Sprinkler temperatures after activation

[Geoff Smith]

I am running a simulation to estimate sprinkler activation times for a given fire scenario. Sprinklers appear to be heating up and activating correctly at the prescribed temperature, but upon activation sprinkler temperature has either increased quickly or plummeted back to ambient. What should be the expected sprinkler temperature behavior in FDS upon activation, and what might be causing problems if my results do not go as expected?

[TimMcD]

My understanding is that when a sprinkler activates, its temperature reverts back to the ambient.

You could compare the times that each individual sprinkler activatea with the times that each individual sprinkler reverts back to the ambient temperature to check this.

I am not sure why the temperatures of the other (non-activated?) sprinklers increases quickly. Do these temperatures reach or get close to the activation temperature of the sprinklers? Could it just be that the fire is developing fast enough to outweigh the effect of the first few activating sprinklers??

(It is hard to generalise, but I would expect to see temperatures of non-activating sprinklers to plateau (i.e. very slow increase or slow decline) after activation of the initial sprinklers.)

[dr_jfloyd]

Sprinklers do not revert back to ambient after activating.  The DEVC is still evaluated. 

If you read the Tech Guide, you will see that there is a correction factor to the RTI equation to account for water droplets in the gas that can cool the sprinkler link.  This cooling term can cause a temperature decrease.

If you have simply prescribed a fire using HRRPUA, sprinklers will not do anything to change the size of the fire.  Also, the relatively large drops from a standard sprinkler are not very effective at cooling the gas. 

Look at the gas temperatures and where droplets are moving in your simulation.  This should provide you with clues as to what is causing a particular sprinkler to behave as it is.  I 

[Geoff Smith]

Attached is a graph showing the behavior of several sprinkler devices in the simulation using 20cm mesh cell size. Activation temp is 69 deg C. You can see that upon activation some heads (e.g. SPR_12) quickly return to ambient temperature, while others (e.g. SPR_6, SPR_8) increase. In other simulations with different mesh cell sizes, all the activated DEVCs seem to begin measuring gas temperature and increased in temperature (10cm mesh) or dropped back to ambient temperature (5cm mesh).

As dr_jfloyd said, it makes sense that after activation the DEVC is still evaluated, and sprinklers 6 and 8 appear to begin measuring upper layer gas temperatures instead of the sprinkler element temperature.

Is there a reason for this variable behavior following sprinkler device activation?

[Kevin]

Is this FDS 5 or 6? If it is 5, I suggest you try 6 because the rise
in temperature may be related to sudden changes in the flow divergence
due to the rapid release of droplets. Alot of work has gone into
improving the handling of spray nozzles in FDS 6.

>  Variable sprinkler behavior after activation 20cm mesh.jpg
> 61KViewDownload

[dr_jfloyd]

Varied behavior is seen since, presumably, the sprinklers are not all located in the same spot.  Apply different inputs to the RTI equation for the sprinkler and you will get different outputs.  The sprinkler response is a function of the local gas temperature and velocity and the local droplet density.  Look at the gas temperatures and velocities next to the sprinkler heads and look to see where the sprinkler drops are going.  This should reveal why the sprinklers are behaving the way they are.

[TimMcD]

I guess the initial question still remains? Why are the temperatures at sprinklers 6 and 8 still increasing well above the activation temperation. Have these sprinklers activated? If not, why is there no (significant) cooling effect.

[Kevin]

See my last post. This might be a bug, or a limitation of the
governing equations. If you are using FDS 5, see if the results of FDS
6 are similar. If they are, create a small test case and post to the
Issue Tracker. If FDS 6 does not exhibit this behavior, then we would
recommend that you use FDS 6 when it is released. If we determine that
it is a bug or a defect of some sort in FDS 5, and that FDS 6 "fixes"
it, we would naturally recommend that you use FDS 6.