Thunderhead Engineering Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Forum moved to https://forum.thunderheadeng.com

Author Topic: Difficulty modelling sprinkler actuation times  (Read 4358 times)

dambook

  • New Member
  • *
  • Posts: 8
    • View Profile
Difficulty modelling sprinkler actuation times
« on: June 06, 2015, 06:09:21 pm »

Hi all

Quick one here. I'm struggling to get sprinklers within this scenario to activate within the 300 s I'd like the model to run for but the sprinkler heads never reach the activation temperature of 68 C despite the 10 MW fire burning underneath (I was expecting them to set off around 90 s to 100 s realistically ). I know the mesh size is completely out of order but even with much smaller cell sizes still not getting favourable times if anything at all. Is there something I'm not doing right here? No vents, doors or open surfaces exists so can't include any of those. Any suggestions would be much appreciated. Thank you all.

Regards

Logged

Dave McGill

  • Member
  • ***
  • Posts: 166
    • View Profile
    • firemodel.ca
Re: Difficulty modelling sprinkler actuation times
« Reply #1 on: June 07, 2015, 05:39:35 am »

Hi,

At the moment, they are positioned at the upper surface of the domain. If you lower them (I suggest at least 1/2 cell dimension) they will appear. As you mentioned, the domain has no open vents, so you won't have a 10 MW fire for very long.

Dave
Logged
Dave McGill
firemodel.ca

dambook

  • New Member
  • *
  • Posts: 8
    • View Profile
Re: Difficulty modelling sprinkler actuation times
« Reply #2 on: June 07, 2015, 09:13:43 am »

Hi Dave

Many thanks for your reply. No issues regarding seeing the sprinkler heads etc. The domain is very large and checking the HRR file it indicates no issues with regards to maintaining that size fire for a 3 minute period. The issue is the sprinkler temperature links not reaching 68 degrees C in the sort of time they would in a real scenario, say around the 90 to 100 second mark. I was wondering if there was something I was doing wrong in terms of setting up the environment. Many thanks again.

Regards
Logged

dambook

  • New Member
  • *
  • Posts: 8
    • View Profile
Re: Difficulty modelling sprinkler actuation times
« Reply #3 on: June 09, 2015, 03:33:58 pm »

Hello all

I'm still struggling to model this scenario to an accurate enough specification resembling a realistic outcome. Am I wrong in thinking that a 10 MW fire would trigger quick response sprinklers with an RTI of around 50 much earlier than the 180 seconds or so this set up is indicating towards? Any words of wisdom or advice will be much appreciated. Thank you in advance.

Kind regards

Logged

Dave McGill

  • Member
  • ***
  • Posts: 166
    • View Profile
    • firemodel.ca
Re: Difficulty modelling sprinkler actuation times
« Reply #4 on: June 09, 2015, 06:26:05 pm »

It's a 10 MW fire with a growth ramp. How big is the fire when the first sprinkler goes off? With that ceiling height, I would expect the sprinklers to go off when the fire is at least several MW in size. The cell size is very coarse. Keep cutting your cell size in 1/2 and see if the activation times converge.

Dave
Logged
Dave McGill
firemodel.ca

dambook

  • New Member
  • *
  • Posts: 8
    • View Profile
Re: Difficulty modelling sprinkler actuation times
« Reply #5 on: June 09, 2015, 07:26:50 pm »

Hi Dave

Thank you for your reply. I have simulated a few runs with much finer mesh but still not achieving anything near realistic times to be honest. In almost every one of the runs there's no activation in the 300 seconds they are performed for.
Logged

dambook

  • New Member
  • *
  • Posts: 8
    • View Profile
Re: Difficulty modelling sprinkler actuation times
« Reply #6 on: June 10, 2015, 02:59:01 pm »

Hi again

I lowered the sprinkler heads down onto the fire load so they stand at a height of around 3 meters but still only activating at around 180 seconds! This has to be rather strange......
Logged

Dave McGill

  • Member
  • ***
  • Posts: 166
    • View Profile
    • firemodel.ca
Re: Difficulty modelling sprinkler actuation times
« Reply #7 on: June 10, 2015, 05:25:03 pm »

I think there are two issues at play.

1.  The domain is a sealed box, so a normal plume that entrains cool air and displaces the hot, less dense air to the top of the domain is not occurring.

2. The grid is too coarse.

I simplified your model into a fire, a sprinkler link, a ceiling, and open sides. I ran it with cells sizes of: 2 m, 1 m, 0,5 m, and 0.25 m. I plotted the sprinkler link temperatures.

See the graph. The temperatures converge, and indicate that a cell size of 0.5 m, or less, is required, along with a domain that is not a sealed box.

The graph and the PyroSim files can be downloaded from dropbox.

https://www.dropbox.com/sh/927zseah74zvcxz/AAAMS3oNR9A33kEVR2EWbj6Ia?dl=0

Dave
Logged
Dave McGill
firemodel.ca

dambook

  • New Member
  • *
  • Posts: 8
    • View Profile
Re: Difficulty modelling sprinkler actuation times
« Reply #8 on: June 10, 2015, 06:04:42 pm »

Hi Dave

Thank you for your response. I can't have any openings within the model as it is such a large volume of space and the worst case scenario which is being depicted here calls for all doors to be deemed as closed at time of fire and there are no ceiling vents or any other form of ventilation thus their exclusion from the model. I understand the need for finer cell sizes but still not managing the realistic results I'm looking for. I'm going to perform a few more runs with finer cell sizes again and hopefully I'll achieve more realistic results. Thank you again for the advice and time you've spent on this, appreciate that very much.

Kind regards
« Last Edit: June 12, 2015, 08:27:02 am by dambook »
Logged