No abstract

SUMMARYTo accurately predict the structural performance of a wood member, knowledge is required of the rate at which it chars and the temperature distribution in the residual load-bearing section. The charring rate and temperature distribution can be calculated with a model that predicts the thermal degradation or pyrolysis of wood exposed to a high-temperature environment. More than 50 wood pyrolysis models have been developed since World War II. They range from simple analytical expressions to complex systems of coupled partial differential equations that describe the heat and mass transfer through wood and char.This paper presents a brief overview of the aforementioned models and provides a more detailed description of a new model. This model is referred to by the acronym CROW (Charring Rate Of Wood). Although the intent was to keep CROW as simple as possible, the model accounts for the four major factors that affect the thermal degradation of wood: dry density of the wood; moisture content of the wood; lignin content of the wood; char contraction.The predictive capability of CROW was evaluated on the basis of ASTM E 119 furnace data obtained for a Douglas fir glulam beam tested under different loads. CROW predictions, with some adjustment for moisture effects, are in reasonable agreement with the measurements. The model will be most useful to predict performance of wood members exposed under thermal conditions that deviate from the standard fire (natural or parametric fires) and/or members that are protected by a membrane.

The physical phenomenon of piloted ignition of a material is described. A number of mathematical models of this phenomenon are presented in order of decreasing complexity. The most sophisticated models include gas-phase phenomena. Simple models neglect all chemical effects and are purely thermal. The most commonly used criteria for piloted ignition are discussed. Correlations used in piloted ignition studies from the past 40-50 years are summarized. Many investigators have been successful in correlating piloted ignition data of wood using a simplified thermal model in combination with a critical surface temperature criterion. Emphasis of this review is therefore on this approach. The paper concludes with a detailed analysis of various factors affecting piloted ignition of wood. Some of the factors are related to the experimental conditions, others are associated with the characteristics of the material.

A modified cone calorimeter with an enclosure has been developed for measuring the yield of combustion products including CO and smoke under vitiated conditions. The CO yields of methane, propane, PMMA, ABS, polyethylene, and Douglas fir are found to increase by at least a factor of two as the oxygen concentration is decreased from 21% to 14%, while the smoke yields are found to be insensitive to vitiation for the solid materials (less than 30% change).Results for air vitiated separately by nitrogen and by carbon dioxide suggest that the CO yield for a given fuel in a free burn is mainly controlled by the flame temperature. For ambient conditions, the CO yields for the solid samples are about a factor of 2.3 smaller than the smoke yields for all the solid materials studied.

Oxygen consumption calorimetry has been used for more than twenty years to determine the heat release rate from experimental fires. Several investigators have quantified the theoretical uncertainty of heat release rate measurements. Estimates of the precision of heat release tests based on round robin results suggest that in practice the uncertainty may be much greater, in particular for intermediate and large-scale tests. The objectives of this paper are to examine the reasons for this discrepancy, and to propose actions that will improve the precision of heat release rate measurements to an acceptable level without raising costs beyond the reach of commercial fire testing laboratories. The proposed actions include the establishment of a proficiency program to obtain a realistic measure of the uncertainty of heat release rate measurements in fire tests.