Assessment of wind speeds based on damage observations from the Angus (Ontario) Tornado of 17 June 2014

Abstract

A tornado occurred in Angus, Ontario, during the late afternoon hours of 17 June 2014. The authors conducted a damage investigation on the morning following the storm. The damage indicators support the classification of the tornado as an EF-2 on the Enhanced Fujita Scale, including the observation of several complete roof failures of recently-constructed, woodframe houses. Most of the damage to residential homes was contained along two streets, with the tornado appearing to have gone down the backyards between the two. In total, 101 houses were observed to have sustained some level of damage. The evidence suggests that the quality of construction likely affected the performance of failed roofs. A detailed fragility analysis was conducted to assess wind speeds associated with these failures of the roof-to-wall connections. An overturned and lofted box truck provided the opportunity to correlate this failure with adjacent, repetitive failures of roof sheathing, shingles, and garage doors.

Framing Failures in Wood-Frame Hip Roofs under Extreme Wind Loads

“…In combination with improved load sharing within trussed hip roofs, these factors are generally believed to make hip roofs significantly more resistant to damage through the common modes of roof failure. This is also supported by fragility analyses (Kopp et al, 2016;Gavanski and Kopp, 2017).…”

“…Many instances of partial hip roof failures are identified in the data. As with the findings of the fragility analyses in Kopp et al (2016), the observed failures evoke additional questions regarding the likelihood and conditions under which partial hip roof failures may occur. Selected examples of the observed failures from Moore are shown in Figure 3 and discussed below.…”

“…The applied 71.5 mph wind speed is much lower than the failure wind speeds estimated by the fragility analyses by Kopp et al (2016) and Gavanski and Kopp (2017). Both studies considered load sharing and found that at the median probability of failure, the wind speed causing RTWC failure in a hip roof is nearly 155 mph (250 km/h).…”

“…Hip roofs are generally known to perform better than other roof shapes. Fragility analyses by Kopp et al (2016) and Gavanski and Kopp (2017) have even suggested that a single DI for residential structures in the EF-Scale may be inadequate due to significant variations in the estimated failure in wind speeds for different roof shapes, although this has not been quantified in damage surveys.…”

“…Several past studies have investigated the superior performance of hip-roofed homes (Meecham et al, 1991;Meecham, 1992), with some more recent works directly investigating hip roof behavior with regard to roof sheathing (DOD-4) and RTWC (DOD-6) performance (Henderson et al, 2013;Kopp et al, 2016). Meecham et al (1991) performed wind tunnel testing to enhance the technical understanding of hip roof performance and found that there is an important relationship between the pressure distribution and underlying framing configuration in wood-frame roofs.…”