ABSTRACT
EXTENDED ABSTRACT: Truck weight data have been gathered in several countries during the last two decades by using medium span bridges as weighing-in-motion (WIM) devices. Several methods are used to obtain the axle weight and gross vehicle weight (GVW) data from observed bridge responses; all these methods lead to unavoidable errors. While errors in the estimation of GVW, being as large as 10%, might be considered acceptable, errors in estimating axle weights by most methods are unacceptably large. Recently, Japanese researchers have tested a WIM method, called the reaction force method, in which vertical stiffeners on support diaphragms of steel girder bridges are used to measure axle weights. This method, which also requires knowledge of vehicle speed, is yet to be improved to give axle weight data of acceptable accuracy. A test on a Canadian steel girder bridge provided an opportunity to verify the Japanese reaction force method as well as to develop another method to measure truck axle weights. The Shell River Bridge in Manitoba has four steel girders, with T-section stiffeners on both sides of each girder above its bearings. All these stiffeners were installed with sensors to measure vertical strains. In addition, all four girders had strain gauges at the mid-span and at sections near the simple supports to measure longitudinal strains. An attempt has been made in the paper to examine the Japanese reaction force method and the other proposed method so as to suggest possible improvements in the methods, which will lead to economy in the amount of collected data and improvement in the accuracy of predictions.
