Matti Kokkala

The cone calorimeter used for predictions of the full-scale burning behaviour of upholstered furniture

In the EC-sponsored CBUF (Combustion Behaviour of Upholstered Furniture) study three different models were developed for using Cone Calorimeter results to predict full-scale furniture burning. Model I is a correlation-based approach and has the most comprehensive coverage of styles and furniture types. Model II is based on area-convolution. It has been developed, so far, for several of the common upholstered furniture styles. Model III is based on an extension of a thermal flame spread model and is here developed only for mattresses. Models II and III offer the prediction of the burning items time history, while Model I computes the peak HRR, the time to peak, and several other primary characteristics. The predictability of all three models is good. At present, the models presented do not consider furniture with plastic frames, and only a limited predictability is offered for office-type furnitures items which contain a hard-plastic shell. The models offer a very significant improvement over the existing state of the art.

Experimental Study Of Boilover In Crude Oil Fires

An experimental study of boilover phenomena in connection with crude oil and kerosene pool fires was carried out in the large scale test facilities of the Fire Research Institute of Japan. Arabian light crude oil was burned in steel pans ranging from a circular one with a diameter of 0.3 m to a square one with a 2.7 m side. An oil layer of 10 mm to 100 mm was floating on water in the pans. The intensity of boilover was found to increase with increasing fuel thickness. An almost tenfold increase in the amount of unburned residue was found to occur as the layer thickness was increased from 40 mm to 50 mm, Resul ts of measurements of liquid and gas temperatures, burning rates and radiative heat outputs are reported.

Upward Flame Spread On Wooden Surface Products: Experiments And Numerical Modelling

A series of upward flame spread tests on different wood products was carried out on 2.4 m and 7.2 m high wall assemblies. The ignition source was a 1.2 m wide and 0.1 m deep propane diffusion burner applied usually at a rate of 100 kW. The measurements included gas and surface temperatures, heat fluxes to the specimen, rate of heat release, etc. After an initial growing fire, a period of decay was observed until the lower parts of the wall started to burn through. If the rear side of the board was insulated, a second phase of increased flame spread was observed, whereas in the case of a conducting substrate the intensity of the fire remained low. A thermal flame spread model was successfully applied to simulate the rate of heat release as a function of time.

Measurements of the characteristic lengths of smoke detectors

The dynamic performance of point-type smoke detectors is described by a simple model including two independent parameters, the static response threshold, and the characteristic length of the detector. An experimental system with constant rate of increase of smoke density under varying velocity conditions is used to determine the model parameters.

Determination of dynamic model parameters of smoke detectors

A simple model for the dynamic response of point-type smoke detectors is described. The model is based on two independent parameters: the characteristic length and the static response threshold. Experimentally determined values of the parameters for six commonly used point-type smoke detectors are reported. In the tests the free stream flow velocity was varied in the range of 0.2-0.8 m/s and the rate of rise of smoke density in the range of 0.1-2 dB/ms. The model parameters were determined from the experimental data by using the least squares fitting with regularisation. The characteristic length was found to vary in the range of 2-10 m indicating a smoke entry time constant in the range of 10-50 s at a flow rate of 0.2 m/s typical of smouldering fires. The response of photoelectric detectors was described by the model reasonably well. In the case of ionisation detectors, significant deviations were found. A method of using the model in fire safety engineering calculations is also presented.

Time-Dependent Event-tree Method For Fire Risk Analysis: Tentative Results

A fire risk analysis method based on the use of time-dependent event trees is presented. To concretize the method, a simple example case, a property-loss risk analysis for a fire in a one-storey industrial hall, is presented. The method incorporates explicitly the time dependence of the fire and its consequences. This is achieved so that the fire incident is divided to time intervals and the events within each time interval are analyzed with a separate event tree. The event trees form the basic structure of the analysis and it is shown here how to connect them to a description of the evolution of the fire. This description can be expressed as a stochastic Markov process and it is shown how to link the branching probabilities of the event trees to the transfer matrices of the Markov chain.