Field-based evaluation of local scour in complex piers of a bridge on the Dhaleswari river

Abstract

Bangladesh is a land of rivers. Bridge, being an essential hydraulic structure for road and rail communicationsof the country,the Government of Bangladesh have already constructed a large number of bridges through its concern engineering departments. Many bridges would also need to be constructed in future. But this very essential and expensive bridge for road and rail communications could be at risk of failure due to inadequate consideration of local pier scour. Over-prediction of scour also has cost implication. After the historical flood of 1998, Bangladesh experienced a bridge-failure incident near Dhaka due to local scour. After the cyclone Sidr in 2007,a part of a bridge was collapsed in Patuakhali District. So, the bridge planners, designers and implementers require predictionof local scour as closely as possible. Practicing engineers in Bangladesh have chosen a wide range of empiricalformulae for predicting local scour. Most of these formulae are experimental in nature. It is not yet known how good they are in the context of our alluvial river and seasonal hydrology. In this study, performance of some selected empirical formulae used in Bangladesh for scour estimation is assessed with respect to the field data under simple pier and complex pier considerations. The study also evaluates the suitability of the HEC-RAS model for scour prediction in the context of Bangladesh. For this purpose, local scour, water level, velocity, etc.were monitored throughout the pre-monsoon, monsoon and post-monsoon seasons of 2017 at a newly constructed bridge over the Dhaleswari River in Manikganj District. A total of five primary data set was collected. Other relevant secondary data required in the study was collected from relevant government and private organizations. The findings of the study reveal that the discharge in the river varied from 250 m3/s on 20 May to 2267 m3/s on 19 August to 772 m3/s on 16 October, 2017. The 20-year flood discharge was 2040 m3/s and the estimated peak discharge was about 2267 m3/s. Thus, the 2017 flood was close to the design discharge condition of the bridge. The observed scour levels varied from -3.12 m PWD on 20 May to -4.98 m PWD on 19 August to -2.81 m PWD on 16 October, 2017. The maximum scour occurred at Pier No. 4 from the left bank of the river on 19 August during the peak flood. The maximum predicted scour level was -8.94 m PWDbased on simple pier formulation usingLacey’s equation. The maximum predicted scour level was -12.70 m PWD based on complex pier formulation using Jain and Fischer’s equation. All the selected empirical equations as well as the HEC-RAS model gave higher values of pier scours than those measured in the field. The discrepanciesbetween the measured and estimated values were more if complex pier formulation was used. The modified Lacey’s equation estimated local scours better performed for simple pier consideration and over-estimated for complex pier consideration compared to original Lacey’s equation. Estimated values of local scour followed the field observed trend of temporal changes for simple pier whereas estimated values had different trend while considering complex pier. Estimated values of local scour by using Melville’s equation and considering simple pier provided the same value for all the data set for being considered asa narrow pier. On the other hand, the Melville’s equation followed the pattern of field observed local scour except for peak flow when the pier was considered as a narrow pier. Among the equations, the FHWA method predicted the scour values (e.g., -5.03 m PWD on 19 August for simple pier) which were closer to the field observed values for both simple and complex pier considerations. The HEC-RAS model gavethe closestscours (e.g., -6.94 m PWD on 19 August) in complex pier consideration. This indicates that theHEC-RAS model and the FHWA method are suitable for local scour prediction for alluvial rivers in Bangladesh and should be duly considered in pier scour estimation.