FIRE AND MATERIALS, VOL. zyxwvutsr 20, 173-181 (1996) zyxwvuts Thermogravimetric Studies of the Pyrolytic Behaviour in Air of Selected Tropical Timbers Musa Momoh A. N. Eboatu Nnamdi Azikiwe University, Awka, Nigeria E. G. Kolawole Ahmadu Bello University, Samaru-Zaria, Nigeria A. R. Horrocks Faculty of Science, Usmanu Danfodiyo University, PMB 2346, Sokoto, Nigeria Bolton Institute, Deane Road, Bolton, Lancashire BL3 5AB, UK A number of common tropical timbers have been subjected to thermal analytical investigation as part of a wider research programme under air atmospheric conditions. Thermal parameters obtained were correlated with ovendry densities of the timbers. Two well-defined pyrolysis stages have been observed which occur over the temperature ranges 201-426°C and 397457°C. The kinetics of the thermal degradation of the timbers were obtained using Broido's analytical procedure. These results were interpreted on the basis of the known mechanism of pyrolysis and the calculated kinetic parameters were discussed within the context of other published values for cellulose. INTRODUCTION Wood, a major material for use in buildings and con- struction, is frequently implicated in fire in~idence.'-~ Therefore many studies have been conducted on wood combustion and a considerable number of papers have been p~blished.~-~ Previous related investigations of wood pyrolysis include isothermal s t ~ d i e s ~ - ~ and dy- namic thermogravimetric a n a l y ~ i s ? - ~ . ~ * ~ Most of these studies tend to centre on temperate and subtropical tim- bers which find use in the construction industries. How- ever, within Africa, hardwood and tropical woods have commercial potential although their combustion behav- iours are little understood. We report here on the thermoanalytical behaviour of ten tropical timbers when heated in air from ambient to 600°C. The work forms a part of a larger study of the burning behaviour of hardwoods to be reported else- where. '' EXPERIMENTAL Materials The timbers were procured from a Nigerian timber fac- tory and their names and types are shown in Table 1 along with their densities." Thermogravimetry (TG) experiments SmalI shavings of each wood (6-10 mg) were placed in the furnace of a Mettler TA 3000fTG 50 under static air atmosphere and heated at lOKmin-'. By means of a computer incorporated into the analyser, onset, peak and end temperatures of pyrolysis/combustion were ob- tained. In addition, stepwise decomposition profiles of each of the timbers were determined. Calculation of kinetic parameters Many analyses of the kinetics of pyrolysis from TG results have been based on the assumption that an over- all first-order reaction is valid. The first-order kinetic relation which is commonly used is given by" where zyxwv m is the sample mass fraction at any time zy t and temperature T(K) and R is the gas constant. Stamm" and Akita4 have used this method to determine the Arrhenius kinetic parameters, namely the frequency fac- tor zyxwvu Z and the activation energy, E,. The techniques of Freeman and Carroll' and Anderson and for measuring the kinetics of pyrolysis are based on determining the order of reaction from the weight loss of a sample exposed to an increasing temperature environ- ment. In this work, the Broido method' is preferred because there is no need to make an assumption about the order of reaction and the equation E, RZ ln[ln(l/m)] = - -In - . T,' RT EaB is used to calculate E,. B is 'the heating rate (e.g. lOKmin-') and T, is the DTG peak temperature for a given decomposition stage. After the activation energy CCC 0308-0501/96/040173-09 zyxwvutsr 0 1996 by John Wiley & Sons, Ltd Received October I994 Accepted (revised) 7 March I996