Abstract

Modern timber construction techniques rely on metallic connectors. Their failure during fires is related to shear failure, which can happen at lower temperatures than those leading to pyrolysis. Based on this information, a series of tests were carried out exposing wood specimens to various heat fluxes in a Cone Calorimeter on a vertical configuration with no pilot ignition and a normal air atmosphere. The aim of this investigation was to develop a dependable temperature measurement methodology in wood samples exposed to high heat fluxes and to build a data-base in order to assess different wood pyrolysis models. The experiments yielded highly repeatable results that are hard to achieve with wood, although at lower heat fluxes the repeatability was poorer. The data were then normalized using a simple, inert, heat transfer model. It will be used as a tool to assess the importance of different terms that are incorporated in models that predict in-depth temperature profiles for wood. The samples behaved almost like an inert solid when subjected to low heat fluxes. Poorer repeat ability at the lower levels of heat flux is justified by the beginning of pyrolytic reactions in wood at about 10 kW m(-2). (C) 2007 Elsevier Inc. All rights reserved.