By Greg Marks, P.E. - Department Head, Forensic EngineeringA 100 foot by 80 foot section of the overframed roof at the rear had peeled off and curled over onto the front half of the roof. Jeff was looking at the underside of that overframing!
It was late in the afternoon and the early evening sky just didn't look right. Jeff, the facilities manager for a large distribution warehouse had seen storms come up quickly before, but this one felt different. The wind was the first thing he noticed – strong gusts that sent him running to his car. As he sat in the parking lot, he had a full view of Building #2 and no idea of what he was about to witness. The storm was upon him in seconds.
Building #2 is a 160 foot by 540 foot steel warehouse building, with an eave height of about 40 feet. The original construction dates from the 1960’s and consists of a steel frame with a flat roof. For maintenance reasons, it was decided around 1994 to construct a conventional gable style roof on top of the existing flat roof. This over framed roof consists of 2x4 walls at 10 feet on center supporting 2x6 rafters and corrugated metal decking. All members are nailed together, except the wall sole plates, which are screwed to the roof deck and the roof deck which is screwed to the rafters. According to Jeff, this roof had served the building well, surviving the ‘blizzard of 1996’ and other heavy snows. A good investment by any appraisal.
As Jeff looked on and the wind picked up, he thought he must be seeing things. It appeared to him that something was flapping at the rear-left side of the building – something big. At first, since visibility was poor, he didn’t think much of it. But, as he looked closer, he realized what he was seeing. A 100 foot by 80 foot section of the overframed roof at the rear had peeled off and curled over onto the front half of the roof. Jeff was looking at the underside of that overframing!
Looking to the right, Jeff noticed the same flapping motion on the right side of the roof. This time, an 80 foot by 80 foot section of the roof on the front side, not only lifted, but was completely blown off. The section of roof landed on the adjacent roof and continued sliding off the side of the building. About 10 tons of lumber and decking crashed through power lines and landed on the tractor trailers parked at the dock doors. Luckily, Jeff was parked on the other side of the parking lot. In 10 minutes, the storm was over and the sun was shining.
How could this have happened? The engineering assessment revealed a classic wind uplift failure. The over framing system was exceedingly light and, therefore, provided very little dead load resistance to uplift forces. Further, the nailed connections also provided very little resistance to uplift forces. The screwed connections, however, had not failed! It was clear that there had been no engineering consideration of these forces in the connections. Indeed, in today’s construction of wood-framed roofs, it is increasingly unlikely to see connections that do not utilize gage metal uplift connectors (commonly called hurricane ties). This storm, as it turned out, was not hurricane-force although gusts at the nearby airport where noted to exceed 60 mph. It is conceivable that a micro-burst produced winds in excess of code design wind loads but, based on the connections observed, winds this strong would not have been required to cause the failure.
As strong as the roof proved in resisting snow loads, it was no match for the uplift pressures of a strong windstorm. Fortunately, in this case, there were no injuries but luck was indeed at work. Had the portion of roof on the left side blown off, it would have easily cleared the property line fence and landed on a busy street and possibly some homes.
In my experience, the least respected design force (besides seismic if you live on East Coast!) tends to be wind uplift. There is something about the suction power of wind that does not register with the owner of a building or his well-meaning contractor. It is not intuitive. Extremely light roof systems tend to be the rule rather than the exception. Couple that with minimal connections and you have a recipe for failure. In this case, the rebuild incorporated a complete connection retrofit using gage metal hurricane ties.
Greg Marks, P.E. heads the Forensic Engineering Department of Barry Isett & Associates which provides 24/7 response in emergencies. Greg’s team of structural engineers investigates materials, products, structures or components that fail or do not function as intended. Greg’s clients include other design professionals, project owners, insurance companies, construction firms. For more information, contact Greg directly at 610-398-0904 or firstname.lastname@example.org