Apparent Thermal Conductivity Measurements by an Unguarded Technique

Abstract

An unguarded longitudinal heat flow apparatus for measuring the apparent thermal conductivity (λ a) of insulations was tested with mean specimen temper atures from 300 to 330°K on samples up to 0.91 m wide, 1.52 m long, and 0.15 m thick. Heat flow is provided by a horizontal electrically heated Nichrome screen that is sandwiched between test samples that are bounded by tempera ture controlled copper plates and 9 cm of mineral fiber insulation. A deter minate error analysis shows λ a measurement uncertainty to be less than ±1.7% for insulating materials as thin as 3 cm. Three-dimensional thermal modeling indicates negligible error in λ a due to edge loss for insulations up to 7.62 cm thick when the temperature difference across the sample is measured at the screen center. System repeatability and reproducibility were determined to be ±0.2%. Differences of λ a results from the screen tester and results from the National Bureau of Standards were 0.1% for a 10-kg/m3 Calibration Transfer Standard and 0.9% for 127-kg/m3 fibrous glass board (SRM 1450b). Measurements on fiber glass and rock wool batt insulations showed the dependence of λ a on density, temperature, temperature difference, plate emittance, and heat flow direction. Results obtained for λ a as a function of density at 24°C differed by less than 2% from values obtained with a guarded hot plate. These results demonstrate that this simple technique has the accuracy and sensitivity needed for useful λ a measurements on thermal insulating materials.