Study of vehicle roll-over stability in strong winds

High-sided, lightly loaded vehicles are known to be prone to accidents in strong, gusty winds. Although wind-related roadway accidents may occur anywhere, vehicles are more sensitive to wind while passing over bridges, given the higher road elevations, exposure, and possible speed-up effects at various bridge locations compared to ground level roads. Our investigation set out to develop a mathematical engineering model that would quantify the roll-over conditions encountered by various types of vehicles, specifically on long-span bridges.

We began by reviewing videos of wind-roll-over events, we determined that such accidents are typically due to direct gust loads, with little contribution from other aeroelastic effects, such as lift attributable to vehicle speed. The videos indicated just prior the tip-over, one side of the vehicle is typically slightly tilted for as long as 0.5 seconds, while the driver struggles to steer back the vehicle, before it finally surrenders and slowly rolls over.

We identified that the two primary variables involved in determining the effect of wind on vehicle stability are the specific vehicle characteristics and the aerodynamic effects of the bridge superstructure, which vary with vehicle location and the orientation of the wind. We analyzed general wind effects on four representative vehicle types – a 48-foot-long tractor-trailer truck, a large intercity bus, a courier van, and a full-sized SUV.

To estimate critical wind speeds, we tested scale models of the vehicles on 1:70 scale models of two different types of bridges, a double-deck suspension bridge and a multi-span box girder structure. Each bridge model was rigidly attached to the turntable in our wind tunnel, which enabled us to test for winds coming in any direction at the scaled deck elevation. We built representative wood and plastic models of the four vehicle types at the same scale as the bridge models, then ran wind tunnel tests with the vehicles at various locations on the two bridges.

We applied turbulent flow in all the wind tunnel tests, which was important in simulating the small-scale turbulence on each bridge type and properly replicating the shear-layers separating and reattaching on the deck and vehicle models. We also ran ground tests to provide a base reference.

High-precision six-axis transducers measured the aerodynamic forces and moments experienced by each vehicle at a high sampling rate – 500 measurements per second over a period of 60 seconds.

We also built a 15-foot model of the main span of Confederation Bridge, which connects the Canadian provinces of Prince Edward Island with mainland New Brunswick, in the same 1:70 scale, and conducted wind tunnel testing with the model of the 48-foot-long tractor-trailer truck on the bridge for calibration purposes. The Confederation Bridge testing provided comparative tests and analyses for this vehicle that enabled us to compare results of the wind tunnel tests with the bridge’s policy for traffic control in strong winds. The critical roll-over speed found on Confederation Bridge for the empty tractor-trailer is about 40 mph, which is close to the adopted bridge traffic restriction speed of 44 mph (70 kph). Given that the Confederation Bridge has more than 15 years of records, this comparison has become a useful calibration of the analysis and an appropriate safety factor that was adopted to generate recommendations.

This results of this study were used to develop traffic restriction guidelines for a specific bridge. The study also yielded general findings that continue to provide valuable guidance for both the design and operation of safe roadways and bridges:

1. Using this approach, critical “roll-over” wind speeds on a bridge at various locations and for specific types of vehicles can be predicted with a precision that was not possible before. For example, we found the highest wind loads in the outside lanes on the upper deck of the two-level bridge. For the lower deck, wind loads were generally lower than those on the upper deck. However, we also observed the large edge-girder trusses offering additional protection from the wind for vehicles in the outer lanes.
2. Reducing vehicle speed does not necessarily help to improve roll-over stability – it remains more or less the same, although the critical wind direction may change.
3. Critical wind speeds depend heavily on deck cross-section and vehicle location. They also differ from those found on roads in open terrain. This highlights the need for conducting bridge-specific wind tunnel studies in order to develop reliable traffic restriction guidelines.

All newly constructed long-span bridges need to establish guidelines that dictate when to restrict traffic under high winds. The methodology we developed enables us to develop bridge-specific guidelines that balance safety with the need to keep the bridge open.