Firebrand attack on ceramic tile roofing assemblies: Fid-Bau Portal

Firebrand attack on ceramic tile roofing assemblies

Manzello, Sanuel L. / Hayashi, Yoshihiko / Yoneki, Takefumi / Yamamoto, Yu

The present study investigated the ignition of ceramic tile roofing assembles (Spanish tile roofing) to a controlled firebrand attack using the Firebrand Generator. Aged or weathered ceramic tile roofing assemblies were simulated by not installing tar paper. For simulated aged ceramic tile roof assemblies, without the installation of bird stops, the firebrands were observed to be blown under the ceramic tiles. Eventually, several firebrands would collect and would produce smoldering ignition (SI) within the OSB (oriented strand board) base layer. With continued application of the airflow, holes were formed within the OSB and eventually the SI would transition to flaming ignition (FI). Simulated aged ceramic tile roof assemblies, with bird stops installed, were also constructed for testing. Even though bird stops were installed, many firebrands were able to penetrate the gaps that exist between the ceramic tiles and the bird stops. These firebrands were observed to produce SI within the OSB base layer; holes were observed in some cases within the OSB base layer. The SI ignition never transitioned to FI when bird stops were applied. The use of tar paper was then used to simulate a newly constructed ceramic tile roof assembly. With the application of tar paper, experiments were conducted first without bird stops installed. Once again, firebrands were blown under the ceramic tiles. The firebrands were able to burn several holes within the tar paper and produce SI within the OSB base layer. The SI was not intense enough to result in the production of holes within the OSB base layer. Tests were then conducted that considered the application of tar paper with bird stops installed. These conditions resulted in no ignition in the tar paper and thus no ignition within the OSB layer. It appears that the combination of the bird stops coupled with the tar paper provided a substantial barrier to ignition. These results are the first ever experiments to ascertain the vulnerabilities of ceramic tile roofing assemblies. The experiments using the Firebrand Generator are extremely conservative; the firebrand attack lasted for six minutes. In real WUI (Wildland-Urban Interface) fires and urban fires, firebrand attack has been observed for several hours and with winds in excess of 20 m/s. Even under such conservative conditions in the present experiments, ceramic tile roofing assemblies were vulnerable to ignition from firebrand showers. It is important to note that ceramic tile roof assemblies found on top of many homes also contain poorly aligned tiles. All of the present data is presented for neatly aligned tiles. Additional experiments are in progress to investigate the influence of gaps within the tiles on the ignition propensity of ceramic tile roofing assemblies.