Indirect ice load monitoring and strength analysis of a steel gate considering uncertainties

Abstract

One of the most significant challenges in analysing steel gate strength is to obtain plausible ice loads that are exerted on the gate. The indirect monitoring of ice-induced loads is an inverse problem of structural mechanics, and various numerical models have been developed over the years for estimating ice loads with reasonable accuracy. However, the mathematical models used to estimate ice loads are usually affected by uncertainty factors such as uncertainties in the transfer matrix, which reduces the reliability of the gate strength analyses. In this study, a D-optimal design combined with a sequential exchange algorithm or C-optimal design method was used to identify 13 candidate arrangement schemes of strain gauges based on an assumption that the transfer matrix is accurate. Every candidate arrangement scheme of the strain gauges on the steel gate corresponds to a certain number of strain gauges. An experimental structure was designed, and a benchmark experiment was conducted to estimate the error of the transfer matrix—i.e., the uncertainties in the mathematical model. Monte Carlo simulations were performed to investigate the effects of the uncertainties of the mathematical model on the estimated ice loads, and the optimal strain gauge locations were then determined. Based on the optimal arrangement scheme, the upper and lower limits of the estimated ice loads corresponding to different probabilities were obtained with the structural monitoring data of a steel gate in a cold region using the Monte Carlo simulation method. Finally, the gate strength was analysed based on the upper limits of the estimated ice loads corresponding to a probability of 99.74%. The analysis results indicate that the ice load did not threaten the gate safety during the structural monitoring period.