Fds and ozone

[Brijesh]

Hi,

I am trying to do a project which involves modelling using ozone software and fds and comparing the results. Simple cases are only considered as of now. I made a 3x3 room and modelled it in ozone. The same room was done using ozone, where I am unable to get the expected results. The temperature data I am getting is much off when compared to ozone. I am unable to find out the mistake I am making. Can anyone help me out. I am new this fds and still learning only. The comparison of temp data obtained from both ozone an fds is attached.

The code I am using is below.

&HEAD CHID='3x3_room', TITLE='checking 3 x 3 room' /

REM room size 3x3x3m, door size 1x2m, Window size 1x1m, burning object loc - middle of room, 

&MESH XB=0, 3, 0, 3, 0, 3, IJK= 30, 30, 30 /
&TIME T_END=2042. /

&VENT XB= 1, 2, 0, 0, 0, 2, SURF_ID='OPEN' / door
&VENT XB= 3, 3, 2.25, 0.75, 0.5, 2.5, SURF_ID='OPEN' / window

&REAC FUEL = 'PROPANE' /

&OBST XB= 1, 2, 1, 2, 0, 0.5, COLOR='RED', SURF_ID='fire1' /

&SURF ID='fire1', HRRPUA=250 /

&DEVC ID='T air 1', XYZ=1, 2, 2, QUANTITY='TEMPERATURE' /
&SLCF PBY = 2, QUANTITY='TEMPERATURE'/

&TAIL /

[dr_jfloyd]

I am not familiar with ozone. A couple of observations:

1. Your T air DEVC is over the edge of the burner. Temperature at this location is going to be sensitive to the details of the combustion and turbulence. Small changes in plume width between two models could result in larger differences at this location. A better location would be further away from the burner. Also, comparing one model against another doesn't tell you much. If you get different answers you don't necessarily know which model is correct. A better approach would be to take a simple experiment and model it with the two models. Then you at least know the experimental results.  The Steckler tests in the validation guide are a similar geometry to what you have here.

2. Your fire definition in FDS is a fire that ramps to 250 kW over a period of 1 s and then stays at 250 kW.  With the TC at the edge of the burner the FDS model shows a rapid rise to a constant temperature. Since this TC is probably in the fire plume, this result makes sense since you have a constant fire size and compartment with two openings.  The ozone model shows 1500 s to reach 800 C and then dropping quickly after that. Does this make sense to you to given basic fire dynamics? Is 800 C the temperature you would expect in a 3 x 3 x 3 compartment with close to 5 m2 of openings?  Why is the temperature dropping in ozone, this suggests the fire is dying out.  Would you expect a 250 kW fire to die out in a compartment with an open door and window?

[Brijesh]

1. Ozone follows the euro code. Heating phase- steady phase-cooling phase. In fds also I want a similar study only. The fire starts and then dies down as it practically should(if the fuel is exhausted).

2. Can you suggest the modification in the code required so that the above scenario can be simulated. 

[dr_jfloyd]

The Eurocode compartment temperature curve isn't based on a single specific fire. It is meant to be representative of a compartment fire temperature profile. If you want FDS to replicate the Eurocode temperature profile you will need to spend some time iterating with the input file to figure out what combination of time dependent heat release rate and wall thermal boundary conditions result in that temperature profile.

[fde]

Ozone is 2 zone model and averages the upper layer temperature. If you are calculating heat transfer to the steel element, then uses Alpert's correlations for ceiling jet. 

A remark for possible future work: Not that you have utilized yet but Ozone radiation model assumes a constant fuel bed area even the prescribed fire is a transient and growing fire. That underestimates the radiation. (v.3.0.4)

You can also check the documentation of Ozone about its calculation model. 

https://research.bauforumstahl.de/nl/fire-safety-1/locafi-temperature-assessment-of-a-vertical-member-subjected-to-localised-fire-10/