This paper analyzes theoretical and experimental approaches for studying the thermal behavior of a living unit subjected to simulated climatic conditions.

The experimental setup is described first, then a study under steady-state conditions of temperature, which enabled the sensitivity of the model at different parameters to be defined. The value of the surface heat-transfer coefficient, spatial heterogeneity of temperatures caused by ventilation, and existence of thermal bridges were studied using finite differences and finite elements methods and by I.R. thermography.

Finally, under non-steady-state conditions, two computer models (finite differences and weighting factors), with different principles and application fields, were compared, and it was established that there was a good agreement between them.

The computer results compared with the experimental ones showed the weakness of all modeling related to certain hypotheses. A necessary modification of the air capacity, of the surface heat-transfer coefficient, or of the power enables a better agreement to be reached between the model and the short-term experiments.