资源描述
英文原文The performance of pressure cells for sprayed concrete tunnel liningsC . R . I . C L AY TO N,J. P. VA N D E R B E R G , G . H E Y M A N N, A . V. D. B I C A a n d V. S . H O P EAbstract:The paper examines the factors that affect the performance of tangential cells embedded in shotcrete tunnel linings. New data, derived from field monitoring, numerical modelling,and calibration tests carried out to simulate the embedment and crimping processes, are presented. These suggest that although well-designed embedded total pressure cells will have cell action factors close to unity, they cannot be assumed to provide reasonable estimates of the stresses within sprayed concrete linings, unless the influences of installation effects, temperature changes, shrinkage and subsequent crimping can be taken into account.Keywords: field instrumentation; tunnels.IntroductionThe pressure cells used for measuring the compressive stresses in shotcrete tunnel linings generally consist of two stainless steel plates with a thin fluid- filled cavity between them. The cavity is connected either to a membrane-type bypass valve or to a vibrating-wire pressure transducer. The use of other direct-stress instruments has been reported in the literature, although infrequently. Pressure cells are typically installed in one of two orientations: radial, to record the stress between the sprayed concrete and the ground surrounding the tunnel; and tangential, to record the hoop stress within the tunnel lining itself. This paper considers only tangential cells.Despite their widespread use in practice, there has been very little research reported in the literature on the use and behaviour of shotcrete pressure cells. Many practitioners remain doubtful of the ability of embedded pressure cells to measure the actual stresses in concrete tunnel linings. In a previous paper reviewing instrumentation for sprayed concrete lined tunnels the present authors noted some of the potential difficulties, stating that it was extremely unlikely that embedded cells be used for monitoring the actual stress in a tunnel lining. Yet, potentially, pressure cells are a valuable source of information that might be used to assess whether tunnel design assumptions are justified, and this paper therefore reports the findings of our further research into this important topic.Factors affecting the pressures recorded by tangential pressure cells in tunnel liningsDirect stress measurement within any medium is made difficult by the many factors that can affect the results. In the case of tangential pressure cells embedded in shotcrete our recent experiences during tunnel monitoring suggest that these are as follows.Cell propertiesThe cell should be constructed so that the stresses in the shotcrete are not significantly modified by its presence. Since the compressibility of the fluid in the cell is less than the surrounding materials it will under-read, but this can largely be compensated for by making the cell wide and thin. The use of cell fluids such as mercury or oil will affect not only the compressibility of cells but also their temperature sensitivity. Changes in temperature will expand the fluid against the surrounding, relatively rigid cell metal and surrounding concrete, and will produce a change in measured stress.Installation effectsThe inadvertent formation of cavities around the cell during shotcreting will lead to a soft measurement system, which will subsequently under-read. Incorrect positioning of the cell within the lining, rotating it towards the radial direction, can also cause it to under-read somewhat, because radial stresses are typically less than 10% of tangential. Indeed the actual thickness of the lining at the point of installation will also affect the interpretation of the stress measurements.Post-installation factorsAs noted above, temperature changes can be expected to lead to changes in measured stresses. Shrinkage during the early life of the shotcrete will result in changes in the recorded stress that are not due to external stress changes. Crimping, which is often undertaken to ensure that pressure cells are properly bedded within the shotcrete , can provide a significant offset to the measured pressures.Numerical and physical experiments, and results from monitoringNumerical modelling and physical simulation have been carried out to assess the actual performance of some stress cells used in practice, and to place their performance in the context of other cell designs.Numerical modelling to assess the effects of cell fluidTo examine the effect of cell fluid on cell performance two idealised circular cells embedded in a block of concrete were modelled under axisymmetric conditions using the finite element package LUSAS. The geometry of the cells and the material properties modelled are shown in Fig. 1. The 160 mm diameter cell is somewhat larger than many of the cells currently in use, whereas the 80 mm cell is smaller, and was considered by the authors to be likely to have an excessive T/D ratio. In the first numerical experiment the effect of the bulk modulus of the cell fluid was investigated, by applying a constant external axial stress and varying the cell cavity pressure . The bulk modulus equivalent to each cell action factor was calculated by integrating the displacements along the surface of the cell cavity. Fig. 2 shows the considerable influence of bulk modulus on cell action factor, but it also shows that when reasonable cell geometries are used cell action factors remain tolerably close to unity when oil is substituted for mercury.Fig. 1. Geometry of cells and properties of materials used during numerical modelling: (a) idealised pressure cell; (b) geometry modelled; (c) material propertiesFig. 2. Effect of bulk modulus of cell fluid on cell action factorPhysical simulationCalibration tests were conducted to evaluate the performance of the vibrating-wire mercury-filled pressure cells, in three phases:(a) During the first phase the manufacturers calibration of the pressure cells was checked by conducting an air pressure calibration on all the cells used.This was done in a 1 m diameter chamber, in the laboratory.(b) The second phase of the experimental work was conducted to investigate whether tangential cells installed under ideal and controlled conditions could produce reliable results. This was done by installing two pressure cells in a precast concrete slab, constructed in the laboratory.(c) The final phase of the experimental work was designed to investigate the performance of the tangential cells under working conditions, as well as to investigate ways of installing the cells to improve their performance. This was done by installing tangential pressure cells in shotcrete slabs, formed in a tunnel under working conditions.The cells used for the experimental work were mercury-filled vibrating-wire cells supplied by Geokon, and with a full-scale range of 20 MPa (Fig. 3). All the pressure measurements were calculated using the temperature correction supplied by the manufacturer.Fig. 3. Vibrating-wire mercury-filled concrete stress cellThe second phase of the calibration testing was carried out to investigate the performance of the tangential pressure cells under ideal and controlled conditions. For this experiment two cells were embedded in a 25 MPa ready-mix concrete slab, 1.0 m high, 1.0 m wide and 0.3 m thick. The two cells were tied to a cage constructed from reinforcing bar meshes, which were identical to those in use in the Heathrow Terminal 4 station tunnels. The experimental set-up is shown in Fig. 4, except that in the first set of experiments two bare cells were used. One bare cell and one precast cell were used in a subsequent experiment, described later in this paper.Fig. 4. Layout of cells embedded in concrete panelDuring the curing period the slab temperature increased to about 32C, and afterwards decreased slowly over several days. Monitoring was carried out until the cell temperatures reached equilibrium with the laboratory environment (Fig. 5).Fig. 5. Temperatures measured after casting cells in ready-mixed concreteAt each load increment the cell readings were observed to stabilise rapidly: creep effects were not apparent. The value of cell action factor was subsequently calculated from pressure cell readings and average applied vertical stresses .The final phase of the experimental work consisted of the evaluation of the tangential pressure cells installed under working conditions. This was done by placing two pressure cells in each of two slabs, similar to the above but constructed using shotcrete in a tunnel at Heathrow Terminal 4. Because the results of the experiment on the cells installed in the concrete slab showed that under ideal conditions the cells seem to perform satisfactorily, it was argued that failure under working conditions might result from installation effects. A major difficulty with the installation of tangential pressure cells is ensuring that no voids are formed in shadow zones around the cell during shotcreting. It was therefore decided to precast one of the cells in each of the two slabs in a tapered concrete block, designed to prevent the formation of shadow zones (Fig. 6).Fig. 6. Detail of cell cast in precast concreteDiscussionTheoretical considerations suggest that a well-designed embedded cell, with high stiffness and a low aspect ratio (T/D), should have a cell action factor close to unity. Our experiments on commercially available cells support this view. The ability of a cell to measure the applied pressure correctly is dependent upon additional factors, however. Offsets due to temperature, crimping and shrinkage must be properly taken into account. The quality of cell installation must be assessed.Although manufacturers routinely supply temperature correction factors for vibrating-wire cells, these correct for the sensitivity of the transducer alone. For a 160 mm radial cell filled with mercury the authors numerical modelling results suggest that a temperature change of 20C will produce a pressure change of the order of 2 MPa. Oil-filled cells can be expected to be more temperature sensitive. A 20C temperature change might produce about a 3 MPa pressure increase, which is of the same order as the tangential stress found in many completed shotcrete tunnel linings. To the authors knowledge, no estimates of the increase in measured stress induced by shrinkage have ever been reported in the literature. In order to make an initial estimate the data from the authors laboratory experiments were reprocessed, using only those data obtained when the concrete slabs were unloaded. If crimping is carried out then the zero offset of the cell is permanently altered. In a real installation the absolute pressure can be recovered only if the pressure change during crimping is carefully recorded. It is suggested that, although crimping is unnecessary if cells are well-designed and installation is good, the initial gradient of the crimping curve should provide a good guide to the cell action factor of the installed cell, with high crimping gradients indicating satisfactory cells. The pressure increases caused by crimping can be eliminated by subtracting them from the values subsequently recorded.ConclusionsUsing numerical and physical experiments, coupled with field observations, this paper has for the first time attempted a rational assessment of the many factors that may lead to embedded shotcrete pressure cells misreading.The data suggest that, unless installation defects are present, the cell action factors of well-designed shotcrete pressure cells are likely to be near to 1. However, other factors need to be taken into account before embedded pressure cell data can be used to determine the true stress in a tunnel lining. Temperature changes immediately after cell placement will be large, and, coupled with the high rate of shrinkage that occurs during the early life of shotcrete, will prevent satisfactory stress measurement during this period. Seasonal temperature changes will cause further changes in pressure cell readings. Strains due to shrinkage of the shotcrete may also significantly increase the measured stresses. Our data suggest that it is possible to predict the temperature sensitivity of the shell/shotcrete system using numerical modelling. Field data, laboratory measurement and estimates based upon an analytical approach are in good agreement.After temperature changes due to cement hydration have ceased, the shape of the crimping curve can be used to assess the quality of cell installation. However, the crimping procedure will generate offset pressures that may be of the same order of magnitude as the actual pressure to be measured in most shotcrete linings. Measurements made after the crimping procedure is completed must be corrected by subtracting the pressure increase observed during crimping. If shadow zones cannot be prevented during installation then the use of precast cells may be advantageous, although the cell action factor of the installation will be modified.The various factors influencing the measurement of the absolute stress in a shotcrete lining cannot be taken account of when routinely interpreting pressure cell data, without the benefit of careful calibration, numerical estimates of thermal sensitivity, and experimental determinations of the effects of shrinkage. We have shown that the effects of temperature, shrinkage and crimping will probably be large, and of the order of the stresses to be measured. However, the cell action factors of well-designed and well-installed pressure cells will be close to unity and, as we have shown, it should be possible to take account of the effects of temperature changes and shrinkage, to estimate the quality of the installation, and to correct forcrimping offset. Despite the potential difficulties the authors believe that tangential pressure cells can still be useful in many tunnelling applications, but only provided great care is taken in the interpretation of their measurements.AcknowledgementsThe authors gratefully acknowledge the support of Heathrow Express, Mott MacDonald Consulting Engineers, and the Engineering and Physical Sciences Research Council of the UK, and the help of Mr J. Barrie Sellers, President of Geokon, Inc.,USA, in reviewing the manuscript. The work described in this paper forms part of a wider research programme now being carried out by the University of Southampton, UK, into the behaviour of SCL tunnels.中文译文喷射混凝土巷道应力测量仪的性能C.R.I.克莱顿，J.P.范德伯格，G.霍曼，A.V.D.哈曼和V.S.霍普摘要：本文研究了影响喷射混凝土巷道应力测量仪性能的因素。新数据通过实地检测、数字模拟、模拟埋设标定试验和褶曲变化进程等进行派生的表达。这些数据表明，虽然精心设计的应力测量仪测出的数据很接近围岩整体运动规律，但是它们不能被假定为对喷射混凝土巷道围岩应力提供了合理的估计，除非将安装影响、温度变化、围岩收缩及后续的褶曲变化等因素考虑在内。关键词：实地检测；巷道1前言用来测量喷射混凝土巷道压缩应力的应力测量仪（以后简称测力仪）通常由两个不锈钢板组成，两个板之间有一个充满液体的管。这个管一般和一个膜式旁路阀门或一个弦式压力传感器连接。在文献中，也提到了一些其他测量应力的器材的作用，虽然不常见。测力仪通常安装在两个方向之一：径向，记录喷射的混凝土与巷道围岩之间的应力；切向，记录巷道内的切应力。本文认为只有切向应力。虽然它们在实践中广泛使用，但在研究报告文献中很少有介绍喷射混凝土巷道测力仪的使用和作用。很多学者仍然保留着对嵌入式测力仪能够测量喷射混凝土巷道实际应力的性能的怀疑。在前一篇审查喷射混凝土巷道测力仪性能的文章中，作者指出了一些潜在困难，特别说明嵌入式测力仪用于监测巷道实际应力是极不可能的。然而，潜在的测力仪是一种宝贵的信息，可能被用来评估巷道设计假设是否合理，因此，这篇文章在这个重要的项目中报告了我们进一步研究的成果。2切向测力仪测量巷道应力的影响因素由于很多因素可以影响结果，所以在任何媒体中直接测量都很困难。在切向测力仪嵌入混凝土的情况下，在巷道监测过程中，我们最近的经验表明如下所示。2.1测力仪特性测力仪应该被改进，以便巷道中的压力不会被它的存在而受到明显改变。当测力仪中的流体的可压缩性小于周围材料，它会读不出数据，但这可以在很大程度上通过将测力仪变宽变长的方法来弥补。测力仪中的流体（如汞或油）的使用不仅会影响测力仪的可压缩性，而且会影响他们的温度敏感性。温度的变化将促使流体对抗环境、相对刚性的测力仪金属和周围的混凝土，并且会使测量的应力产生变化。2.2安装影响喷浆过程中，在测力仪周围无意形成的空隙会导致一个松软的测量环境，这就会使测量仪读不出数据。不正确安装的测量仪，旋转它往径向，有时也会导致读不出数据，这是因为径向应力一般比切向应力小10%。事实上，实际安装时的衬砌厚度也会影响应力测量的结果。2.3安装后的影响因素如上所述，温度的变化将会导致实测应力变化。早期喷浆中的收缩过程导致应力记录的变化不取决于外部应力的变化。经常进行褶曲以确保测力仪正常嵌入巷道内，可以提供一个明显的偏移测量应力。3数字模拟与物理实验和检测结果数字和物理模拟实验已进行，用来评估一些测力仪在实践中的性能，并且将这些性能用于一些其它测力仪的设计中去。3.1数字模拟实验评估压力计流体的影响为了检验测力仪中流体对测力仪性能的影响，参照实验将两个圆形测力仪以轴对称方式嵌入混凝土块中，使用有限元包LUSAS进行测定。测力仪的几何结构和材料属性参照图如图1所示。160毫米直径的测力仪比许多目前正使用的测力仪要大一点，而80毫米直径的要小一点，作者认为可能是T/D比较大。在第一个数字模拟实验中，通过应用一个恒定的外部轴向应力和不同的测力仪管内应力，调查了测力仪流体体积弹性模量的影响。体积弹性模量相当于通过整体的测力仪管曲面位移计算每一个应力单元。图2显示了应力变化因素体积弹性模量的相当大的影响力，但同时还显示了当应用合理的几何形状时，在用汞替代油的情况下，应力单元变化情况仍和整体一致。图1 测力仪的几何结构和用于实验的材料属性图2 测力仪流体体积弹性模量的影响因素3.2物理模拟实验用来评估弦式汞应力测力仪性能的校准测试分为三个阶段：1）在第一阶段，通过对所有使用的测力仪进行一个空气压力试验来检查制造商校准过的测力仪。这个实验是在实验室中一个1米直径的空间中进行的。2）实验工作的第二阶段是研究测力仪在理想和能控制的条件下安装，到底哪个能得出可靠的结果。这通过在实验室中构建的预制混凝土板中安装两个测力仪来进行。3）最后一个阶段是调查切向测力仪在工作条件下的性能，以及研究如何通过改变安装方法来改善测力仪的性能。这通过在处于工作条件下的巷道中形成的混凝土板中安装切向测力仪来实现。用于实验的弦式汞测力仪都质优价廉，而且测量范围达到20兆帕（如图3）。所有的应力测量都通过制造商提供的温度修正来进行计算。图3 弦式汞应力测量仪校准实验的第二阶段是研究切向测力仪在理想和现实条件下的性能。在这个实验中，将两个测力仪嵌入拌好的混凝土板中，这个板高1.0米，宽1.0米，厚0.3米。两个测力仪被绑定到一个用钢筋加固的网格笼中，这个笼和希思罗机场4号站隧道中所使用的是完全相同的。实验设置如图4所示，只是在第一组实验中使用了两个原装的测力仪。本文稍后会介绍，在随后的实验中使用了一个原装测力仪和一个预制的测力仪。在固化阶段，钢坯温度增长到32C，在随后的几天，温度缓缓下降。监测一直进行到测力仪温度和实验室环境平衡为止（如图5）。图4 嵌入混凝土板中测力仪的布局图5 测力仪嵌入混凝土后的温度变化随着每个增量的加载，测力仪的读数迅速稳定：蠕变影响不明显。随后通过测力仪读数和提供的垂直应力的平均值来计算测力仪单位应力值。实验的最后阶段包括对处于工作条件下的切向测力仪性能的评估。将两个测力仪分别放在两个板中，和上述的相似，但使用的是希思罗机场4号站隧道的混凝土板。因为这个实验的结果表明在理想条件下，测力仪似乎表现得更让人满意，有人认为在工作条件下会失败可能是安装影响所导致的。安装切向测力仪的一大困难是确保在处于混凝土中的测力仪周围没有空隙所形成的“阴影区”。因此，当局决定在一个圆锥形的混过凝土块中预制两个测力仪中的一个，旨在防止形成“阴影区”（如图6）。图6 嵌入混凝土中的测力仪的详细信息4讨论理论思考表明一个精心设计的具有高刚性和低长宽比（T/D）的嵌入式测力仪应该有一个单元接近整体。我们对商业上可用的测力仪的实验支持这一观点。然而，一个测力仪能正确测量所施加的应力的能力取决于其它因素。除了温度的影响，褶曲和压缩的偏移量也必须适当地考虑在内。必须评估测力仪的安装质量。虽然制造商经常为弦式测力仪提供温度修正，但他们仅仅是对传感器灵敏度进行了校正。对于160毫米直径的汞测力仪，作者的数字模拟实验结果表明温度每变化20C，测出的结果会有2兆帕的变化。油测力仪有望有更高的温度敏感性。20C的温度变化可能产生约3兆帕应力的增加，这是在很多已完成的喷射混凝土巷道中发现的切应力变化的共同规律。据作者所知，在已报告的文献中，没有因收缩导致已测应力增加的评估。为了使作者从实验室实验中得到的初步估计数据得到进一步的处理，当混凝土板卸载时仅用这些提到的数据。如果进行褶曲，则测力仪的零偏移量会永久改变。在实际安装中，只有仔细记录褶曲过程中的应力变化，才能使绝对应力恢复。有人建议，虽然当测力仪经过精心设计，且安装不出错时，褶曲是不必要的，但是褶曲曲线的起始梯度应该为安装好的测力仪提供一个很好的指南，并且褶曲高梯度标示令人满意的测力仪。褶曲产生的应力增量可以通过随后的记录值减去增量来消除。5结论通过数字和物理模拟实验，再加上实地观测，本文第一次尝试对众多影响因素进行理性的评估，这些因素可能会导致对喷射混凝土巷道测力仪做出错误判断。数据表明，除非安装缺陷存在，精心设计的测力仪的应力单元值才可能和1接近。然而，只有将其它因素考虑在内之后，嵌入式测力仪得到的数据才能用于确定巷道中的真实应力。在测力仪布局增大后，温度会马上发生变化，再加上早期发生在混凝土中的高频率收缩运动，温度变化将会给这个时期令人满意的测量造成困扰。季节温度变化将会对测力仪读数造成进一步的影响。由于混凝土收缩产生的应力也可能使测量的应力大大增加。我们的数据表明，预测使用数字模拟的喷混凝土系统的温度敏感性是可能的。实测数据、实验室测量和基于一种解析方法的估算达成了一致。在水泥水化引起的温度变化停止后，褶曲曲线的形状可以用来评估测力仪的安装质量。然而，褶曲过程将会产生可能和多数巷道中实测应力同一个数量级的应力偏移量。褶曲过程结束后，测量的数据必须通过减去观测褶曲过程中产生的应力增量进行修正。如果测力仪安装过程中不能阻止阴影区的形成，那么使用预制测力仪会更有利，虽然这会使测力仪的安装因素改变。当按惯例解释测力仪数据时，影响喷射混凝土巷道绝对应力测量的各种数据可以不用考虑在内，不包括小心校准、热灵敏度的数据估算及收缩影响的实验测定的好处。我们已经说明了温度、收缩和褶曲的影响可能会很大，包括要测量的应力的顺序。不过，精心设计和安装合理的测力仪应力单元会接近整体，如我们所示，应该尽可能将温度和收缩的影响考虑在内，评估安装质量，修正褶曲偏移量。尽管有很多的潜在困难，作者扔坚信切向测力仪在很多巷道的应用程序中是有作用的，但只有在测量中非常认真才能让它实用。6鸣谢作者衷心感谢希思罗机场快递、莫特 麦克唐纳咨询工程师、英国工程与自然科学研究委员会的支持，以及J.巴里 塞勒先生的帮助，感谢他们对本文的大力支持。本文描述的实验形成了一个更广泛的研究项目，目前被英国南安普顿大学提出，用于研究SCL巷道的表现形式。
展开阅读全文