_jfloyd:

[dr_jfloyd]

From the thread:

You could try something like this:

Set TMPA to the outside temperature.

Make the inside of the building a different zone

Use the localized leakage to add multiple leakage paths over the height of the building.

Use INIT to set the inside temperature of the building

Use TMP_INNER to set the steady state wall temperature profile for the exterior and interior surfaces (otherwise they will initialize at TMPA)

You will want to add a set of constant mass flow supply and exhaust vents inside the building. Aim for a typical building air change rate and with that mass flow set the temperature to compensate for the expected leakage flow into the building plus losses through the walls. 

Without a fire and the natural ventilation openings closed run this for a period of time until the leakage flows and interior temperature stabilize.

Then start the fire.