Radiation emission from a heating coil or a halogen lamp on a semitransparent sample

The radiation emission of the heating coil of a Cone Calorimeter and the one of the halogen lamp of a Fire Propagation Apparatus have been studied experimentally for varying power settings. These are two standard apparatuses used for fire calorimetry. The objective is to characterize and compare the radiative flux spectrum received by a fuel sample during pyrolysis experiments. The deviation from the standard assumption of black or gray emission is discussed. It is observed that the emission of the heating coil can be approximated well to an ideal blackbody, especially in the infrared range. On the contrary, the halogen lamp emission is more complex, non gray, with an important contribution in the visible and in the near infrared ranges. The flux received by a sample exposed to these emitters is predicted using ray tracing simulations. This shows that the irradiation flux and spectrum from the cone can be accurately calculated if the coil temperature is known. The non Lambertian irradiation flux from the lamp is modeled with a combination of diffuse and collimated intensities, representing the direct emission from the lamp itself and the reflection by the mirror at the rear side. For both emitters, the irradiation is confirmed to be approximately uniform over the surface of a sample 5 cm large (maximum deviation of 2% on the incident flux). The uniformity decreases for larger samples, but the ratio of the flux at the center to average flux is still 1.04 for standard 10 cm 10 cm samples under the cone. For illustration purposes, the influence of the spectral characteristics of the emitter is studied in the case of a sample of PMMA, a non gray translucent medium. Using recently published measurements of PMMA absorptivity, the absorbed flux by a 3 cm thick sample is predicted. In the case of an incident flux of 20 kW/m2, the calculated average absorptivity of the sample is 0.91 under the cone, while it is 0.32 under the FPA lamp. These calculations involve absorption data of a virgin sample at room temperature and consequently the numerical results only hold for the initial instants of irradiation. However, the very large differences in radiative behavior show that important discrepancies in the pyrolysis behavior are expected between the two emitters. This might have consequences for fire testing and inter comparisons of flammability results worth further investigation. 2013 Elsevier Masson SAS. All rights reserved.