Hens H, Janssens A
Bibliographic info:
USA, Atlanta, American Society of Heating, Refrigerating and Air Conditioning Engineers (ASHRAE), 1998, proceedings of "Thermal performance of the exterior envelopes of buildings VII" a conference held Sheraton Sand Key Hotel, Clearwater Beach, Florida

In cool and cold climates, sloped roofs with cathedral ceilings are quite sensitive to moisture damages caused by built-in moisture and prolonged concealed condensation of water vapor produced inside. Conventional solutions are to leave a cavity between the thermal insulation and the sheathing and vent it with outside air and/or to include a vapor barrier below the insulation layer. An alternative, however, is the self-drying roof. This concept was evaluated experimentally. For that purpose, three well insulated roof sections, all covered with shingles and lined inside with a gypsum board, were tested in a hot box. The first had an airflow and vapor-tight polyethylene film between the glass fiber insulation and the gypsum board internal lining. The second had the gypsum board only as an airflow retarder, and the third had a new type of organic, glass-fiber, fabric reinforced felt as airflow and vapor retarder. The plywood deck under the shingles contained a known amount of built-in moisture. The sections were exposed to a sunny period first, followed by a steady-state cold period afterwards. Section 1 remained wet, with the moisture moving from the plywood to the polyethylene during the sunny period and back to the plywood during the cold period. Section 2 dried during the sunny period but turned wet again during the cold period. Section 3 finally dried during the sunny period and got some wetness during the cold period; however, it got less than roof 2. Apparently, roof 3 came closest to the concept of self drying. In order to evaluate to what extent simple engineering tools and simplified models could predict the measured response, the tests were also simulated using three such models.