One important question concerns the significance of hygrothermic characteristic values. Whilst the strength and stiffness values for safety-related building materials are determined over the course of large series of tests and statistically evaluated, most hygrothermic characteristic values are determined through random sampling on a few randomly selected samples.
Due to the wide diversification of the material parameters and the differing installation situations, in which different respective material parameters can be critical, it cannot be ensured in each case that the generic datasets achieve less favorable results than the measured data sets for a specific product. In the case of application in a sloped, ventilated roof, the moisture in the materials results from, for example, the high humidity level of the outdoor climate in winter. The water content increases when the level of the sorption isotherm is raised. In flat roofs, however, the moisture comes from the interior area via the vapor barrier. The quantity barely changes in relation to the properties of the materials above the vapor barrier. The previously more critical high level of the sorption isotherm now results in the same quantity of moisture leading to a smaller increase in the relative humidity in the outer region of the component and the generic materials remaining slightly more favorable. A generation of material datasets, which are critical in all cases, is therefore not possible. Furthermore, the respective generic datasets could only be generated and verified on the basis of one to three measured datasets.
The generic datasets can be utilized for the dimensioning of constructions and should produce representative results in the majority of cases. Particularly in the case of hygrothermically-favorable results, slight differences between the generic and individual specific materials can be expected at most. In contrast, if results are obtained which are close to critical limits, moisture levels will be reached in the materials in which the differences between the generic and individual measured data can be greater. If, in individual cases, it is not possible to improve the component in an appropriate manner, the use of specific material characteristic values for the real product should be attempted in order to enable a more reliable evaluation.