The Relevance of Surface Resistances on the Conductive Thermal Resistance of Lightweight Steel-Framed Walls: A Numerical Simulation Study

Abstract

The accurate evaluation of the thermal performance of building envelope components (e.g., facade walls) is crucial for the reliable evaluation of their energy efficiency. There are several methods available to quantify their thermal resistance, such as analytical formulations (e.g., ISO 6946 simplified calculation method), numerical simulations (e.g., using finite element method), experimental measurements under lab-controlled conditions or in situ. Regarding measurements, when using the heat flow meter (HFM) method, very often, the measured value is based on surface conditions (e.g., temperature and heat flux), achieving in this way the so-called surface-to-surface or conductive thermal resistance (Rcond). When the building components are made of homogeneous layers, their Rcond values are constant, regardless of their internal and external surface boundary conditions. However, whenever this element is composed of inhomogeneous layers, such as in lightweight steel-framed (LSF) walls, their Rcond values are no longer constant, depending on their thermal surface resistance. In the literature, such systematic research into how these Rcond values vary is not available. In this study, the values of four LSF walls were computed, with different levels of thermal conductivity inhomogeneity, making use of four finite elements’ numerical simulation tools. Six external thermal surface resistances (Rse) were modelled, ranging from 0.00 up to 0.20 m2·K/W. The average temperature of the partition LSF walls is 15 °C, while for the facade LSF walls it is 10 °C. It was found that the accuracy values of all evaluated numerical software are very high and similar, the Rcond values being nearly constant for walls with homogeneous layers, as expected. However, the variation in the Rcond value depends on the level of inhomogeneity in the LSF wall layers, increasing up to 8%, i.e., +0.123 m2·K/W, for the evaluated Rse values.