Unbalanced Snow Loading and the Structural Integrity of Circular Arched Roofs

As structural engineers, we rely on meteorologists to provide data on snowfall, rainfall, and wind velocity, but the effects of snow and wind on structures are greatly dependent on structural form. Our knowledge of these effects has been based on intuition, experience, and, more recently, model testing. Professor Allan Davenport, the distinguished authority on wind effects on structures, has observed that the longer these natural phenomena are recorded, the greater the measured maxima will be. A probabilistic approach is required to establish “design values.”

Although the National Research Council has made systematic measurements of roof snow loads on a sampling of buildings across Canada for many years, only by happenstance would a snowfall be combined with critical wind direction, intensity, and duration to produce a critical loading condition on a measured building. Unfortunately, we seem to have been following the failures instead of predicting their probable occurrence, but this is only one aspect revealed by the authors’ experience. They also demonstrate the need to consider more carefully the effect of different forms of snow accumulation, when defining snow load distributions on roofs, because of the sensitivity of some structures, such as trusses, to shear force. It is shown that some structures have failed even though the ground snowfall was less than the design snowload.

The authors mention the tragic collapse of a small-town hockey arena in Ontario and refer to the subsequent examination and check of adequacy of all such structures in Ontario. During this process, I examined 5 or 6 hockey arenas and curling rinks. I do not recall any that met the required standards, but it was possible to reinforce some of them rather than replace the roof entirely.

John Springfield, P.E.

Structural Design Forum Editor