Sprinkler temperature keeps raising after freeze control is activated in FDS model

[SHERRY WANG]

Hi,

I have a question about sprinkler temperature in an FDS model. Thanks for taking the time to read my question below:

I am doing a sensitivity check for my FDS models with different mesh sizes of 0.1m, 0.2m and 0.4m.
I am using a quick response sprinkler, the activation time of the sprinkler is 68C, the code is below:

&PROP ID='Commercial Spray',
QUANTITY='SPRINKLER LINK TEMPERATURE',
ACTIVATION_TEMPERATURE=68.0,
RTI=50.0,
C_FACTOR=0.65,
PART_ID='Water',
FLOW_RATE=56.1,
PARTICLE_VELOCITY=5.0,
SPRAY_ANGLE=60.0,75.0/

I also have a freeze time control to stop the heat release rate (HRR) from increasing once the sprinkler is activated at around 68C.

In the 0.2m and 0.4m mesh size model, all the sprinklers' temp reduce to the ambient temperature of around 20C immediately after sprinklers reach close to 68C. 

However, in my 0.1m mesh size model, after 68C, the sprinkler temperature still keeps raising to around 85C then remains unchanged, and it does not decrease to the ambient temperature. And I saw the freeze time control is activated when the sprinkler temp reaches around 67C as the HRR stops increasing from this point. So except for the sprinkler temp, all the other things look reasonable. 

Does anyone know the reason the sprinkler temp keeps raising to 80-90C and is not reduced to the ambient temperature?

Many thanks!

Sherry

Fire Engineer

[fde]

I would suggest adding a temperature device at the identical location of sprinkler and watch the temperature at this device as well. And see if temperature drops at this device as it should.

[Kevin McGrattan]

Are you referring to the sprinkler "link" temperature or the air temperature near the sprinkler? In either case, the change in grid size will change the air and droplet flows in the grid cell that contains the sprinkler. I suggest that you zoom in close to a single sprinkler and observe the flow vectors, temperatures, and so on. By the way, in actual experiments, the temperature near an activated sprinkler is erratic. That is not surprising.

[dr_jfloyd]

One contributing factor might be that the default OFFSET for a sprinkler is 10 cm. With the 20 cm and 40 cm mesh all the drops are being placed in the cell with the sprinkler head. With a 10 cm mesh, the drops are being placed in the grid cells adjacent to the cell with the sprinkler head. 

[SHERRY WANG]

Thanks all,

I have a 'link' temperature.

I checked the layouts of these three models with different mesh sizes. In the 0.2m and 0.4m mesh model, some sprinkler heads are on the boundary between two cells, and some sprinkler heads are placed in the cell.

In the 0.1m mesh model, all the sprinkler heads in the point of the cross. 

The flow vectors show the flows near the sprinkler heads are erratic. But after the sprinkler was activated, I can obviously see the flow downward and form an umbrella shape.

Thanks again for your suggestions!

Regards,

Sherry

[SHERRY WANG]

Hi, thanks!

I added a heat detector and thermocouple, they are 10mm lower than the sprinkler head. After the time of sprinkler activation, the temperatures measured by the heat detector and thermocouple have the similar trend as the sprinkler head, they keep going up for a while then remain unchanged with a few fluctuations,  almost parallel to the sprinkler temperature and lower than it. 

Regards,

Sherry