实测车流作用下大跨桥梁荷载效应极值外推法
Extrapolating method of extreme load effects on long-span bridge under actual traffic loads

提出多种密度随机车流作用效应极值的概率叠加方法,外推了公路实测车流作用下大跨桥梁的车载效应极值; 阐述了基于Rice公式的界限跨阈率叠加原理,并验证了其正确性; 基于中国某高速公路长期监测车流数据模拟了稀疏、一般和密集3种状态的随机车流,应用界限跨阈率叠加模型估算了某混凝土斜拉桥主梁的最大弯矩。研究结果表明:根据某高速公路实测车流数据,稀疏、一般、密集车流密度分别约为1.7、5.0、8.3 veh？min-1; 在数值算例中,车辆质量为45 t车型占有率由0增加至80%导致最大车辆质量仅下降1.2%,而车辆质量为50 t的车型占有率由0增加到20%导致最大车辆质量增加14.4%,说明多个平稳随机过程组合而成的非平稳随机过程的极值主要是由数值较大的随机过程决定; 采用跨阈率叠加方法对某混合车流车辆最大质量的外推误差为2.55%,验证了将实际混合车流数据进行“车流离散”和“极值概率叠加”后得到的车载效应极值的方法的可行性; 密集车流占有率从0逐渐增加至5%导致斜拉桥主梁弯矩极值增幅为33.45%; 某斜拉桥设计年限内的年均交通量增长率为2.8%,对应的主梁跨中正弯矩极值超越设计标准值的概率为0.83,高于设计要求,需对桥梁车流量采取管控措施。
A probabilistic superposition approach was proposed for investigating the extreme load effects of stochastic traffic flows with multiple densities. The approach was utilized to extrapolate the extreme values of vehicle load effects on the long-span bridges using the measured traffic data. The principle of superposing level-crossing rate based on the Rice's formula was explained, and its validity was proved. Three types of stochastic traffic flows, including the sparse flow, normal flow and dense flow, were simulated based on the long-term monitored traffic data of a highway bridge in China, and the maximum bending moment of a concrete cable-stayed bridge was analyzed based on the level-crossing superposition model. Analysis result shows that based on the measured traffic data of a highway, the densities of free flow, normal flow and busy flow are 1.7, 5.0 and 8.3 veh？min-1, respectively. In the numerical example, when the occupancy of 45 t-vehicles increases from 0 to 80%, the maximum vehicle mass decreases by only 1.2%. But when the occupancy of 50 t-vehicles increases from 0 to 20%, the maximum vehicle mass decreases by 14.4%. This phenomenon indicates that the extreme value of a non-equilibrium random process composed by some stationary random processes is mostly depended on the random processes with higher values. The maximum extrapolating error of the maximum vehicle mass is 2.55% for the mixed traffic flow using the level-crossing superposition approach, which explains that the extrapolating approach of extreme load effects on long-span bridges based on the principle of vehicle dispersion and superposition of extreme value probability is feasible. The increase of occupancy of dense traffic flow from 0 to 5% leads to an amplification of 33.45% for the maximum bending moment of the girder of a cable-stayed bridge. When the annual traffic growth rate of a cable-stayed bridge is 2.8% in the design lifetime, the probability that the mid-span extreme bending moment of the bridge girder exceeds the design standard value is 0.83