Small and large-scale experimental studies of soil-arch interaction in masonry bridges
Small and large-scale experimental studies of soil-arch interaction in masonry bridges
复制标题
砌体桥梁土拱相互作用的小型和大型试验研究
DOI：
--
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
J. Wang
中科院分区：
文献类型：
--
作者：
M. Gilbert;C. C. Smith;J. Wang
拱顶、回填土和任何表面荷载之间的相互作用是复杂的，到目前为止，只进行了有限数量的相关试验研究。最近，小规模试验被用来试图分离荷载分散和被动约束的影响，而全尺寸模型桥(3米跨度，安置在一个大的，8.3米长×2.1米高，极其坚硬的试验室中)被用来提供高质量的试验数据，这些数据可以与并行开发的数值模型相关联。小型和全尺寸桥梁都在装有沿一个面的大观测窗口的室内进行测试，以便能够对土壤和拱形运动进行数字成像和定量测量，从而能够随后与数值模型进行关联。给出了主要结果，包括用碎石灰岩和/或软粘土填充的足尺模型桥的结果。382 ARCH‘07-第五届国际拱桥会议将使用小比例模型，但为了最大限度地减少应力水平相关问题的影响，拱筒将由由摩擦界面隔开的刚性块组成，并且在这种比例下将只使用摩擦土。已经使用该设备进行了几项研究(例如，Hulet et al.2006)，并在本文中简要描述了与被动效应和荷载分散效应直接分离有关的研究。要获得真正高质量的数据，没有什么可以替代全尺寸试验，新的试验基础设施已经开发出来，使3m跨度桥梁能够在实验室中进行试验；本文简要介绍了前两次试验情况。由于对平面应变土-拱相互作用问题的了解还很少，所以本文研究的是平面应变土-拱相互作用问题，而不是更复杂的三维问题，并特别努力确保有效地刚性和无摩擦的边边界墙。此外，还确定了平面内拱基固结条件应完全可控。此外，在以前的实验室测试中(例如，墨尔本和吉尔伯特1995年)，经常发现解释土壤压力传感器的输出可能是有问题的。例如，经常从彼此接近的两个压力室获得非常不同的压力历史，但原因往往不清楚。如果也能辨别土体的运动，解释就会容易得多。虽然这是不可能的，如果使用传统的天台墙，这是可能的，如果取而代之的是在桥梁的边缘提供观察窗。因此，这一特征在小型和大型试验台的设计中都有所规定，还允许使用数字成像和粒子图像测速(PIV)技术来有效地自动化收集土体运动的过程。然后，可以将这些运动与并行开发的详细数值模型的结果相关联(例如Gilbert等人。2007)。在简要介绍试验设备的同时，还对粘土和石灰石碎石填土的全桥试验结果进行了比较和对比。表1：小型和大型试验桥参数对比实桥几何跨度(Mm)380 3000仰角(Mm)85 750筒厚(Mm)28 215冠填深度(Mm)36 305横向宽度(Mm)125拱：1010；回填：1045回转高度(Mm)22 220回转长度(Mm)41 320加载布置加载螺旋液压千斤顶加载长度(Mm)上回填：38，上拱：2220加载位置(Mm)94.5桥梁1：750；桥梁2：720回填材料砂碎石灰岩/粘土容重(kN/m)16.5石灰岩：19.1，粘土：22.1phi(度)43.8石灰岩：54.5，粘土：粘聚力(kN/m)0石灰岩：3.3，粘土：78个砖石单位类型铸造丙烯酸固体粘土工程(CL.A)抗压强度(N/mm)154个单位重量(KN/m)13.7 23.2砂浆类型非1：2：9(水泥：石灰：沙)抗压强度(N/mm)1.9单位重量(KN/m)14.4-15.4 2被动约束/活荷载分散分离试验
The interaction between the arch, its backfill and any applied surface load is complex, and only a limited amount of relevant experimental research has been undertaken to date. Small-scale tests, which are inexpensive and quick to perform, have recently been used to try to separate the effects of load-spreading and passive restraint, whilst full-scale model bridges (3m span and housed in a large, 8.3 m long × 2.1 m high and extremely stiff test chamber), have been used to furnish high quality test data which can be correlated against numerical models, being developed in parallel. Both small and full-scale bridges are being tested in chambers incorporating large observation windows along one face to permit digital imaging and quantitative measurements of the soil and arch movements, allowing subsequent correlation with numerical models. Key results are presented, including those from full-scale model bridges filled with crushed limestone and/or soft clay. 382 ARCH’07 – 5th International Conference on Arch Bridges small-scale models would be used, but that, to minimise the influence of stress-level related issues, the arch barrel would be composed of rigid blocks separated by frictional interfaces, and that only frictional soil would be used at this scale. Several studies have been performed using this apparatus (e.g. Hulet et al. 2006) and in this paper a study concerned with direct separation of passive and load spreading effects is briefly described. To obtain really high quality data there is no substitute for full-scale tests, and new test infrastructure has been developed to enable 3m span bridges to be tested in the laboratory; the first two pilot tests performed are briefly described in this paper. Since the plane strain soil-arch interaction problem is still poorly understood this, rather than the more complex 3D problem, was investigated, with special efforts made to ensure effectively rigid and frictionless side boundary walls. Additionally it was decided that the in-plane arch abutment fixity conditions should be fully controllable. Furthermore, in previous laboratory tests (e.g. Melbourne and Gilbert 1995) it was often found that interpreting the output from soil pressure cells could be problematic. For example, very different pressure histories were frequently obtained from two pressure cells in close proximity to each other, but the reason for this was often unclear. Interpretation would be much easier if soil mass movements could also be discerned. Whilst this is not possible if conventional spandrel walls are used, this becomes possible if observation windows are instead provided at the edges of the bridge. This feature was therefore specified in the design of both the small and large-scale test rigs, also permitting use of digital imaging and particle image velocimetry (PIV) techniques to effectively automate the process of collecting soil mass movements. Such movements can then be correlated against the results from detailed numerical models, which are being developed in parallel (e.g. Gilbert et al. 2007). As well as briefly describing the test apparatus, the results collected from pilot full-scale bridge tests involving clay and crushed limestone fill are compared and contrasted. The main features of the small-scale and full-scale bridges which will be referred to in the paper are summarised in Table 1. Table 1 : Comparison of small and large-scale test bridgs Parameter Small-scale bridges Full-scale bridges Geometry Span (mm) 380 3000 Rise (mm) 85 750 Barrel thickness (mm) 28 215 Crown fill depth (mm) 36 305 Transverse width (mm) 125 Arch: 1010; Fill: 1045 Skewback height (mm) 22 220 Skewback length (mm) 41 320 Loading Arrangement Loading screw Hydraulic jacks Loaded length (mm) On fill: 38, On arch: 2 220 Load position (mm) 94.5 Bridge 1: 750; Bridge 2: 720 Backfill Material Sand Crushed limestone / clay Unit weight (kN/m) 16.5 Limestone: 19.1, Clay: 22.1 Phi (degrees) 43.8 Limestone: 54.5, Clay: Cohesion (kN/m) 0 Limestone: 3.3, Clay: 78 Masonry units Type Cast acrylic Solid clay engineering (cl.A) Crushing strength (N/mm) 154 Unit weight (kN/m) 13.7 23.2 Mortar Type None 1:2:9 (cement: lime: sand ) Crushing strength (N/mm) 1.9 Unit weight (kN/m) 14.4 – 15.4 2 PASSIVE RESTRAINT / LIVE LOAD SPREADING SEPARATION TESTS