| Résumé | Hygrothermal simulations are a common method for assessing the moisture performance of walls. With climate change, it is crucial to evaluate how walls will respond to future conditions. However, future climatic data is subject to uncertainties, and evaluating wall responses with these uncertainties require a large number of simulations. To address this challenge, a meta-model was developed based on Partial Least Squares (PLS) regression as an alternative to hygrothermal simulations. A PLS model was firstly developed for a brick cladding wall and this study demonstrates the applicability of the model to cover other wall configurations and climatic uncertainties due to the initial conditions assumed in the climate model. A modified brick model was proposed for stucco cladding walls and its performance was verified with the simulation results for three Canadian cities under two climate periods: historical (1986–2016) and future (2062–2092), considering the worst-case scenario when global temperature is projected to increase by 3.5 °C at the end of the century based on the RCP 8.5 emission scenario. The results showed that the PLS model results in a similar performance as simulations for stucco cladding wall and different climate runs. The predicted results from the meta-model can be used as a screening measure to limit the number of simulations to cases where the predicted hygrothermal performance is above a certain threshold set by the user. The general approach will be applied to additional wall claddings such as wood-siding, other emission scenarios, and additional geographical locations representing different weather characteristics. |
|---|
