Recent research conducted at the University of Florida has demonstrated that closed-cell spray foam (ccSPF) applied to the underside of roof decking effectively bonds the sheathing to the framing, significantly increasing uplift resistance. The study, conducted by Dr. David O. Prevatt and funded by Honeywell and Huntsman, two makers of ingredients that go into ccSPF, found that 3 inches of the foam sprayed between framing members provided a threefold increase in uplift resistance as compared with traditionally installed roof sheathing panels. While these results sound impressive, Dr. Prevatt points out that the increase provides the same benefits as increasing the nailing schedule to a 6/6 schedule (every 6 inches along panel edges and every 6 inches in the field) from the usual 6/12 schedule. What was perhaps most impressive is that using only spray foam to glue the sheathing to the framing provided almost as much resistance (178 to 209 psf) to uplift as does 8d common nails (205 psf) installed at the 6/6 schedule. This suggests what may be the biggest structural advantage of a foamed roof assembly — reducing the likelihood of a roof blowoff when the sheathing doesn't get nailed off with enough nails or when too many nails miss their mark.

A test panel (left) in a study at the University of Florida simulates a roof assembly consisting of 1/2-inch OSB fastened to 2x4 framing at 24-inch centers. The framing bays have been filled with closed-cell spray foam.
UNIVERSITY OF FLORIDA CIVIL AND COASTAL ENGINEERING A test panel (left) in a study at the University of Florida simulates a roof assembly consisting of 1/2-inch OSB fastened to 2x4 framing at 24-inch centers. The framing bays have been filled with closed-cell spray foam.
During the study, the assembly was placed on a pressure chamber and a vacuum pump drew a vacuum that was increased in 15-psf intervals until the assembly failed and the sheathing popped off the framing. For the fully foamed assemblies, this occurred at around 240 psf. The assemblies that had ccSPF fillets installed failed at 160 psf. The assemblies with sheathing alone nailed only with nails (6/12 schedule) failed at about 75 psf.
UNIVERSITY OF FLORIDA CIVIL AND COASTAL ENGINEERING During the study, the assembly was placed on a pressure chamber and a vacuum pump drew a vacuum that was increased in 15-psf intervals until the assembly failed and the sheathing popped off the framing. For the fully foamed assemblies, this occurred at around 240 psf. The assemblies that had ccSPF fillets installed failed at 160 psf. The assemblies with sheathing alone nailed only with nails (6/12 schedule) failed at about 75 psf.
When a roof is not likely to be replaced anytime soon and the sheathing nailing can't be verified (on a tile roof in good condition, for example), contractors in Florida are beginning to employ the "fillet method." This practice uses closed-cell spray foam to help bond the roof sheathing to existing framing and provide a secondary water barrier.
GARY’S ROOFING SERVICE When a roof is not likely to be replaced anytime soon and the sheathing nailing can't be verified (on a tile roof in good condition, for example), contractors in Florida are beginning to employ the "fillet method." This practice uses closed-cell spray foam to help bond the roof sheathing to existing framing and provide a secondary water barrier.

The uplift study also evaluated the benefit of installing a "fillet": a 3x5-inch bead of ccSPF in the corners between the sheathing and the roof framing. The fillet method effectively doubled the uplift resistance of the baseline assembly of 2x4 framing on 24-inch centers sheathed with 1/2-inch OSB nailed on a 6/12 schedule.

The uplift study is one of several recent studies of the structural properties of ccSPF. Tests conducted by Building Science Corporation (BSC) to evaluate the impact resistance of wall systems showed that conventional wood-framed walls do not have the same impact resistance as impact-resistant windows. (That is, walls consisting of studs, 1/2-inch OSB sheathing, housewrap, and siding cannot sustain the impact required by the ASTM E1886 and E1996 missile test, which hurls a 9-pound 2x4 at 50 feet per second.) The only test panel in the BSC demonstration capable of resisting the required impact load included a layer of 1/2-inch OSB sheathing between 1-inch foam insulating sheathing and 2 inches of ccSPF sprayed between 2x6 studs. Surprisingly, BSC found that a wall with foam sheathing, housewrap, and ccSPF (no OSB) performed better in impact tests than a wall with housewrap and OSB sheathing.

The BSC study notes that walls may not have to be built to the same standard as windows, despite these surprising results. When a window fails under impact, the resulting hole in the wall (the entire window) is relatively large, providing a big enough hole to internally pressurize a home, which often leads to catastrophic failure. When a wall fails, the zone of impact is marginally bigger than the impacting face of the projectile. Such an opening may not be large enough to have a catastrophic effect