Vacuum insulation panels (VIPs) are composed of an evacuated nanoporous core material, and a barrier envelope to atmospheric gases. Their apparent thermal conductivity after manufacturing is approximately 5 mW/(m.K), five to eight times lower than that of conventional insulation materials. Compared to conventional insulation materials, the most important issue remains in the duality performance / durability. The answer lies in the understanding and modeling of heat and mass transfer in VIPs. Many experimental and numerical studies about heat transfer led to a semi-empirical model for the apparent thermal conductivity of a VIP, taking into account the different transfer modes in the panel. This model highlights the main mechanism of VIPs aging: due to gas permeation through the envelope, pressure and humidity in the panel increase gradually over time, which causes an increase of the apparent thermal conductivity. The study of mass transfer through the gas barrier envelope is particularly difficult for three main reasons. First, the permeation modeling of VIPs envelopes (multilayer membranes with metalized polymer films) has to take into account micro-defects in the metallic layers, which play a key role in the total permeation rate. Second, the permeances to be measured are too low for many conventional methods, especially for dry air. Third, a literature analysis shows that the classical model for mass transfer through barrier envelopes does not take into account the combined influence of temperature and relative humidity, which is a great lacuna. From experimental data available in the literature, the role of relative humidity on the barrier properties of the VIPs envelopes is highlighted. The existence of a coupling phenomenon between the mass flows of the various gases is hypothesized to start the establishment of a new gas permeation model, which takes into account partial pressure and total pressure, and thus the molar concentration of each gas in the mixture. The predictions of this model are compared with predictions based on the classical model for gas permeation, and the differences between the two models are analyzed. Two experimental campaign are then implemented to measure envelope permeance, through whole VIPs aging in climatic boxes and through direct measurement of the permeance on membrane samples (manometric method). These measurements are carried out at fixed temperature and relative humidity (T = 48 °C, φ = 65 % HR), but with a total pressure ranging from 80 mbar to 1 bar. This exploratory measurement campaign shows no significant influence of the total pressure on the apparent permeability to water vapor. These results are used to draw first conclusions on the respective roles of the partial pressure and the total pressure, and suggest some outlooks to the experimental approach initiated in this study.
