doorfantest in FDS6 - air leakage through construction

[Liv Astrid]

I have been trying to simulate a doorfantest experiment in FDS.

In doorfantest you mount a fan in a door. You can measure flow through the fan and the pressure difference between the inside and the outside of the room. From this test you get a value called Air leakage rate (often referred to as n50). Air leakage rate(n50) tells you how many times the volume of air is exchanges per hour if there is a 50 Pa pressure difference between the inside and the outside of the room. The value is given as "Air changes per hour" [ACH].

 

I cannot seem to get results in FDS that are comparable to the results I got in my experiment. When I try to simulate the test the leak area is larger than the tests with small n50-values(13ACH) and smaller than the tests in large n50-values(40ACH). The leak area is approximately equal at n50-values close to 26ACH.

 

Input file is found here: http://pastebin.com/vJBmHUGu

 

test interval: n50 (13-40 ACH), Volume 29.4 m3, leak area was at least (0.0049-0.0412 m2) in the experiments.

 

FDS version: FDS6 Release Candidate 4 Serial. Compilation Date : Mon, 28 Oct 2013

 

Does anyone have experience simulating this type of experiment in FDS, or perhaps know what my problem might be?

[dr_jfloyd]

It is not clear what you have measured. Were all these tests done in the same compartment? For the same compartment the leak area should be fixed (unless you really over pressurize the compartment and make the leaks bigger) and what you should see as you vary fan flow rate is a change in the equilibrium pressure (and one expects the flow rate to vary with something close to the square root of the pressure difference).

[Liv Astrid]

I conducted 5 practical doorfan tests where I changed the air leakage in the room (I regulated the air leakage by opening and closing holes, thereby changing the leak area). The air leakage rate for the 5 different scenarios varied from an n50 value of 13 ACH to an n50 value of 40 ACH. The n50-value is defined at a 50 Pa pressure difference, so the equilibrium did not have larger pressure difference than 50 Pa in neither the tests nor the simulations. 

 

In the tests I mounted a doorfan in a door. The doorfan was connected to a pressure reader measuring the pressure difference between the inside and the outside of the room. The doorfan also measured the airflow through the fan. The computer software connected to my doorfan calculated the air leakage rate on the basis of the airflow through the doorfan when the pressure difference was 50 Pa.

 

In the computer simulation I aimed at creating equilibrium in the room at 50 Pa. I did this by calculating the flow at a certain air leakage rate. I used VOLUME_FLOW to make sure the FDS fan would blow the correct amount of air. Then I regulated LEAK_AREA until I managed to create equilibrium at 50Pa.

 

Example1:

An n50 of 13.4ACH in a room with a volume of 29 m3 gives a flow of 0.10943 m³/s through a fan with a surface area of 1 m2. In the practical test the leak area was at least 0.0049 m2 (most likely larger, due to an leaky construction). In FDS I get a LEAK_AREA 0.012 m2, which is 2.48 of minimum leakage area in the tests. 

 

Example 2:

An n50 of 38.7ACH in a room with a volume of 29 m3 gives a flow of 0.26868 m³/s through a fan with a surface area of 1 m2. In the practical test the leak area was at least 0.041 m2 (most likely larger, due to an leaky construction). In FDS I get a LEAK_AREA 0.035 m2, which is 0.85 of minimum leakage area in the tests. 

 

I find it strange that the leak area would be smaller in the simulation. I would expect it to be larger. The leak area in the tests were measured physical holes. Which means that the total leakage area (A_total=A_holes+A_leaky_construction).  

[dr_jfloyd]

At steady-state you have delta_P = K/2 rho U^2 where K is a loss coefficient that depends upon the size and shape of the flow path. The leakage model in FDS presumes you are leaking through small cracks and gaps and not a large contiguous opening. I would not expect that the effective flow losses would be the same for the fixed room leakage vs. the large openings you make.

Using your smallest vent size (13 ACH) and assuming FDS is perfectly modeling your compartment then your 0.012 m^2 FDS area means an fixed leakage of 0.0071 m^2 which is on the order of 0.01 % of the compartment surface area. Not an unreasonable value for typical construction.

A better approach would be to 1st determine the fixed leakage (do the test with no additional openings). Then for each of the openings you make you can determine the effect flow loss for those openings and define a duct to represent those openings.

[Gregor]

For this test cases I use the HVAC-model, too. With Jasons information, you get good results. It's like Yahtzee ;-)

dr_jfloyd: