Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Doctor of Philosophy

Program

Civil and Environmental Engineering

Supervisor

Hong,Hanping

Abstract

Wind load and snow load are two of the environmental loads that need to be considered for structural design. This study compared the statistical characteristics of the annual maximum thunderstorm and non-thunderstorm winds based on the updated wind speed dataset and carried out the extreme value analysis for wind load by treating the synoptic and thunderstorm winds separately. The results indicated that for southern Ontario and the southern part of the Prairies, the extreme winds could be dominated by thunderstorm winds. The consideration of both thunderstorm winds and synoptic winds separately to recommend the reference wind velocity pressure for the codified structural design was made.

Extreme value analysis for snow depth and snow load was also carried out. The statistics of and probabilistic models for the annual maximum ground snow depth and ground snow load (as well as snowpack bulk density) were assessed. The rain component of the roof snow load was evaluated, and the correlation between the snow component and rain component of the annual maximum roof snow load was investigated. It turns out that the correlation between the snow component and rain component of the annual maximum roof snow load was negligible. This suggested that the use of the sum of the ground snow load and rain load to evaluate the roof snow load could be conservative, and the use of the well-known square-root-of-the-sum-of-squares (SRSS) rule to evaluate the roof snow load based on the snow component and rain component was recommended.

By using the newly developed statistics of the wind and snow, a reliability-based design code calibration for the National Building Code of Canada was carried out. The analysis considered the stationary extremes derived from observed meteorological data and included the nonstationary climate change effects. New design load factors and companion load factors for the ultimate limit state and serviceability limit state by considering wind load and snow load were recommended. Scaling factors that account for climate change effects were also suggested.

Summary for Lay Audience

Wind load and snow load are two of the environmental loads that need to be considered for structural design. This study investigates the largest wind and snow loads that could affect the performance of structures. It also evaluates the effect of the structures in a notional sense. Moreover, it evaluates the potential climate change effects of the largest wind and snow loads and on structural safety.

The largest wind loads are assessed in terms of extreme wind speed by considering the thunderstorm and non-thunderstorm winds separately. Such an assessment was not available in the literature for Canada. The assessment allowed us to map the wind hazard and to identify regions that are prone to thunderstorm winds. Maps are developed and recommended for practical use.

The evaluation of the snow loads considers both the snow component and rain component as indicated in the National Building Code. The statistics of and probabilistic models for the annual maximum ground snow depth and ground snow load (as well as snowpack bulk density) were assessed and presented. An evaluation for rain and snow components of the roof snow load indicates that extreme rain and snow events are less likely to happen concurrently. This result indicated a modification for the current snow load evaluation method that leads to a relatively conservative estimation.

With the newly developed statistics of the wind and snow, new wind and snow load factors were suggested for the National Building Code of Canada to ensure the structural reliability achieves the required safety level. In addition, a new practical approach in defining the design wind load and snow load by considering climate change effects is recommended.

Available for download on Wednesday, September 01, 2021

Share

COinS