Wind Uplift Capacity of Foam-Retrofitted Roof Sheathing Panels Subjected to Rainwater Intrusion

Abstract

Widespread damage to residential roofing systems during Hurricane Andrew in 1992 highlighted the structural inadequacy of wood roof sheathing attachment in existing Florida homes. Prior research found that retrofitting wood decks with closed-cell spray-applied polyurethane foam (ccSPF) could provide adequate wind resistance to pre-1994 roofs. However, it is unknown whether this retrofit method increases the potential for moisture to accumulate in the roof sheathing, and if so, what effect this has on the strength of the bond between the ccSPF and the wood substrate. Five full-scale roofs were subjected to natural weathering for 3 months, and 131 small-scale laboratory samples of ccSPF bonded to wood sheathing were intermittently sprayed with water and stored in a high-humidity environment for up to 16 weeks. These tests enabled systematic assessment of the effects of water leakage on ccSPF-retrofitted roof structures. The attic roofs were retrofitted with two ccSPF configurations and covered with three-tab asphalt-shingle roofing subjected to extreme leakage conditions. A control attic roof was not retrofitted with ccSPF. Moisture content and temperature were continuously monitored in the wood truss members of the roofs over 3 months of exposure to both natural and simulated rainfall conditions. At the conclusion of the exposure period, sheathing panels were harvested and subjected to wind uplift structural tests. Tensile tests were conducted on the small-scale samples after 0, 1, 2, 4, 8, 12, and 16 weeks of exposure. The results demonstrated that with extreme leakage conditions, ccSPF-retrofitted roofs experienced significant moisture buildup in the wood substrates. However, the ccSPF-retrofitted sheathing panels maintained high wind uplift failure capacities. The tensile tests on small-scale samples demonstrated as much as 54% reduction in tensile strength of the wood-ccSPF bond with time as a result of moisture-induced bond degradation, although the strength of this bond after 16-week exposure remained above design wind pressures by a minimum factor of 3.75.