Flammability assessment of fire-retarded Nordic Spruce wood using thermogravimetric analyses and cone calorimetry

Experimental techniques such as the cone calorimeter, representing realistic fire conditions, and the thermogravimetric analyser (TGA) combined with evolved gas analysis (EGA) can be used to determine flammability and degradation properties of materials. The desire is to correlate the flammability properties measured in the cone calorimeter for samples of size 100 mm x 100 mm with those measured or deduced from TGA combined with EGA for milligram samples. Such an achievement will allow the design of fire-safe materials by quickly assessing (a) the fire safety of materials in their earliest milligram formulation and (b) the dependence of their flammability properties on the molecular structure of the material. In the present study, a cone calorimeter and TGA investigation is conducted for commercial Nordic Spruce wood impregnated by mono-ammonium phosphate (fire retardant, FR) through a vacuum pressure method. Experiments in the cone calorimeter with increasing FIR concentrations indicated that (a) the char yields increased, (b) the apparent ignition temperature increased, (c) time to piloted ignition increased, (d) the total amount of heat released was reduced, (e) the peak heat release rate was reduced and (f) the carbon monoxide and smoke yields increased especially before ignition occurred. By comparison, char yields also increased with FR content in the TGA degradation experiments in nitrogen. The increase in the char yield with FIR content explains quantitatively the decrease in the heat release in the cone calorimeter. By contrast, the onset temperatures measured in TGA decreased, whereas the ignition temperature deduced in the cone calorimeter increased with FIR content. This difference is attributed to reduced yield of levoglucosan (reported in recent literature using TGA/EGA) with increased char yield as well as to the presence of phosphorous containing moieties in the volatiles, which both can quench piloted ignition. Finally, the TGA measurements showed that the FIR concentrations decreased for milligram samples at different distances from the surface of the wood used in the cone calorimeter measurements. The variation of FIR retardant with depth needs to be considered when using TGA data to interpret cone calorimeter measurements and the fire performance of the FR wood in approval tests such as the single burning item (SBI). (C) 2009 Elsevier Ltd. All rights reserved.