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: Simulate a fire by using spread rate  (Read 9054 times)

Sofia

  • Member
  • **
  • Posts: 12
    • View Profile
Simulate a fire by using spread rate
« on: March 25, 2013, 09:09:16 am »

Hi,

I would like to simulate a fire by using the spread rate function. The fire will peak at a HRR of 1300 kW at t=160 s, during the next 60 s the HRR will be kept constant and after that the HRR will be reduced to 433 kW during the following 60 s and then the fire will be kept constant at this value.  I tried to use the ramp-up time function, see the values in the attached file. The first 160 s turned out as intended, the following part of the fire development did not turned out as intented, please see the the HHR curve in attached file. As you can see the reduction of the HRR accures after approximately 160 s and not 60 s as intentional.

Is there anyone who knows what I'm doing wrong?

I am grateful for all the help I can get.

Kind regards,

Sofia
Logged

Charlie Thornton

  • Thunderhead
  • *****
  • Posts: 851
    • View Profile
Re: Simulate a fire by using spread rate
« Reply #1 on: March 25, 2013, 10:18:14 am »

I tried to reproduce the error using a model with a 1m^2 100kW burner with the same ramp pattern (though, the times were much quicker) and I wasn't able to reproduce the sluggish decrease in HRR coming down from the peak. There might be some physical reason that it comes down slowly, but I don't know. Are you able to share your PSM or FDS file so we can try to reproduce/understand the issue with exactly the same data?
Logged

Dave McGill

  • Member
  • ***
  • Posts: 134
    • View Profile
    • firemodel.ca
Re: Simulate a fire by using spread rate
« Reply #2 on: March 25, 2013, 11:29:22 am »

Hi Sofia,

The ramp you have described will jump to full strength in the first second, hold steady at that rate until 220 s, and then decline to 1/3 of its strength between 220 s and 280 s. It remain at 1/3 strength for the rest of the simulation.

To duplicate the fire you have described, the ramp settings should be as follows.

0,0
160,1
220,1
280, 0.3
900, 0.3

The growth and decline will be linear.

Dave
Logged
Dave McGill
firemodel.ca

Dave McGill

  • Member
  • ***
  • Posts: 134
    • View Profile
    • firemodel.ca
Re: Simulate a fire by using spread rate
« Reply #3 on: March 25, 2013, 11:31:07 am »

Charlie,

Is there an issue with her using commas instead of periods to indicate decimals?

Dave
Logged
Dave McGill
firemodel.ca

Charlie Thornton

  • Thunderhead
  • *****
  • Posts: 851
    • View Profile
Re: Simulate a fire by using spread rate
« Reply #4 on: March 25, 2013, 02:29:33 pm »

Commas are used in the UI as decimal separators when PyroSim has identified that they are locale appropriate. If everything is working properly, the FDS input file will have:

Code: [Select]
&RAMP ID='my_ramp', T=160.0, F=1.0/
...even though the UI is using commas as the decimal separator. Still, it would be nice to see the FDS input file because we have occasionally messed that up.
Logged

Sofia

  • Member
  • **
  • Posts: 12
    • View Profile
Re: Simulate a fire by using spread rate
« Reply #5 on: March 27, 2013, 06:28:39 am »

The FDS file for my fire scenario was to big, so I created a smaller file with a larger grid, which I've attached.
The concept for this fire is the same; The fire with max HRR of 1440 kW will peak at t=175 s. It is held constant for 60 s and then reduced to 1/3 the following 60 s, and after that the fire will be held constant at this value. I have the same problem with this fire, the reduction of the HRR occures to late (after approx 300 s instead of 235 s).
Logged

Dave McGill

  • Member
  • ***
  • Posts: 134
    • View Profile
    • firemodel.ca
Re: Simulate a fire by using spread rate
« Reply #6 on: March 27, 2013, 07:40:59 am »

Hi Sofia,

The problem is the combination of a ramp and a spread-rate. When you have just a ramp, all the cells that comprise the burner follows the timeline of the ramp beginning at t=0. When you combine a ramp with a spread-rate the ramp clock starts on each cell individually as the spreading fire reaches it. For example, if the cell furthest away from your fire is ignited at 60 s, then the ramp sequence for that particular cell will begin then, and 60 s is added to all the time values of the ramp for that particular cell.


In addition, the ramp is still not specified correctly. Your file contains,

&RAMP ID='Fire_RAMP_Q', T=0.0, F=0.0/
&RAMP ID='Fire_RAMP_Q', T=1.0, F=1.0/  this causes the ramp to peak at 1 s
&RAMP ID='Fire_RAMP_Q', T=175.0, F=1.0/
&RAMP ID='Fire_RAMP_Q', T=235.0, F=1.0/
&RAMP ID='Fire_RAMP_Q', T=295.0, F=0.33/
&RAMP ID='Fire_RAMP_Q', T=900.0, F=0.33/


Dave
« Last Edit: March 27, 2013, 07:47:48 am by Dave McGill »
Logged
Dave McGill
firemodel.ca

Sofia

  • Member
  • **
  • Posts: 12
    • View Profile
Re: Simulate a fire by using spread rate
« Reply #7 on: March 27, 2013, 08:01:52 am »

Thanks Dave. Ok. So it's not possiple to combine the spread rate function with ramp up?  For example, if I would like to simulate the kind of fire I earlier described, which reduces after a particular time I could'nt use the spread rate function?
Logged

Dave McGill

  • Member
  • ***
  • Posts: 134
    • View Profile
    • firemodel.ca
Re: Simulate a fire by using spread rate
« Reply #8 on: March 27, 2013, 09:14:06 am »

Hi Sofia,

You can combine the spread rate with the ramp, the problem is that you can't prescribe the overall HRR easily. If following an exact HRR is important, then I  suggest you forget about the spread rate. In your sample, the fire is barely more that 1 m^2. Does it really matter if the fire is applied to that surface uniformly instead of spreading?

Dave
Logged
Dave McGill
firemodel.ca