In calculations related to simulating of fire development in a room, the precise thermophysical parameters of the enclosing structures may not be known. In this article we show the use of field simulation to compare the temperature dynamics and thermal radiation levels during fires in rooms of identical size, with walls made of inert material, brick, polished steel, and glass.
This experiment has demonstrated that the temperature and radiation dynamics in a room with a combustion source are primarily influenced by the radiation reflectivity coefficient of the walls. The experiment has also shown that opaque materials almost completely block the transfer of heat to adjacent rooms. The density, heat capacity, and thermal conductivity of the wall materials minimally impact the dynamics of air temperature and thermal radiation.
Introduction
Integral and zone models are widely used to study the dynamics of dangerous factors in fire scenarios [1]. Although these models do not fully utilize information about room geometry, wall materials, etc., the results are generally acceptable. Increased computational power allows for the use of more sophisticated field modeling methods for fire calculations [1]. Generally, a field model enables the inclusion of numerous parameters of the simulated scene in the calculations. Certain parameters that characterize material combustion are listed in reference books [2]. However, the thermophysical parameters of the walls often cannot be determined with high precision. It could have scientific value to evaluate how significantly these parameters impact the results of the simulation.
Thermal conductivity, heat capacity, density, emissivity, and material absorption rates influence the rate at which walls heat up because of thermal radiation and from hot air. Thus, wall materials affect the heat loss from the room through the walls. The thermal energy inside a room is characterized by the air temperature and the intensity of thermal radiation. Thus, when you set a specific combustion intensity, wall materials affect only air temperature and thermal radiation, and do not influence the release of soot and toxic reaction products.
This study explores the impact of wall material properties on the dynamics of temperature and thermal radiation inside a room during a fire, using a field model. The simulation was conducted using the FDS 6.5.3 environment [3].
Simulated Materials
Heat loss through walls from a room with an active fire occurs via two heat transfer mechanisms: heat radiation and convective heat exchange between gas flows and solid surfaces. Both mechanisms become increasingly significant as the internal temperature rises. Thus, at the initial stages of a fire, when temperatures are relatively low [4], wall materials exert minimal influence on the dynamics of temperature and thermal radiation.
We have decided to simulate combustion using typical room materials, “furniture + household products” rated for low to medium fire resistance [2].
Wall materials primarily differ in terms of density, thermal conductivity, heat capacity, emissivity (ε), and absorption. To succinctly yet effectively compare them, the following materials were selected: inert, brick, polished steel, and transparent glass. The “inert” material is an idealized concept, consistently at 20°C, opaque, and has an emissivity of 0.9. Since this material’s temperature remains constant, its density, thermal conductivity, and heat capacity are irrelevant. It is expected to facilitate the greatest heat loss, acting as a coolant. The term “inert” will henceforth be used without quotes. In the FDS system, this material is the default for walls; it can be explicitly defined with the property SURF_ID=‘INERT’.
In FDS, the default Absorption Coefficient (ABSORPTION_COEFFICIENT) is 50,000 m-1, indicating that the material is opaque unless otherwise specified.
The “brick” material possesses the following parameters: density = 1800 kg/m3, thermal conductivity = 0.87 W/(mK), heat capacity = 0.88 kJ/(kgK), and ε = 0.9. It is opaque. In FDS, this is achieved by covering all walls, partitions, floors, and ceilings with a surface possessing the following properties:
&MATL ID = 'BRICK', EMISSIVITY = 0.9, DENSITY = 1800, CONDUCTIVITY = 0.87, SPECIFIC\_HEAT = 0.88/
SURF ID='BRICK WALL', MATL\_ID='BRICK', BACKING='EXPOSED', THICKNESS=0.25 /
The “polished steel” material has the following characteristics: density = 7800 kg/m3, thermal conductivity = 47 W/(mK), heat capacity = 0.5 kJ/(kgK), and ε = 0.1. It is opaque and serves as an example of a material with a mirror-like surface. In FDS, reflectivity is calculated as 1-ε. The properties are defined as follows:
&MATL ID = 'STEEL', EMISSIVITY = 0.1, DENSITY = 7800, CONDUCTIVITY = 47, SPECIFIC\_HEAT = 0.5/
&SURF ID='STEEL WALL', MATL\_ID='STEEL', BACKING='EXPOSED', THICKNESS=0.02 /
The “transparent glass” material parameters are: density = 2500 kg/m3, thermal conductivity = 1 W/(mK), heat capacity = 0.8 kJ/(kgK), and ε = 0.9. It is transparent with an absorption rate of 0 m-1. Its properties in FDS are defined as follows:
&MATL ID = 'GLASS', EMISSIVITY = 0.9, DENSITY = 2500, CONDUCTIVITY = 1, SPECIFIC\_HEAT = 0.8, ABSORPTION\_COEFFICIENT = 0/
&SURF ID='GLASS WALL', MATL\_ID='GLASS', BACKING='EXPOSED', THICKNESS=0.02 /
This material set includes a range of thermal conductivities, different emissivities, both opaque and transparent materials, and a conventional material consistently at a constant temperature. It is noteworthy that a material with an emissivity (ε) absorbs exactly the same amount of ε, of the incident radiation it receives.
Description Of The Premises
Let’s consider a premises with internal dimensions of 5x15 meters, divided by partitions into three rooms measuring 5x4.75 meters, 5x5 meters, and 5x4.75 meters respectively. The partitions between the rooms, as well as the exterior walls separating the room from the outdoors, are 0.25 meters thick for “inert” and “brick” materials, and 0.02 meters thick for “polished steel” and “transparent glass” materials. An open doorway, 1 meter wide and 2 meters high, is located in the middle of one of the partitions. The outer rooms feature open doorways to the outside. The floor plan is depicted in Image 1. For convenience, the rooms are numbered from left to right.
In the central room (2), along a solid partition and 0.5 meters above the floor, there is a fire source measuring 2x1 meters. In the middle of each wall in every room, at a height of 1.7 meters, a temperature measuring device is positioned. Additionally, at points in the room marked with black dots and labeled with lowercase letters on Image 1, there are temperature and heat radiation measuring devices. All measuring devices are located at 1.7 meters high. Each heat radiation measuring device is equipped with four directional radiation sensors, oriented to the cardinal directions.
We will simulate the burning for 300 seconds.
Analysis of the Results
Wall temperature
Let’s discuss the temperature dynamics of walls for different materials. It’s important to note that with the inert material, the wall temperature always remained at 20°C.
In room 1, our focus is only on the eastern wall, as it acts as a partition with room 2, where combustion occurs. The highest temperature, 37°C, was recorded for the steel wall due to its high thermal conductivity.
In the other rooms, the walls reached their highest temperatures when made of brick. This is because brick has both low thermal conductivity and high emissivity, which translates to a high absorption coefficient. Meanwhile, polished steel, which predominantly reflects thermal radiation, heats up less: it conducts heat quickly and cools down rapidly. Glass in our model does not absorb radiation at all, and thus also heats up slowly.
The material of the walls does influence their temperature during a fire; however, wall temperature itself is not considered a hazardous factor. Further we will demonstrate that the wall temperature is not indicative of the air temperature within the room.
Air temperature
Now, let’s analyze the air temperature graphs at various points. In room 1, the temperature did not significantly deviate from 20°C for any wall material, indicating that even a thin, airtight partition effectively shields against high temperatures.
In room 2, the temperature graphs at points a,b,c,d,e,f are very close to each other. A similar closeness is observed in room 3 at the same points. Therefore, to streamline the presentation, we will only consider the temperature at point “c” in rooms 2 and 3.
Image 2 shows the temperature graphs at point “c” in room 2 for simulations with different materials. The air was hottest in the simulation with polished steel walls. In other cases, the temperatures were nearly identical. The threshold of 70°C was exceeded in all four scenarios after approximately the same duration: around 60 seconds.
Temperature graphs in the doorway between rooms 2 and 3, as well as at point “c” in room 3, exhibited similar dynamics. It is noteworthy that the temperature in the doorway was slightly higher than at point “c” in room 2. However, the temperature at point “c” in room 3 was significantly lower than in the doorway.
The significant rise in temperature in the simulation with polished steel walls is due to the mirror surface trapping radiation within the room, effectively disabling one of the two heat removal mechanisms. The approximate equality of the graphs for the other materials is due to these materials allowing radiation to escape more effectively from the room. As the simulation shows, it is not critical whether the radiation was absorbed by the wall (inert material and brick) or passed through (glass). Nor does it matter much if the wall material itself heats up significantly. It was previously mentioned that the temperatures of walls made from different materials varied greatly, but they had almost no impact on the air temperature in the rooms.
Furthermore, although the walls made of polished steel were not the hottest, the air temperature in this simulation was the highest. The primary influence on air temperature in the room comes from the walls’ reflectivity. In the FDS system, the reflectivity is calculated as 1-ε, with polished steel having a reflectivity of 0.9, compared to 0.1 for other materials.
Thermal radiation
Now let’s look at the thermal radiation graphs. A noticeable difference from the background radiation values in room 1 at points a and b was only seen with glass walls (Image 3).
This radiation came from the fire source, having passed through the wall. Glass is the only material among those considered where the thermal radiation from a combustion source behind a partition can be significant. For the other, opaque materials, the partition did not heat up enough for its own radiation to become hazardous.
In the other rooms, we consider the heat radiation from the west, as that is where the combustion occurs. The radiation flux at points a,b,c,d,e,f varies. The highest flux is always observed at point “c”, as it is closest to the fire, and the measuring device there directly faces the flame. Image 4, as an example, shows the graphs of the heat radiation flux at six points in room 2 for a simulation with brick walls. Moving forward, to reduce the amount of illustrative material, we will compare fluxes only at points “c”.
Image 5 displays the graphs of the thermal radiation flux at point “c” in room 2 for simulations with different materials.
The radiation power in the simulation with mirrored walls was more than twice that of the other simulations. The roughly equal radiation power in simulations with non-mirrored materials is mainly determined by their equal reflectivity. Differences in material temperatures only caused slight variations, as the average wall temperature is much lower than that of the fire source and the combustion area, which emits the hottest radiation. Thus, the main source of heat radiation during a fire is the fire itself, and the temperature of the walls plays a relatively minor role. However, the reflectivity coefficients of wall materials are extremely important.
The same characteristic relationships for the heat radiation flux were observed in the doorway as in the previous image. The value of 1.4 kW/m2 in a doorway with steel walls was exceeded after 70 seconds, while with inert, brick, and glass walls it was not exceeded at all.
Image 6 shows the graphs of the thermal radiation flux at point “c” in room 3. Here, in the case of steel walls, the radiation was the strongest, as in room 2.
However, the graphs for the other three materials diverge slightly more than in room 2, because in the case of opaque materials, the partition blocks a significant part of the radiation.
The radiation flux in room 3 when the walls are made of glass exceeds the flux for the case with brick walls by about a third, on average. This is due to the transparency of the glass partition between rooms 2 and 3. However, in the case of polished steel walls, where radiation was trapped in the room, the heat flux was much higher than with glass walls, despite the partition being opaque and rooms 2 and 3 being connected only by a doorway.
Conclusion
Comparing four fundamentally different wall materials for the same room configuration has led us to conlude that wall reflectivity significantly impacts both temperature and thermal radiation flux in a room with an fire source (and in adjacent rooms with open doors). Other material properties have minimal impact. In the FDS system, you can use the EMISSIVITY parameter to set a material’s emissivity and reflectivity is calculated as 1-EMISSIVITY. If the emissivity of the material used in the simulation corresponds to reality, then using the properties of an inert material is generally acceptable.
In rooms separated from the fire source by an opaque partition of any significant thickness, and without open doorways connecting them, the choice of wall materials is essentially irrelevant.
Transparent walls affect temperature and radiation inside a room with an fire source almost as little as inert materials. However, they significantly influence radiation in adjacent rooms. It is important to ensure that the transparency of these materials is accounted for in simulations. In the FDS system, this is done by setting the material property ABSORPTION_COEFFICIENT to 0 (1/m). A realistic value can be used if known.
In scenarios lacking transparent or mirrored walls, it is often reasonable to assume wall emissivity is close to 0.9 and use the INERT material. The simulation demonstrated that the difference in temperature dynamics and thermal radiation flux between models using inert and brick materials does not exceed 20% throughout the simulation period.
The study simulated the combustion of a material with moderate heat release and smoke production. In scenarios involving the burning of high-energy materials, where the air temperature or concentration of hazardous substances exceeds levels safe for humans mere seconds after combustion - before the walls can significantly heat up — the influence of wall materials on temperature and thermal radiation becomes largely irrelevant.
Literature
- Order of the Ministry of Emergency Situations of the Russian Federation of June 30, 2009 No. 382 “On approval of the methodology for determining the calculated values of fire risk in buildings, structures and structures of various classes of functional fire hazard” (as amended on December 12, 2011 No. 749, and December 2, 2015 No. 632). Appendix No. 6. The procedure for calculations and mathematical models for determining the time for escape routes being blocked by dangerous fire factors;
- Koshmarov, Yu. A. Predicting Hazardous Fire Factors on Premises: A Textbook. Moscow: Academy of the State Fire Service of the Ministry of Internal Affairs of Russia, 2000, 118 p (in Russian);
- https://pages.nist.gov/fds-smv/;
- Koshmarov, Yu. A. A Mathematical Model of the Initial Stage of Fire in Premises upon Ignition of Combustible Liquid, Fire and Explosion Risk, Vol. 10, No. 5, 2004, pp. 70-80 (in Russian).