Precision of instruments that measure heat transfer through windows

The purpose of this study was to determine the precision of instruments that measure thermal transmission through a window by determining: (1) the precision of each instrument under specified test conditions, and (2) the differential behavior of each instrument under different temperature levels.

Seven conventional window treatments and three temperature levels were randomly selected and evaluated using the Window Thermal Transmission Instrument (WITI). A 7 x 3 fully crossed, balanced, factorial experimental design with ten replications was used. Measurements of voltage and temperature differential were taken and used to calculate thermal transmission coefficients (U, W/m²K). In addition, data was collected from three published studies (three different instruments) and analyzed in terms of research design and precision of instrument.

Analysis of variance was used to determine the precision of the WTTI; how well one was able to reproduce thermal transmission, and how the magnitude of variation changed as the specified test conditions were altered. Precision was found to be better at the higher temperature level than at either of the other two temperature levels. As temperature level increased, the variation in thermal transmission coefficients decreased. Temperature level was associated with 57.31% of the variation in the thermal transmission variation, while window treatments were associated with 63.15% of the variation in thermal transmission variation. There was no difference in the precision of the instrument when six of the conventional window treatments were tested. There was a difference when the foam backed drapery was tested. This treatment has produced a larger variation in the thermal transmission coefficients than all other window treatments evaluated.

Statistical quality control charts were used to determine the precision of the Thermally Insulated Chamber, Twin Room Facility, and Window-In-Wall Instrument because the experimental design used with each instrument prohibited the application of analysis of variance. The charts indicated no significant differences in precision of the instruments. A comparison of the control charts for these three instruments and the WTTI showed no significant differences, however, the WTTI had the smallest average standard deviation.