Jesse S. Lozano

An Investigation of Crown Fuel Bulk Density Effects on the Dynamics of Crown Fire Initiation in Shrublands 1

Crown fire initiation is studied by using a simple experimental and detailed physical modeling based on Large Eddy Simulation (LES). Experiments conducted thus far reveal that crown fuel ignition via surface fire occurs when the crown base is within the continuous flame region and does not occur when the crown base is located in the hot plume gas region of the surface fire. Accordingly, the focus in this article is on crown fuel ignition when the crown base is situated within the intermittent flame region. In this region, the flame shape and height changes with time over the course of pulsation. This causes the flame to impinge on the crown fuel base and the hot gas is forced through the crown fuel matrix. Under certain conditions, it is observed that the crown fuel bulk density affects the impingement of flame and the ignition of crown fire. The crown fuel properties used were estimated for live chamise (Adenostoma fasciculatum) with a fuel moisture content of 44% (dry basis). As the crown fuel bulk density is increased from 0.75 kg·m−3 to 1.75 kg·m−3, it is observed that the average hot gas velocity inside the crown matrix decreases from 0.70 m·s−1 to 0.52 m·s−1, thus, resulting in less entrained air passing through the crown fuel and more energy accumulation inside the crown fuel matrix. Higher bulk density also influences the surface fire. As the hot gas flows into the crown fuel matrix is retarded, the average hot gas temperature at the crown fuel base increases from 768 K to 1,205 K. This is because the mixing rate of air and combustible gas around the base of crown fuel increases. Although higher fuel bulk density means more fuel must be heated, the increase in accumulated energy per unit volume within the crown fuel matrix is higher than the additional heat needed by the fuel. Thus, the average crown fuel temperature increases and ignition occurs at higher bulk density.

Fluid Dynamic Structures in a Fire Environment Observed in Laboratory-Scale Experiments

Particle Image Velocimetry (PIV) measurements were performed in laboratory-scale experimental fires spreading across horizontal fuel beds composed of aspen (Populus tremuloides Michx) excelsior. The continuous flame, intermittent flame, and thermal plume regions of a fire were investigated. Utilizing a PIV system, instantaneous velocity fields for the three regions were measured and special attention was given to the coherent fluid dynamic structures that are present in a propagating fire environment. Measurements were performed inside the fire itself and in the surrounding environment. From the PIV data the formation of vortex structures in front of the fire were observed. For the 3 flame regions, instantaneous velocity field data was analyzed to determine existing vortex diameters and vorticity values. The presented results of the detailed and measured velocity field within a propagating fire are likely the first of its type.

EXPERIMENTAL AND NUMERICAL MODELING OF SHRUB CROWN FIRE INITIATION

The transition of fire from dry surface fuels to wet shrub crown fuels was studied using laboratory experiments and a simple physical model to gain a better understanding of the transition process. In the experiments, we investigated the effects of varying vertical distances between surface and crown fuels (crown base height), and of the wind speed on crown fire initiation. The experimental setup was designed to model an isolated clump of crown fuel such as a single tree or group of shrubs. Three wind velocities (0, 1.5, and 1.8 m · s−1) and three crown base heights (0.20, 0.30, and 0.40 m) were used. Crown fuel (solid) and the air temperature within the elevated fuel bed were measured. Crown bulk density and fuel moisture content were held constant in all the experiments. As crown base height increased, crown fire initiation success decreased. Non-zero wind speeds reduced crown fire initiation success because of reduced heating. A simple physical model based on convective and radiative heat exchanges was developed to predict crown fire initiation above a surface fire. The predicted results for different wind speeds and crown base heights were in good agreement with the experimental measurements. Because of its relative simplicity and inclusion of basic physics, it is anticipated that the model can be readily applied and/or adapted to model diverse fuel configurations.

Experimental modelling of crown fire initiation in open and closed shrubland systems

The transition of surface fire to live shrub crown fuels was studied through a simplified laboratory experiment using an open-topped wind tunnel. Respective surface and crown fuels used were excelsior (shredded Populus tremuloides wood) and live chamise (Adenostoma fasciculatum, including branches and foliage). A high crown fuel bulk density of 6.8 kg m-3 with a low crown fuel base height of 0.20 m was selected to ensure successful crown fire initiation. Diagnostics included flame height and surface fire evolution. Experimental results were compared with similar experiments performed in an open environment, in which the side walls of the wind tunnel were removed. The effect of varying wind speed in the range 0-1.8 m s-1, representing a Froude number range of 0-1.1, on crown fire initiation was investigated. The suppression of lateral entrainment due to wind tunnel walls influenced surface fire behaviour. When wind speed increased from 1.5 to 1.8 m s-1, the rate of spread of surface fire and surface fire depth increased from 5.5 to 12.0 cm s-1 and 0.61 to 1.02 m. As a result, the residence time of convective heating significantly increased from 16.0 to 24.0 s and the hot gas temperature at the crown base increased from 994 to 1141 K. The change in surface fire characteristics significantly affected the convective energy transfer process. Thus, the net energy transfer to the crown fuel increased so the propensity for crown fire initiation increased. In contrast, increasing wind speed decreased the tendency for crown fuel initiation in an open environment because of the cooling effect from fresh air entrainment via the lateral sides of surface fire.

CFD Study of Huge Oil Depot Fires - Generation of Fire Merging and Fire Whirl in (7 x 7) Arrayed Oil Tanks

One of the largest industrial fire disasters may occur in oil tank depots which store large amounts of oil. Many previous studies on the fire safety of oil tank depots have been related to the fire propagation from one single oil tank fire to the adjacent tank via radiation. However, single oil tank fire may cause a fire whirl in windy conditions, entraining much more ambient air and enhancing flame radiation, which may increase the possibility of fire propagation toward the neighboring tanks. In addition, when an oil depot storing a large amount of oil in tanks is subject to destructive earthquakes, merging fires and fire whirls may be generated, leading to disastrous consequences. In this work, the authors examined the fire merging and fire whirl behaviors in multiple huge oil tank fires by CFD simulations using FDS v4. The constant heat release rate model was employed and the effects of tank-to-tank distance, wind speed and heat release rate were examined. It was found that these parameters are important to cause fire merging and fire whirls, and at the same time, the conditions to cause fire merging and fire whirls lie in a limited range. Some relevant correlations were established. The results are expected to be useful for mitigating the disasters due to fire merging and fire whirls.

Experimental Investigation of the Velocity Field in a Controlled Wind-aided Propagating Fire Using Particle Image Velocimetry

Experiments on wind-aided firespread across an array of small diameter discrete fuel elements using a Particle Image Velocimetry (PIV) and thermocouple system were carried out in a specially designed wind tunnel at Northrop Grumman Space Technology. The rate of spread is determined using data gathered from thermocouples that were uniformly spaced within the fuel elements along the primary fire spread direction. The fuel consisted of a two dimensional array of identical, evenly spaced, wooden elements, positioned vertically in holes in a ceramic substrate. Utilization of this fuel bed allowed for repeatability of experiments. The effect on firespread rate was investigated for two different fuel loadings. A PIV system was used to investigate the various fluid dynamic structures present within a propagating fire front, and how firespread was influenced. From analysis of the thermocouple and PIV data it was determined that in the high fuel loading situation (total dry fuel mass per unit fuel bed area, m = 3.12 kg/m 2 ) firespread was assisted by intense forced turbulent convection, which preheated the fuel elements downwind of the fire front. For the low fuel loading case (m = 0.78 kg/m 2 ) radiation played a more significant role in preheating of the unburned fuel elements.

CFD Simulations of Urban and Wildland Fire Spread Among Discrete Fuels Under Effect of Wind

It is important to investigate the urban and wildland fire behavior to mitigate the fire hazards. There have been many studies on such fires, but the need of real time fire simulations has recent increased and a demand to predict fire spread patterns in urban and wildland regions for decision-making strategies against fires has emerged. However, the knowledge of fire spread behavior is still insufficient, particularly for the condition of discrete fuel distributions. Under this condition the fire spread behavior shows high complexity due to the significant interactions between the radiation, conduction and convection heat transfer, especially under significant ambient wind effects. This paper investigates urban and wildland fire spread behavior by utilizing CFD simulations for two types of fuels under the effect of wind. A 15×15 square array, consisting of 225 fuel sources, is used to simulate the discrete fuel distribution, with varying fuel spacing and wind speed. The simulation method is similar to that used in our previous study, but with different ignition heaters. The comparison of the simulated results for the reduced and real scale models is reasonable, as verified by the similarity law. The critical fire spread distance, the wind effect upon fire spread, and the variation of fire spread rate for the two types of fuels are extensively investigated.Copyright