Simulated temperature much lower than experiment

[GetPython2020]

Hi,

I am new to FDS, and I am facing an issue when using PyroSim to simulate a very simple fire situation:

1. Open space, a pile of wood pallets (6ft tall, with HRRPUA 1000 kW/m2), peak HRR is about 7,500 kW/m2, fire burns on all surfaces of the wood pile
2. REAC: Radiative Fraction is set to 0.7 (set it a high value to test the temperature change, when using default value, the temperature only change 2-3C)
3. Two temperatures are measured: one is placed 15 feet away from wood pallet edge (measure air temperature), and another one is 30 feet away to measure a adiabatic wall temperature, both at a height of the wood pallets top
4. The mesh is 0.5ft0.5ft0.5ft, simulation time is 230s

Question is, both measured temperature are below 30.C, which is much lower than the experiment temperature 48 C. Anyone has ideas on how to set up the simulation to better reflect the experiment? Thank you in advance.

[drjfloyd]

No one knows what experiment your are trying to replicate
No one knows exactly what inputs you actually specified in your FDS input file.

There is little that anyone can do to make specific suggestions.

Are you using FDS outputs that reflect the actual measurements made in an experiment? (TEMPERATURE is not likely to reflect the actual experimental measurement)?
Are you accurately representing the experiment?
Have you verified you are getting the fire you intended?
Have you done a grid study?

[GetPython2020]

Thank you for the response. I am trying to replicate the experiment: burn a stack of wood pallet and measure the temperature on a wall 30 feet away, with a 10mph wind blowing the fire towards the shed. The distance in the following picture is 15 feet, but I did the same experiment with 30 feet.

I re-ran the simulation this morning and get the following results:



Temperature at the bottome middle of the wall:

Air temperature 15ft away:

I think this time my simulation might be correct, I used a finer mesh and a 60s simulation took over 2 hours. I may just need to run a longer time.
Really appreciate if you can review the following FDS, or any advice will be appreciated. Thank you.

&HEAD CHID='SDS_simulation_3'/
&TIME T_END=60.0/
&DUMP DT_RESTART=300.0, DT_SL3D=0.25/

&MESH ID='Mesh01', IJK=140,40,30, XB=0.0,14.0,0.0,4.0,0.0,3.0/

&SPEC ID='SFPE WOOD_PINE_fuel', FORMULA='C1.0H1.7O0.83'/

&REAC ID='SFPE WOOD_PINE',
FYI='SFPE Handbook, 5th Ed, Tables 36.11 and A.39 "pine"',
FUEL='SFPE WOOD_PINE_fuel',
CO_YIELD=5.0E-3,
SOOT_YIELD=0.03,
HEAT_OF_COMBUSTION=1.24E+4,
RADIATIVE_FRACTION=0.4/

&DEVC ID='AirTemp', QUANTITY='TEMPERATURE', XYZ=6.096,1.524,2.1336, ORIENTATION=-1.0,0.0,0.0/
&DEVC ID='WallTemp_Top', QUANTITY='WALL TEMPERATURE', XYZ=10.668,1.331117,2.154229, IOR=-1/
&DEVC ID='Radiant Heat Flux_Middle', QUANTITY='RADIATIVE HEAT FLUX GAS', XYZ=6.096,1.524,1.524, ORIENTATION=-1.0,0.0,0.0/
&DEVC ID='WallTemp_Bottom', QUANTITY='WALL TEMPERATURE', XYZ=10.668,1.413805,0.3048, IOR=-1/
&DEVC ID='Radiant Heat Flux_Wall Front', QUANTITY='RADIATIVE HEAT FLUX GAS', XYZ=10.6,1.524,1.524, ORIENTATION=-1.0,0.0,0.0/

&MATL ID='YELLOW PINE',
FYI='Quintiere, Fire Behavior - NIST NRC Validation',
SPECIFIC_HEAT=2.850001,
CONDUCTIVITY=0.139999,
DENSITY=640.000004/
&MATL ID='GYPSUM',
FYI='NBSIR 88-3752 - ATF NIST Multi-Floor Validation',
SPECIFIC_HEAT=1.09,
CONDUCTIVITY=0.17,
DENSITY=930.0/
&MATL ID='INSULATION',
FYI='Isolatek BLAZE-SHIELD DC/F - WTC FDS5 Validation',
SPECIFIC_HEAT_RAMP='INSULATION_SPECIFIC_HEAT_RAMP',
CONDUCTIVITY_RAMP='INSULATION_CONDUCTIVITY_RAMP',
DENSITY=208.0/
&RAMP ID='INSULATION_CONDUCTIVITY_RAMP', T=20.0, F=0.05/
&RAMP ID='INSULATION_CONDUCTIVITY_RAMP', T=377.0, F=0.1/
&RAMP ID='INSULATION_CONDUCTIVITY_RAMP', T=677.0, F=0.2/
&RAMP ID='INSULATION_SPECIFIC_HEAT_RAMP', T=20.0, F=0.8/
&RAMP ID='INSULATION_SPECIFIC_HEAT_RAMP', T=677.0, F=2.0/

&SURF ID='ADIABATIC',
COLOR='GRAY 80',
ADIABATIC=.TRUE./
&SURF ID='WoodFire',
COLOR='RED',
HRRPUA=433.644,
BURN_AWAY=.TRUE.,
MATL_ID(1,1)='YELLOW PINE',
MATL_MASS_FRACTION(1,1)=1.0,
THICKNESS(1)=1.2/
&SURF ID='WoodShed',
RGB=146,202,166,
MATL_ID(1,1)='GYPSUM',
MATL_ID(2,1)='INSULATION',
MATL_ID(3,1)='GYPSUM',
MATL_MASS_FRACTION(1,1)=1.0,
MATL_MASS_FRACTION(2,1)=1.0,
MATL_MASS_FRACTION(3,1)=1.0,
THICKNESS(1:3)=0.03048,0.03048,0.03048/
&SURF ID='Wind',
RGB=200,248,34,
HEAT_TRANSFER_COEFFICIENT=0.0,
VEL=-4.5/

&OBST ID='Wood Pallets', XB=0.3,1.52,0.9,2.12,0.0,1.8, RGB=252,2,7, SURF_ID='WoodFire'/
&OBST ID='Floor', XB=10.668,13.1064,0.0,3.048,0.0,0.1524, SURF_ID='WoodShed'/
&OBST ID='Ceiling', XB=10.668,13.1064,0.0,3.048,2.7432,2.8956, SURF_ID='WoodShed'/
&OBST ID='Wall', XB=10.7,10.9,0.2,2.8,0.0,2.8, SURF_ID='WoodShed'/
&OBST ID='Wall', XB=10.7,13.1,0.0,0.2,0.0,2.8, SURF_ID='WoodShed'/
&OBST ID='Wall', XB=10.7,13.1,2.8,3.0,0.0,2.8, SURF_ID='WoodShed'/
&OBST ID='Wall', XB=12.9,13.1,0.2,2.8,0.0,2.8, SURF_ID='WoodShed'/

&VENT ID='Mesh Vent: Mesh01 [XMAX]', SURF_ID='OPEN', XB=14.0,14.0,0.0,4.0,0.0,3.0/
&VENT ID='Mesh Vent: Mesh01 [XMIN]', SURF_ID='Wind', XB=0.0,0.0,0.0,4.0,0.0,3.0/
&VENT ID='Mesh Vent: Mesh01 [YMAX]', SURF_ID='OPEN', XB=0.0,14.0,4.0,4.0,0.0,3.0/
&VENT ID='Mesh Vent: Mesh01 [YMIN]', SURF_ID='OPEN', XB=0.0,14.0,0.0,0.0,0.0,3.0/
&VENT ID='Mesh Vent: Mesh01 [ZMAX]', SURF_ID='OPEN', XB=0.0,14.0,0.0,4.0,3.0,3.0/
&VENT ID='Mesh Vent: Mesh01 [ZMIN]', SURF_ID='ADIABATIC', XB=0.0,14.0,0.0,4.0,0.0,0.0/

&TAIL /

[Er9y714]

The experiment has a high wind speed and this might deviate the results significantly. Much larger domain sizes than the shown model are required to accurately capture the wind induced flow. This alone might skew the results if not other parameters also do not contribute.

[GetPython2020]

@drjfloyd I only measured temperature during the experiment, I also have a thermal camera recording the whole process. In this way, do you think a temperature slice at X=... would be enough?

[drjfloyd]

What specific device did you use to measure the temperature?

[GetPython2020]

I used type K thermocouple.

[drjfloyd]

Look at the User's Guide table of contents for the chapter on Outputs.

[GetPython2020]

Thank you for all the advice, I'll definitely read the user's guide and keep revising my model.