Requesting hint about modelling

[Kersal KK]

Hi,
I'm trying to modelling a railway tunnel fire to recreate a fire curve to test tunnel ceiling resistance.
I have recreate a train carriage with all the component inside and I have to simulate several scenarios with different ignitions.

Now, I have two ways:
1) I could assign to each solid that make up the train a material (generally Polyurethane Foam for seats, rubber for floor, etc.) and then put a burner on a seat and get the results
2) Assign to every solid a HRRPUA material previously obtained by a cone calorimeter tests. I know for each material the ignition temperature, so I could set for immediate burn a seat and then see if the others solid start to burn or not.

The second case is the simple way but IMO weak in predictive power. The first case involve a lot of more problems concerning the specific pyrolysis information about all the materials.

My objective is to find a real fire curve, which of two solution appears to be better?

[Kersal KK]

Furthermore,

if I know a HRRPUA Curve like the one in the picture and I know also the mass loss rate peak [g/s] how can I best simulate a burning object in my scenario made from the same material? I have recreat a Surf like below, but is it correct?

&SURF ID='SEAT'
     TMP_INNER=20.000,
     BACKING='EXPOSED',
     RAMP_Q='RAMP_Q_SEAT',
     MATL_ID(1,1)='SEATS',
     MATL_MASS_FRACTION(1,1)=1.00,
     THICKNESS(1)=0.10,
     HRRPUA=200.000,
     IGNITION_TEMPERATURE=352.000/
&RAMP ID='RAMP_Q_SEAT', T=0.000, F=0.750000/
&RAMP ID='RAMP_Q_SEAT', T=10.000, F=1.000000/
&RAMP ID='RAMP_Q_SEAT', T=500.000, F=0.750000/
&RAMP ID='RAMP_Q_SEAT', T=600.000, F=0.000000/




Kind regards
Pietro

[Dave McGill]

Hi Pietro,

If there is enough oxygen, and fuel, then each cell would exhibit the HRR described in your ramp. As an aside, the first values of the ramp should be T=0, F=0.

Regards

Dave

[Kersal KK]

Thanks Dave for the answer.

Aside from the T=0 mistake, my object will stop burning at T=600. But It's correct? I mean HRRPUA is a value of HRR at m² but the burning time follow the same rule? 1 m² will burn on 600sec and will produce 200kW, but 2m² will produce 400kW and will burn in 600 seconds too?

[Dave McGill]

Pietro,

Each cell surface to which the SURF has been applied will follow the ramp. (Again, assuming there is adequate oxygen and fuel; as you have put a thickness value for the fuel, the fuel could be exhausted before 600 s.)

The ignition temperature has been specified. What will be the heat source to get the fire started? This also means each individual cell surface has to reach 352 C before it will start to burn, and the fire will spread gradually over the object, and will burn longer than 600 s. The ramp "clock" will start individually for each cell when it reaches the ignition temperature. (If you had not supplied an ignition temperature, then combustion would begin on all cell surfaces at the same time, without the need for an external ignition source.)

Dave

[Kersal KK]

Thank Dave, it's perfect! It's work and the fire spread inside the train is similar (a bit higher) than the experimental one. For my purpose could be acceptable.

Now, a second question: is the HRRPPUA function affected by REAC? I mean, now i have to simulate the case that a Diesel Fire Pool starts burning outside the train and the possibility that the this fire spread inside the train. But Diesel fuel burns with a different reaction than polyurethane. Considering that my investigation is only about the higher temperature on the ceiling of a tunnel and the time that that temperature remains while all the carriage burns, have I to consider the two different REAC?

Furthermore, it's possible to consider the effect of melted aluminium train walls rather than simple put it off when it reaches 660°C ? 

[Dave McGill]

Hi Pietro,

HRRPUA is not directly affected by the REAC. The reaction determines the consumption of oxygen, and the generation of combustion products. If you had a situation where the fire was bordering on being air-limited, then the choice of reaction might affect the HRRPUA. (A reaction with high oxygen requirements might push the fire into an air-limited situation, and reduce the HRRPUA.

You could use two different reactions. It will slow down the processing considerably, so unless diesel represents a significant portion of the total fuel load, then I wouldn't bother. If you use polyurethane as your single reaction, then all fuel will consume oxygen and produce products of combustion as if they were polyurethane. 

If you haven't done so, you should consider increasing the radiative fraction from the default value of 0.35. I would guess that an appropriate value for polyurethane is closer to 0.50.

If you know the temperature at which holes will form in the aluminum walls, then you could either remove the obstructions making up the walls, or create holes based upon the temperature (adiabatic surface temperature might be the best choice).

Dave

[Kersal KK]

So i have to change my Reaction in:

&REAC ID = 'POLYURETHANE_GM27',
    HEAT_OF_COMBUSTION=23200.00,
     RADIATIVE_FRACTION=0.50,
    SOOT_YIELD=0.19800,
    CO_YIELD=0.04200,
    C=1.00000,
    H=1.70000,
    O=0.30000,
    N=0.08000/

it's right?

For the aluminium walls I had the same idea about removing but you gave me a good idea about adiabatic. Thanks again

Pietro

[Dave McGill]

Hi Pietro,

The reaction looks OK; you don't actually need to specify the HEAT_OF_COMBUSTION; with the molecular formula specified, FDS can calculate the Heat of Combustion.

Dave

[Kersal KK]

Hi Dave,

I have simulated the scenario even with Propan REAC and I have obtained very big differences. It's possible?  I have choose the Polyurethane reaction because I have a lot of seats made of polyurethane foam, rubber on the floor and PPMA on the walls; all the materials have a low flammability hazard due to the presence of fire retardant additives (all materials are classified as 1M). 

The pics attached show the temperature slice at 300s and 500s for propane and polyurethane reaction. 

what is the best way to figure out the right reaction ?

[Dave McGill]

eHi Pietro,

My understanding, which may have been incorrect, was that the only sources of fire were burners using the HRR defined by the ramp in your earlier post.

If you have a variety of potentially combustible surfaces (foam, rubber, PMMA etc.) then it is far more complex, and difficult to predict. 

You've mentioned that all of the materials have fire-resistant additives. Are the material properties, for example, Heat of Combustion, those of  "normal" polyurethane GM27 foam, or the FR variety?

Propane will have a radiative fraction of 0.35 (default) or less, while the polyurethane has a radiative fraction of 0.50. They will have different oxygen requirements. You could put an oxygen slice file down the middle of the rail-car and see if there is a difference in available oxygen.

At the end of the day, you need to pick a single reaction. If the bulk of the fuel is polyurethane GM27 FR, then I would use it's properties for the reaction, assuming you can find them.

Dave  

[Kersal KK]

Hi Dave,
you have understood well: all components of my railcar was defined by HRRPUA (seats, walls and floors) defined by different ramps ,heat of combustion, heat capacity, density and  heat of vaporization. The ignition source is a seat that is defined as "immediate burn". 

I'm investigating the problem to understand if with my FR I have the same  Heat of Combustion if I burn a single seat.

[Kersal KK]

Hi Dave,

the model with Polyurethane reaction was good. I obtained exactly the same results of experimental test (with only two seat put one in front of the others). 

Now I have to put, as ignition, the diesel pool fire. My problem is that I want to simulate even the duration of pool fire, but I have some doubt about the way to do so using HRRPUA way.

Starting from Nureg 1805 i found that a Diesel pool fire of 1m² generates an HRR of 1811,15 kw and the burning duration is 20,40 seconds for 1 litres of diesel (this means that the material have a thickness of 0,001) .But  I can I create a HRRPUA surf that burn with those properties and only for that time?

[Dave McGill]

Hi Pietro,

The simplest way is to create a burner (for example, a 1 m^2 burner with a HRRPUA of 1811 kW) with a ramp that follows the HRR you want. . However, it will burn using the existing reaction (polyurethane ?), and it is a burner, not a liquid fuel fire.

I'm not sure how a fire at track level would ignite the upholstered furniture in the rail car.

Dave

[Kersal KK]

I don't understand:

If I know that 1 m² of diesel (10 litres) burns in 20,4s and give an HRR=1811kw, if I create a surface of 10 m² will burn in 204sec ? Or I have to create a new burner ?

About the train: the diesel  fuel is stored on a thin steel tank under the carriage. The floor of carriage is made as sandwich: a steel sheet, then  polystyrene foam for acoustic purpose, then polyester fiber, and then rubber at the end. 

The scenario might be related to a loss of diesel fuel that creates a pool that for some reason it triggers ( for example, because in contact with hot brakes: but even this is unlikely because the tank is installed away from the wheels ) . In this way the flames outside could, through the windows or to the collapse of the walls in aluminum, come into contact with the material inside .

Another scenario could be related to a derailment which causes the breakage of both the tank ( steel and far from the tip of the train , unlikely ) and of the walls of the train .

However , my goal is to see which could be the stress of the fire of the gallery walls in the worst case .

[Dave McGill]

Pietro,

If you create a 1 m^2 burner with a HRRPUA of 1811 kW and use a ramp so that it burns out in 20.4 s, then enlarge the burner to 10 m^2, it will reach a HHR of 1811 kW x 10, but it will still burn out in 20.4 s. (However, this has nothing to do with diesel as a fuel; it is a burner on a ramp, using the polyureythane reaction.)

Ignore my earlier question about the diesel as an ignition source. I had forgotten that all of the contents of the railcar were burners on ramps.

Dave

[Dave McGill]

Pietro,

To clarify my earlier response, I'm assuming the burner is created with a single vent.

Dave

--
You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+u...@googlegroups.com.
To post to this group, send email to fds...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/e03a9416-f35d-4225-b108-d2b6e41dfb9f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Kersal KK]

Yes, i understood. Now my PC is calculating... tomorrow i will check if i did all the right thing.

Thanks always 

Pietro