Smoke Layer Height

[Austin Appino]

I am having trouble with understanding the smoke layer height and average upper and lower temps in my simulation. I have read 17.10.5 which states how to denote these cases in the .txt file. Although my file runs, the data for smoke layer height and temps do not change over time. Any advice would help. I have attached both the .fds file and the a_devc file.

Thanks,

Austin J. Appino

[dr_jfloyd]

&MESH IJK=14,92,8 XB=16,23,0,46,6,10/

Your overall domain is 10 m tall.

&TIME T_END=50.0 /

You are running your simulation for 50 s. 

&SURF ID='fire', HRRPUA=6400, RAMP_Q='fireramp' /

&RAMP ID='fireramp', T=0.0, F=0.0/

&RAMP ID='fireramp', T=100, F=0.02941/

&VENT XB=19,19.707,22.5,23.207,0.1,0.1, SURF_ID='fire', COLOR='BLUE'/

At 50 s your fire is only 90 kW.  

&DEVC XB=17,17,23,23,0.0,3, QUANTITY='LAYER HEIGHT',ID='smokeheight' /

&DEVC XB=17,17,23,23,0.0,3, QUANTITY='LOWER TEMPERATURE',ID='uptemp' /

&DEVC XB=17,17,23,23,0.0,3, QUANTITY='UPPER TEMPERATURE',ID='lotemp' /

You are telling FDS to compute the layer temperature for the first floor. 

Do you expect that a fire that grows to 90 kW in 50 s will result in the upper layer descending from 10 m down to 3 m?

[Austin Appino]

I think you're right. I let the program run for 1500s over night and the numbers only vary after 50s. Is the layer height number refering to the height of smoke from the floor to the ceiling or from the ceiling to the floor. Thanks again.

[dr_jfloyd]

At t=0 is the layer height in the a_devc.csv file 0 m (which would mean it is measured from the ceiling) or 3 m (which would mean it is measured from the floor)?

[Austin Appino]

Its at 3m. Which I believe means its measured from the floor. Thanks so much.

[Austin Appino]

I have run the simulation for 1200s now and my smoke heights do not change at all during the simulation. I place three "sensors" for smoke layer height, upper and lower temp at locations that I thought were worst case scenarios. The temperatures change but the smoke height is staying at 3m. Do you know what could have gone wrong? It worked before I added the new sensors and moved them to new locations. I have attached the file.

Thanks,

Austin J. Appino

On Tuesday, June 5, 2018 at 7:44:43 AM UTC-5, dr_jfloyd wrote:

[dr_jfloyd]

You are modeling an atrium. You have large openings between floors which means you have one coupled volume that the smoke layer is descending through. I suggest you have one set of layer devices that sit within the atrium opening that span the full height of the atrium. 

[Austin Appino]

Is this new file what you meant?

Thanks,

Austin J. Appino

[dr_jfloyd]

No.  I was suggesting that your z coordinate go from 0 to 10 m rather than 0 to 3, 3 to 6, and 6 to 9.

You have a very coarse grid. On the 3rd floor for example the temperatures are essentially constant away from the fire plume since the coarse grid plus the very large exhaust rates means the layer is very well mixed. The concept behind the layer equations breaks down if you have essentially isothermal conditions from ceiling to floor. Measuring over the full height of the atrium will avoid this. 

[Austin Appino]

Thank you. Will this give me any data of the smoke layer on each floor? That is what we are mainly concerned with.

[dr_jfloyd]

If the layer is 9 m, it is at the ceiling of the 3rd floor. If the layer is 6 m the 3rd floor is full of smoke and the layer is at the ceiling of the 2nd floor. etc.  You have an atrium with large openings between floors, a very coarse grid, and a fairly large exhaust flow rate for the ceiling height. 

[Austin Appino]

I have changed my file to be 3 single meshes with one sensor from 0,9 m in the z. I ran it for 1200s and it still only read the smoke layer at 9m the entire simulation. The temperature changes very little but it does change. Any tips?

[dr_jfloyd]

The layer temperature equation needs a well stratified environment for it to work properly.  Your "upper layer" is barely above ambient. This is not a well stratified environment. 

[Austin Appino]

Is there anyway to have the _devc file output smoke toxicity, upper/lower smoke conditions of each level, or visibility on each level? These are the main data points we are researching for this building.The goal is to have a enough ventilation to create a safe environment for people to evacuate the building in 1200s (although I believe that's an extreme time limit for this situation).

[dr_jfloyd]

You can place either slice files or devices at any planes (slice) or points (devices) you desire for any of the output quantities listed in the user's guide.

[Austin Appino]

Yes, I understand that. I have found a way to output visibility. I am having trouble getting more information on how to understand the smoke layer height through the outputs that are available. Specifically, the upper smoke layer height/toxicity and the lower smoke layer height/toxicity on each floor. I have looked through CH. 17 which talks about these but haven't found specifics. Are there other sections that are more useful?

[dr_jfloyd]

You have an atrium with large floor to floor openings; you don't physically have independently stratified layers on each floor. A floor is either in the lower layer of the atrium, in the upper layer of the atrium, or contains the layer interface. Using the layer averaged quantities will underestimate that hazard on the top floor and over estimate the hazard on lower floors. If you just want to use layer quantities over the height of a tall atrium, you may as well use a zone model. CFD gives you the local data. If you wish to reflect the hazard with as little bias as possible, use the local data that the model is giving you.