No abstract

Lightweight architecture, implementing coated fabric and membrane materials, increasingly adopts double or multiple layer pneumatic cushions. The reason lies in the possibility to guarantee adequate thermal insulation and better indoor comfort, especially in permanent buildings. The cushion thermal transmittance decreases as the number of layers increases. However, the U-value is generally assessed by simple calculations assuming the cushion layers are parallel planes. This way, the center-of-the cushion U-value is taken as the overall U-value, disregarding any effect of the curved geometry of the cushion on the heat transfer across it.In this paper an experimental approach to the evaluation of the overall U-value of multiple layer cushions is proposed. Two sample cushions made up of PVC coated polyester with a surface of about 1 m 2 are built, one double layer and one triple layer.They are tested in vertical position in a double thermal chamber laboratory apparatus, establishing a constant temperature difference equal to 25°C across them. The cushion surfaces exposed to the chambers are divided into thermally homogeneous portions where temperature and heat flow density probes are centrally placed. This way the overall thermal resistance and thermal transmittance of the cushions are measured. In parallel, the U-value for the two samples are calculated using literature correlations for free convection in rectangular cavities and assuming radiative heat transfer between parallel grey surfaces. The experimental results show that, passing from two to three layers, the overall thermal resistance of the cushion almost doubles (+ 91%) and the thermal transmittance reduces by about 30%. At the same time, simple calculations of the U-value are found to underestimate the insulation capacity of the cushions, especially for the double layer one. These outcomes point out the necessity of further investigations to understand the impact of the cushion shape on free convection in the air cavity.