Gunnar Heskestad

Engineering relations for fire plumes

This paper presents a number of engineering relations drawn from the literature for calculating properties of fire plumes. Plume properties considered include flame heights, temperatures, velocities, concentrations of combustion products, and entrainment rates of air from the surroundings. In addition, a brief discussion is presented on the effect of fire growth to demonstrate the validity of the relations set forth. A note on virtual origin is also included.

Luminous heights of turbulent diffusion flames

A general analytical relation for predicting mean luminous heights of buoyancy-controlled, turbulent diffusion flames is established. The relationship is based on a plot of experimental flame heights in correlation coordinates proposed previously, including extensive data recently published.

Virtual origins of fire plumes

A model is proposed for correlating measurements of virtual origins of fire plumes, made possible in conjunction with a relation established recently for predicting mean flame heights. The model is consistent with plume theory and is shown to correlate available determinations of virtual origins quite well. However, precise experiments are needed to fully test the model and to investigate anticipated effects of fuel identity. It is shown that the results of this work lead to a rationale for the success of previously proposed temperature correlations which appear to be valid even into the flaming regions of plumes.

The initial convective flow in fire

This study concerns physical modeling of the initial fire environment generated by fire in an enclosure, which persists up to that time in a fire when recirculation of combustion products begins to influence the further yield of products. It was the primary purpose to investigate experimentally the validity of modeling relations, proposed previously, for the convective flow generated by “power-law” fires, i.e., fires growing in heat-release rate with a specific power of time from ignition. Three wood-crib fires of different fire-growth rates were combined with three different ceiling heights under large flat ceilings for a total of nine experimental configurations. The experimental fires were power-law fires growing with the second power of time. Temperatures and velocities were measured in the hottest gas layer under the ceiling. The data were shown to be well correlated in nondimensional variables of the modeling theory. It was possible to establish analytical expressions for the nondimensional temperature and velocity fields. A useful finding, which appears to be valid also for other kinds of fire growth than that investigated (including steady fires), is that the local gas velocity in the hottest layer can be related directly to the local temperature rise and ceiling clearance, regardless of fire-growth rate and time from ignition.

Quantification of thermal responsiveness of automatic sprinklers including conduction effects

The response time index (RTI) represents the product of the thermal time constant for the heat-responsive element of an automatic sprinkler and the square root of the associated gas velocity. The RTI and sprinkler temperature rating are usually sufficient to predict sprinkler response, provided gas temperatures and velocities generated by the fire at the sprinkler site are known. However, recent evidence has indicated that a companion response parameter may be needed to quantify response for low gas temperatures and velocities and for low-RTI sprinklers in growing fire situations. The companion parameter accounts for heat loss by conduction to the sprinkler mount. The technical basis and methods of measurements of the response parameters are presented, along with preliminary results of room fire tests conducted to verify the refined response model.

Extinction of gas and liquid pool fires with water sprays

Extinction in open space of flames from pool fires by downwardly directed water sprays has been investigated on two linear scales, one three times larger than the other. Circular pool fires were employed as fire sources, mostly in the form of gas discharge (methane) from a horizontal sand surface but also, to a limited extent, in the form of heptane pools. The results are presented in normalized plots based on scaling theory verified in a previous study. Extinction data from the methane fires are insensitive to the initial spray angle of the nozzle discharge. The data are consistent with an engineering relation showing extinction water flow rate approximately proportional to an effective nozzle diameter, and to the 0.4-power of both nozzle height and freeburn heat release rate. This result has been interpreted to indicate that spray-induced dilution of the flammable gas is a major factor in extinguishing fires from gaseous discharge. Extinction data of liquid pool fires from this study (n-heptane) and previous investigations (gasoline, JP-5) are consistent with the methane data, except for somewhat higher water rates at extinction.

Ceiling jets of strong fire plumes

Abstract Measurements are presented of maximum excess temperatures in ceiling jets produced by strong, turbulent, steady fire plumes impinging on flat horizontal ceilings. The ratio of free flame height (flame height in the absence of a ceiling) to ceiling height (flame-height ratio) ranged from approximately 0·3 to 3 . A common type of correlation of ceiling jet excess temperatures, employing the ceiling height as a normalizing length scale for the observation radius, did not correlate the data well. However, an excellent correlation of the temperature data was achieved for flame-height ratios up to approximately 2, using an expression for the plume radius at the ceiling level as the normalizing length scale. Data at the largest flame-height ratio investigated, 3, did not conform with the other data, attributed to significant heat release in the ceiling jet itself.

Scaling the interaction of water sprays and flames

Scaling relations for (dry) compartment fires have been extended to include interactions of water sprays with the gas phase, including flames. All important interactions scale properly, except attenuation of thermal radiation by the sprays. Methods of scaling spray devices to meet requirements are discussed, including application of geometrically similar spray nozzles and, where geometric similarity cannot be assured, the application of an effective spray nozzle diameter. Extinction experiments were conducted in a large space on linear scales differing by a factor of ten, where freeburn heat release rates varied from 1.1 to 1530 kW. The results are well correlated in normalized coordinates based on the scaling theory. There is no evidence, in the extinction scenario investigated, of the improper scaling of thermal radiation attenuation, or a slight inaccuracy in the scaling of mean drop diameter to requirements.

Fire Plumes, Flame Height, and Air Entrainment

Practically all fires go through an important, initial stage in which a coherent, buoyant gas stream rises above a localized volume undergoing combustion into surrounding space of essentially uncontaminated air. This stage begins at ignition, continues through a possible smoldering interval, into a flaming interval, and may be said to end prior to flashover. The buoyant gas stream is generally turbulent, except when the fire source is very small. The buoyant flow, including any flames, is referred to as a fire plume.

Update: The initial convective flow in fire

This brief note updates correlations established previously by the authors for the ceiling flow generated by fires growing with the second power of time, based on knowledge of the actual heat of combustion of wood, the combustible used in the original experiments leading to the correlations. In addition, the correlations are generalized to include combustibles with a significantly different convective fraction of total heat release rate than wood.