Validating numerical model for bridge pier scour estimation through physical modelling – Case study of Kelanisiri Bridge

Abstract

Bridge scour is the erosion of sediment around bridge piers by flowing water, causing a significant risk to bridge stability and safety. Scour depth, the extent of sediment erosion around bridge piers or other structures caused by flowing water, is a critical parameter in bridge engineering. Generally, there are three types of scour phenomena that impact the performance and safety of bridges as local scour, contraction scour, and degradational scour. Understanding these are essential for identifying and reducing any dangers to the stability and safety of bridges. This research aims to investigate the scouring phenomenon around the bridge piers of the Kelanisiri Bridge in Sri Lanka, understand its mechanisms, develop predictive models, and propose effective mitigation measures. The Kelanisiri bridge spans 134 m and is 10.4 m wide, has two cylindrical piers with a diameter of 2.5 m and has undergone scouring around piers, posing a severe danger to its stability. Also, this aims to contribute to the creation of sustainable solutions by examining the scouring processes and assessing the effects of variables like sediment types, flow velocities, and hydrological conditions through a combination of field investigations, laboratory experiments, and numerical modelling techniques. Further, this research will help to improve Sri Lanka's transportation infrastructure by ensuring the country's road network remains resilient and connected in the face of changing hydraulic conditions and sediment transport processes in river systems. In this research, a laboratory scale physical model is developed to validate the results obtained from a numerical model developed using HEC-RAS software. The physical model replicates the bridge pier and surrounding riverbed at a scaled-down size, using similar materials. This allows for direct visualization of the scour process and the collection of detailed scour depth data at various locations around the pier. In the numerical model, HEC-RAS has capability to incorporate sediment transport equations, conduct extensive flow pattern studies, simulate complex river geometries, and offer a broad array of analytical tools. Comparing the scour depths obtained in the physical model with the predictions from the numerical model allows for an assessment of the validity and limitations of the numerical simulations. The performance of the HEC-RAS model in replicating the physical model's scour depth measurements was assessed using R-squared (R²) and Root Mean Squared Error (RMSE). R², ranging from 0 to 1, indicates the proportion of variance in the observed scour depths explained by the model's predictions. A good model exhibits both a high R² (strong correlation) and a low RMSE. The outcomes of the physical model comply well with the HEC-RAS numerical model results, with R2 value of 0.87 and RMSE value of 0.12. Therefore, the accuracy of the numerical model, which had been designed to predict scour depth for the bridge piers at Kelanisiri bridge can be validated through this laboratory-scale experimentation. Finally, these discoveries will be critical in assuring the safety and durability of bridge constructions in Sri Lanka