Describes an attempt to calibrate the Isothermal Planes method for R-value predicting, using previously published results. A set of 84 cases have been collected, for which measured R-values have been published in major publications, together with sufficient detail to allow calculation. This collection includes wood-framed, wood-plank, metal-framed and masonry structures. Comparison of R-values for this set shows good agreement between measured values and those calculated by the Isothermal Planes method, provided that a slightly tighter set of definitions than normal is followed. These include allowance for the small contact resistances between frame and facing (this is significant only with metal frames), as well as better rules about how layers should be defined. The R-value forecasts over the whole set were within +/- 0.1 m2 degC/W of measured values for 81% of the cases, and percentage differences were larger at smaller R-values. The average was within 2% of the measured values, indicating that there was little or no consistent bias. The standard deviations were 10% to 17% for all three material groups. Forecasts of appropriate quality for engineering purposes appear to be obtainable with the minor improvements suggested to the Isothermal Planes method, but the Parallel Flow method is not adequate and cannot be made so. Assurance of correct procedures in practice remains a problem. Continued measurements must still be used as a final arbiter.

KEYWORDS: year 1995, validating, thermal resistance, calibrating, isothermal, measuring, calculating, comparing, accuracy, buildings, masonry, data, thermal bridges, timber framed construction, steel