Riverbed deformation around multiple pile bridge piers under ice-covered conditions - experimental study in a large-scale flume and numerical simulation

在加拿大，一些地区的冰盖可以持续长达六个月。覆冰条件下最大流速位置比无冰条件下更靠近河床。因此，冰盖下桥墩周围的河床冲刷深度大于无冰盖时。过度冲刷可能会导致桥墩基础损坏，例如1966年不列颠哥伦比亚省史密瑟斯16号公路桥梁倒塌和2018年安大略省布鲁斯港桥梁倒塌。该研究将有助于我们更好地了解冰盖下不同尺寸、形状和布置的多个桥墩周围的局部冲刷过程。这项研究将为土木工程师提供可靠的结果，可用于修改考虑冰盖影响的桥梁设计指南。本研究将解决科学文献中的一些重要空白，包括：（a）评估冰盖下多个桥墩周围的局部冲刷过程，（B）调查冰盖下多个桥墩的尺寸/形状/布置对局部冲刷过程的影响，（c）建立描述冲刷过程对水流强度、颗粒尺寸、覆盖条件、多个桥墩布置、和压缩马蹄涡;（d）建立了冰盖下多桥墩局部冲刷过程的数值模拟模型。试验将在长80米、宽2米、深1.3米的大型水槽中进行。将使用五种不同的非均匀砂。泡沫聚苯乙烯板将用作模拟冰盖（光滑、粗糙和非常粗糙）。无阻流和覆冰条件下的实验数据分析将集中在无量纲变量，如剪切雷诺数，希尔兹参数和桥墩尺寸。将建立方程以确定冰盖下的冲刷深度。 模拟冲刷过程将包括模拟一个强耦合系统，包括流场，泥沙输运和河床形态。数值模型将包括三个模块：用于模拟多个桥墩周围流场的水动力模块、用于模拟悬浮泥沙输运的泥沙模块以及用于说明河床地形变化的河床变化模块。本文还将对多桥墩周围的局部冲刷进行数值研究。多桥墩流场模拟的重点之一是解决基础设施周围的涡系。为了求解涡系，将使用非定常RANS（雷诺平均Navier-Stokes方程）或大涡模拟来求解马蹄涡。本文首先采用标准k-ε湍流模型对冰盖下的多桥墩绕流场进行数值模拟。基于RNG的k-ε湍流模型也将用于模拟。

In Canada, ice cover can last up to six months in some regions. Under ice covered conditions, the location of maximum flow velocity is closer to the riverbed than during open flow conditions. Thus, riverbed scour depth around bridge piers under ice cover is greater than in the absence of ice cover. Excessive scour can cause damage to bridge pier foundations, such as occurred with the collapse of the Highway 16 bridge in Smithers, BC in 1966 and the bridge collapse in Port Bruce, ON in 2018. The proposed research will help us better understand local scour processes around multiple piers with different dimensions, shapes and arrangements under ice cover. This study will provide civil engineers with reliable results that can be used to modify bridge design guidelines that take the impact of ice cover into consideration. This research will address some important gaps in the scientific literature, including, (a) assessing local scour processes around multiple piers under ice cover, (b) investigating the impacts of the dimension/shape/arrangement of multiple piers on local scour processes under ice cover, (c) establishing equations describing the dependence of scour processes on flow intensity, particle size, cover conditions, arrangement of multiple piers, and compressed horse-shoe vortex under ice cover, and (d) developing a numerical model for simulating local scour processes around multiple piers under ice cover. Extensive experiments will be carried out in a large scale flume which is 80-meters long, 2-meters wide and 1.3-meters deep. Five different non-uniform sands will be used. Styrofoam sheets will be used as simulated ice cover (smooth, rough and very rough). Analyses of the experimental data under both open-flow and ice-covered conditions will focus on non-dimensional variables, such as shear Reynolds number, Shields parameter and bridge pier dimensions. Equations will be established to determine the scour depth under ice cover. Modeling the scour process will include modeling a strongly coupled system involving flow field, sediment transport and bed morphology. The numerical model will consist of three modules: a hydrodynamic module for the simulation of the flow field around multiple piers, a sediment module for the modeling of suspended sediment transport, and a bed variation module to account for the change in the bed topography. Numerical studies of the local scour around multiple bridge piers will also be undertaken in the present research. One of the priorities for the simulation of flow fields around multiple bridge piers is to solve the vortex systems around the infrastructure. To solve the vortex systems, either unsteady RANS (Reynolds-averaged Navier-Stokes equations) or Large Eddy Simulation will be used to resolve the horseshoe vortex. The standard k-epsilon turbulence model will be used first in the simulation for the flow field around multiple bridge piers under ice cover. The RNG-based k-epsilon turbulence model will also be used in the simulation.