Long-term assessment of degradation processes is a very useful tool for an analysis of building materials performance. Since computational techniques are mostly used for this purpose, hygric properties of involved materials are required as substantial input data. Unfortunately, some construction details or heterogeneous materials have to be solved by means of multi-dimensional modelling which is demanding on computing power and thus the calculations may take a lot of time. The presented paper aims at determination of effective hygric properties of heterogeneous materials which would allow one-dimensional transformation. The parameter identification process is carried out on the basis of results of multi-dimensional modeling, using genetic algorithms. The main objective is to find such effective global moisture transport and accumulation functions that provide in one-dimensional modeling as similar results to multidimensional modeling as possible. The obtained functions give a very good agreement; the investigated relative humidity profiles differ only by 1.48 percentage points in average. The correctness of obtained results is also verified using the Lichtenecker's mixing rule as homogenization technique. The transformation of the original multidimensional problem into one-dimensional is found to substantially contribute to minimization of computational time, which is reduced from weeks to minutes.
Full Text (PDF)