Fire Spread to a solid fuel

[Timea Marton]

Dear all,

I have two boxes in an open enclosure. The first object is set to burn at a specified HRRPUA and RAMP on 5 of its surfaces (4 sides and top, not bottom).

The second object is located 80 cm from the first and has the same geometry and material properties and is set to ignite at 250 degrees and should follow a second RAMP with a specified HRR. 

The second object starts burning on one of its side surfaces, but i have noticed that the indicated HRRPUA is not achieved, as the fire spreads from cell to cell. Is it possible to set somehow the entire surface to burn when the ignition temperature is reached?

I have tried to implement SPREAD_RATE in the SURF line to try to fasten the spread but this did not work.

I have also tried by setting a lower conductivity (k) on the other 4 surfaces of the second object, to try to speed up the spread from oen surface to another.

I have copied the code i am using below.

&MATL ID ='RUBBER TYRE'
 DENSITY = 910.
 CONDUCTIVITY = 0.13
 SPECIFIC_HEAT = 1.88  
 HEAT_OF_COMBUSTION = 32000./
 REFERENCE = 'www.matbase.com/material/polymers/elastomers/natural-rubber/properties'

&SURF ID ='FIRE1'
 COLOR ='RED'
 HRRPUA = 339.052
 RAMP_Q ='fireramp1'
 MATL_ID ='RUBBER TYRE'
 THICKNESS = 0.02/

&RAMP ID='fireramp1', T=0, F=0 /
&RAMP ID='fireramp1', T=240, F=0.168 /
&RAMP ID='fireramp1', T=960, F=0.168 /
&RAMP ID='fireramp1', T=1440, F=0.662 /
&RAMP ID='fireramp1', T=1500, F=1 /
&RAMP ID='fireramp1', T=1620, F=0.542 /
&RAMP ID='fireramp1', T=2280, F=0.12  /
&RAMP ID='fireramp1', T=4200, F=0 /

&SURF ID ='FIRE2'
 COLOR ='CRIMSON'
 HRRPUA = 339.052
 IGNITION_TEMPERATURE = 250
 RAMP_Q ='fireramp2'
 MATL_ID ='RUBBER TYRE'
 THICKNESS = 0.02/

&MATL ID ='RUBBER TYRE2'
 DENSITY = 910.
 CONDUCTIVITY = 0.01
 SPECIFIC_HEAT = 1.88  
 HEAT_OF_COMBUSTION = 32000./
 REFERENCE = 'www.matbase.com/material/polymers/elastomers/natural-rubber/properties'

&SURF ID ='FIRE2.a'
 COLOR ='CRIMSON'
 HRRPUA = 339.052
 IGNITION_TEMPERATURE = 250
 RAMP_Q ='fireramp2'
 MATL_ID ='RUBBER TYRE2'
 THICKNESS = 0.02/

&RAMP ID='fireramp2', T=0, F=0 /
&RAMP ID='fireramp2', T=60, F=0.289156627 /
&RAMP ID='fireramp2', T=600, F=0.289156627 /
&RAMP ID='fireramp2', T=960, F=0.662650602 /
&RAMP ID='fireramp2', T=1020, F=1 /
&RAMP ID='fireramp2', T=1140, F=0.542168675 /
&RAMP ID='fireramp2', T=1800, F=0.120481928 /
&RAMP ID='fireramp2', T=3720, F=0 /

--Objects--
&OBST XB= 0.4,5.2,  3.2,5.0, 0.2,1.4  SURF_IDS ='FIRE1','FIRE1','INERT'/fire starts here
&OBST XB= 0.4,5.2,  0.6,2.4, 0.2,1.4  SURF_ID6 ='FIRE2.a','FIRE2.a','FIRE2.a','FIRE2','INERT','FIRE2.a'/

Any help would be greatly appreciated!

Best regards,

Timi

[dr_jfloyd]

Define the &VENT for the fire with a DEVC_ID.  Have the DEVC be the surface temperature and use STATISTICS to find the maximum over the object.   When the maximum reaches the set point, the entire VENT will turn on.

[Timea Marton]

Hi,

So it would be a line like this for the second object:

&VENT XB= 0.4,5.2,  0.6,2.4, 0.2,1.4, SURF_ID='FIRE2', DEVC_ID='FIRE2',

&DEVC XB= 0.4,5.2,  0.6,2.4, 0.2,1.4, QUANTITY='SURFACE_TEMPERATURE',SETPOINT=250, STATISTICS='MAX', ID='FIRE2'

And thank you for your quick response.

BR,

Timi

[Timea Marton]

oh sorry...WALL_TEMPERATURE instead of SURFACE_TEMPERATURE

[dr_jfloyd]

A VENT needs to be a plane.  Your VENT line is not a plane. 

For the XB for the DEVC you should make it extend a little above and below the plane of the VENT to make sure FDS finds the wall cells (See the discussion in 16.10.10).  For example:

&VENT XB=0.4,5.2,0.6,2.4,1.4,1.4,...

&DEVC XB=0.4,5.2,0.6,2.4,1.3,1.5,... 

On Sunday, August 3, 2014 5:46:04 PM UTC-4, Timea Marton wrote:

oh sorry...WALL_TEMPERATURE instead of SURFACE_TEMPERATURE

[Timea Marton]

Hi,

I have some additional questions. 

When using a burner defined as a VENT and activating it with DEVC, as in this case, is the ignition temperature dependent on the material properties defined in the SURF line? Or is it useless to specify these?

When i extend the edges on XB for the DEVC, does this mean that that the ignition temperature is calculated on a bigger surface, than my actual object?  I am asking this because it seems, if the answer is yes, that this is quite relevant when analyzing fire spread for objects at different distances. 

Thank you. 

Regards,

Timi

&MATL ID ='RUBBER TYRE'
 DENSITY = 910.
 CONDUCTIVITY = 0.13
 SPECIFIC_HEAT = 1.88  
 HEAT_OF_COMBUSTION = 32000./
 REFERENCE = 'www.matbase.com/material/polymers/elastomers/natural-rubber/properties'

&SURF ID ='FIRE1'
 COLOR ='RED'
 HRRPUA = 339.052
 RAMP_Q ='fireramp1'
 MATL_ID ='RUBBER TYRE'
 THICKNESS = 0.02/

&OBST XB=  9.2,14.0, 0.8,2.6, 0.2,1.4  /
&VENT XB=  9.2,14.0, 0.8,0.8, 0.2,1.4, SURF_ID ='FIRE2', DEVC_ID='Burner1' /
&DEVC XB=  9.2,14.0, 0.7,0.9, 0.2,1.4, QUANTITY ='WALL TEMPERATURE',SETPOINT=250, STATISTICS='MAX',   ID='Burner1'/

[dr_jfloyd]

If you are forcing ignition via a device you should not also use an ignition temperature in the material definition.

By extending the DEVC XB, I mean just a small amount on the order of half the grid spacing.

&VENT XB=0,1,0,1,2,2/

If for example, my grid spacing is 0.3 m, then this vent will actually have co-ordinates of 0,0.9,0,0.9,2.1,2.1

In this example, if my device were

&DEVC XB=0,1,0,1,2,2/

Then the search for wall cells to apply the device to will not return anything.  But if my device were

&DEVC XB=0.1,0,1,1.85,2.15/

Then it would find all the wall cells.

[Timea Marton]

Thank you for your answer. It is now clear why the extension should be applied. I have arranged my obstacles such that it fits perfectly the grid so i did not have problems with this. And i thought the extension is useless then. 

It is still not clear however, if material properties are regarded by the DEVC. It makes sense not to include the ignition temperature in the MAT line, as you specify this in the DEVC line. I was wondering about the other properties such as, conductivity, specific heat and density , if these material properties are considered when the DEVC reaches its SETPOINT? Normally the material properties should have an influence on how fast the surface reaches ignition temperature, hence will have an influence on the rate of fire spread. 

Thank you. 

[dr_jfloyd]

If I have understood you correctly, you are looking to have then entire surface burn at some predefined rate once any point on the surface has reached some ignition temperature.  In this case you should define the OBST with the heat capacity, conductivity, etc. of your fuel but not use ignition temperature.  Then have a VENT with an HRRPUA (and perhaps a RAMP) which is activated by a DEVC triggered  on the surface temperature.

If your VENT applied a surface with an ignition temperature, then you would be back to where you are now.  When the OBST reached  the ignition temperature, FDS would change the boundary condition to be the VENT with ignition temperature and you would have to wait for each cell to reach the ignition temperature.

[Timea Marton]

Yes you understood correctly. I want the entire surface to burn once a point on the surface reaches ignition temperature. So i am just curious about this first point.

When you say 'you should define the OBST with the heat capacity, conductivity, etc. of your fuel but not use ignition temperature', do you actually mean that i should add the SURF line to the OBS line or is it fine to just add it to the VENT?  

Previous example : 

&OBST XB=  9.2,14.0, 0.8,2.6, 0.2,1.4  /              
&VENT XB=  9.2,14.0, 0.8,0.8, 0.2,1.4, SURF_ID ='FIRE2', DEVC_ID='Burner1' /
&DEVC XB=  9.2,14.0, 0.7,0.9, 0.2,1.4, QUANTITY ='WALL TEMPERATURE',SETPOINT=250, STATISTICS='MAX',   ID='Burner1'/

Enter code here...

Or you mean like this?

&OBST XB=  9.2,14.0, 0.8,2.6, 0.2,1.4, SURF_IDS='INERT','INERT','FIRE2','INERT','INERT','INERT' /              

&VENT XB=  9.2,14.0, 0.8,0.8, 0.2,1.4, SURF_ID ='FIRE2', DEVC_ID='Burner1' /
&DEVC XB=  9.2,14.0, 0.7,0.9, 0.2,1.4, QUANTITY ='WALL TEMPERATURE',SETPOINT=250, STATISTICS='MAX',   ID='Burner1'/

The point where i get confused is that , when i use the DEVC and add a SETPOINT , it just seems that it would not take account of the material properties, conductivity , density etc., as i am not specifying anything like this on the DEVC line. I hope you understand my question now.

[dr_jfloyd]

The DEVC just monitors the wall temperature.  It does not compute heat transfer, it just looks at the values stored in the array of wall temperatures.

&OBST XB=..., SURF_ID='your material with no ignition temperature'/

&VENT XB=...,SURF_ID='your HRRPUA you want to use',DEVC_ID='the wall temperature device'/

&DEVC ID=...../

Until the SETPOINT is reached the OBST will use the properties defined by its SURF_ID.   When the DEVC sees that one of the wall cells has exceeded the setpoint, the VENT is placed onto the OBST and you get the HRR that you have specified.

[Timea Marton]

Hi.

I am running a different simulation now, and i want to know if there is a way to specify the rate of fire spread with SPREAD_RATE from the first ignited point.

 

So to be clear, before i had a VENT that ignited the whole surface when one point on the surface reached ignition temperature. Now i am want to specify a spread rate, but i am not sure how to code it so that the fire spreads from the first point that reaches ignition temperature. I have only managed to do it for a random point, assuming that the fire will start there. Is there any way to do it differently? 

Thank you,

Timi 

[dr_jfloyd]

You could define a large number of VENT inputs with different starting locations for the SPREAD_RATE.  Have DEVC monitor each location.  Then define some form of control system logic to turn on the first vent whose DEVC reaches the setpoint.

[Timea Marton]

Ok this would be quite time consuming as i have 5 VENT planes on an object, and in total 12 objects. Is there a way to output the location of the cell that ignites first, and then use this output for a second simulation where I dictate the spread rate?  

[dr_jfloyd]

Use smokeview to look at the wall temperature data.

[Timea Marton]

ok will do that. i think getting an approximate point should be fine. 

Thank you for your help