Transient calculation models that predict moisture content in wood and other hygroscopic construction members employ the sorption curve of the materials to express the relationship between relative humidity of air and equilibrium moisture content. However, the sorption curves are somewhat temperature dependent, and they exhibit hysteresis, so the equilibrium state depends on the sorption history. Most transient moisture transport models neglect these facts. This paper will attempt to refine the current models by implementing a sorption “surface” spanned by relative humidity, temperature, and the resulting moisture content.

The paper mainly indicates that taking hysteresis into account will tend to reduce the effective moisture capacity of building materials for short-term moisture variations, while taking the influence of temperature on the sorption properties into account will tend to increase the apparent moisture capacity in practical situations where heating of a building element is normally ensued by its drying.

The paper will investigate the effect of considering moisture content dependency of the vapor permeability, sorption hysteresis, and temperature dependency of the sorption curves in an existing calculation model, which is used to predict the moisture conditions in solid wood constructions. This paper presents measurements of moisture content in almost 5-in.-thick roof elements of solid wood, which are located over an open stable for cattle. A previous comparison between measurement and prediction of the moisture content of the wood had indicated the possibility of using temperature-dependent sorption data as a way to improve the correspondence—particularly the question about the magnitude of the seasonal variations. The calculation results with the new model are presented alongside with experimental results and practical experiences of such constructions.