Comparative study of radiometric and calorimetric methods for total hemispherical emissivity measurements
Résumé
Accurate knowledge of infrared emissivity is important in applications such as surface temperature measurements by infrared thermography or thermal balance for building walls. A comparison of total hemispherical emissivity measurement was performed by two laboratories: the Laboratoire National de Métrologie et d'Essais (LNE) and the Centre d'Études et de Recherche en Thermique, Environnement et Systèmes (CERTES). Both laboratories performed emissivity measurements on four samples, chosen to cover a large range of emissivity values and angular reflectance behaviors. The samples were polished aluminum (highly specular, low emissivity), bulk PVC (slightly specular, high emissivity), sandblasted aluminum (diffuse surface, medium emissivity), and aluminum paint (slightly specular surface, medium emissivity). Results obtained using five measurement techniques were compared. LNE used a calorimetric method for direct total hemispherical emissivity measurement [1], an absolute reflectometric measurement method [2], and a relative reflectometric measurement method. CERTES used two total hemispherical directional reflectometric measurement methods [3, 4]. For indirect techniques by reflectance measurements, the total hemispherical emissivity values were calculated from directional hemispherical reflectance measurement results using spectral integration when required and directional to hemispherical extrapolation. Results were compared, taking into account measurement uncertainties; an added uncertainty was introduced to account for heterogeneity over the surfaces of the samples and between samples. All techniques gave large relative uncertainties for a low emissive and very specular material (polished aluminum), and results were quite scattered. All the indirect techniques by reflectance measurement gave results within ±0.01 for a high emissivity material. A commercial aluminum paint appears to be a good candidate for producing samples with medium level of emissivity (about 0.4) and with good uniformity of emissivity values (within ±0.015).