1、1 INTRODUCTION 1.1 Background Bridges are a major part of the infrastructure system in developed countries. It has been estimated that in the USA about 600,000 bridges (Dunker 1993), in the UK about 150,000 bridges (Woodward et al. 1999), in Germany about 120,000 bridges (Der Prfingenieur 2004) and
2、in China more then 500,000 road bridges (Yan and Shao 2008) exist. Historical stone arch bridges still represent a major part of this multitude. It has been estimated that 60 % of all railway bridges and culverts in Europe are arch bridges (UIC 2005). Recent estimations regarding the number of histo
3、rical railway natural stone arch bridges and culverts in Europe lie between 200,000 (UIC 2005) and 500,000 (Harvey et al. 2007). Also in some regions in Germany about one third of all road bridges are historical arch bridges (Bothe et al. 2004, Bartuschka 1995). Dawen & Jinxiang estimate that 70 % o
4、f all bridges in China are arch bridges. The success of historical natural stone arch bridges - which are often more than 100 years old - is based on the excellent vertical load bearing behaviour (Proske et al. 2006) and the low cost of maintenance (Jackson 2004) - not only in mountainous regions. H
5、owever, changes in loads or new types of loads (Hannawald et al. (2003) have measured 70 tonne trucks on German highways under regular traffic conditions and Pircher et al. have measured 100 tonne trucks) might endanger the safety of such historical structures. Obviously, bridges with an age of more
6、 than 100 years were not designed for motorcars since this mode of transportation has only been in existence for approximately 110 years. The increase of loads does not only include vertical loads but also horizontal loads in the longitudinal direction and perpendicular to the longitudinal direction
7、 of these bridges. For example, the weight of inland waterway ships in Germany has increased dramatically in the last decades, which also corresponds with increasing horizontal ship impact forces (Proske 2003). Furthermore some loads from natural processes such as gravitational processes may not hav
8、e been considered during the design process of the bridges. Especially in mountain regions this Historical stone arch bridges under horizontal debris flow impact Klaudia Ratzinger and Dirk Proske University of Natural Resources and Applied Life Sciences, Vienna, Austria ABSTRACT: Many historical arc
9、h bridges are situated in Mountain regions. Such historical bridges may be exposed to several natural hazards such as flash floods with dead wood and debris flows. For example, in the year 2000 a heavy debris flow destroyed an arch bridge in Log Pod Mangartom, Slovenia and only recently, in Septembe
10、r 2008 an arch bridge was overflowed by a debris flow. A new launched research project at the University of Natural Resources and Applied Life Sciences, Vienna tries to combine advanced numerical models of debris flows with advanced models of historical masonry arch bridges under horizontal loads. T
11、he research project starts with separate finite element modelling of different structural elements of arch bridges such as spandrel walls, the arch itself, roadway slabs, pavements and foundations under single and distributed horizontal loads. Furthermore miniaturized tests are planned to investigat
12、e the behaviour of the overall bridge under debris flow impacts. The results will be used to combine the modelling of the different structural elements considering the interaction during a horizontal loading. Furthermore this bridge model will then be combined with debris flow simulation. Also earli
13、er works considering horizontal ship impacts against historical arch bridges will be used control. The paper will present latest research results. 400 ARCH10 6th International Conference on Arch Bridges gravitational processes (debris flow impacts (Zhang 1993), rock falls (Erismann and Abele 2001) a
14、nd flash floods (Eglit et al. 2007) including water born missiles or avalanches) can cause high horizontal impact loads. 1.2 Historical Events In the year 2000, a debris flow destroyed two bridges in Log Pod Mangartom, Slovenia, one of them was a historical arch bridge. In October 2007 the historica
15、l arch bridge in Beniarbeig, Spain was destroyed by a flash flood. Similarly the Pppelmann arch bridge in Grimma, Germany was destroyed in 2002, in 2007 a farm track and public footpath arch bridge over the River Devon collapsed. Figure 1: Debris flow impact at the Lattenbach (Proske & Hbl, 2007) Fi
16、g.1 shows an example of the historical arch bridge at the Lattenbach, before and after a debris flow event, where the bridge is nearly completely filled with debris. Due to far too expensive solutions or not applicable methods for historical arch bridges it would be very useful if models were availa
17、ble to estimate the load bearing capacity of historical masonry arch bridges for horizontal loads perpendicular to the longitudinal direction. Since intensive research was carried out for the development of models dealing with vertical loads for historical arch bridges, there is an unsurprising lack
18、 of models capable for horizontal impact forces against the superstructure. This might be mainly based on the assumption that horizontal loads are not of major concern for this bridge type due to the great death load of such bridges. The goal of this investigation is the development of engineering m
19、odels describing the behaviour of historical natural stone arch bridges under horizontal forces, mainly debris flow impacts, focused strongly on the behaviour of the superstructure and based on numerical simulations using discrete element models and finite element models. 2 INNOVATIVE ASPECT AND GOA
20、LS 2.1 Innovative Aspects The conservation of historical arch bridges is not only an issue of the preservation of cultural heritage but is also an economic issue since the number of historical bridges in developed countries is huge (Proske 2009). Compared to vertical load cases no models currently e
21、xist for horizontal loads perpendicular to the longitudinal direction. It is therefore required to develop new models dealing with these capacious horizontal loads which include all types of gravitational hazards like avalanches, debris flow, rock falls or flood borne missiles or impacts from modes
22、of transportation. First works related to the development of debris flow design impact forces and the behaviour of arch bridges under such an impact have started already 2007 at the Institute of Alpine Mountain Risk Engineering at the University of Natural Resources and Applied Life Sciences, Vienna
23、 (see Fig.2) Klaudia Ratzinger and Dirk Proske 401 Figure 2 : Examples of the structural behaviour under impacts (left against the pier, right against the arch itself) (Proske and Hbl 2007) This investigation and its results regarding debris flow impact will flow into the development of the new Aust
24、rian code of practice -Norm 24801 for the design of structures exposed to debris flow impacts as well. 2.2 Goal To develop load bearing behavior models of historical natural stone arch bridges under horizontal loads perpendicular to the longitudinal direction, a realistic model of debris flow agains
25、t solid structures has to be implemented indifferent programs. Separate finite element modelling of different structural elements of arch bridges such as spandrel walls, the arch itself, roadway slabs, pavements and foundations under single and distributed horizontal loads are part of this investiga
26、tion. Furthermore miniaturized tests are part of the project to investigate the behaviour of the overall bridge under debris flow impacts. The results will be used to combine the modelling of the different structural elements considering the interaction during a horizontal loading. Furthermore this
27、bridge model will then be combined with debris flow simulation. Also earlier works considering horizontal ship impacts against historical arch bridges will be used. Therefore three models of historical arch bridges are developed: (1) Discrete element program model (PFC), (2) Explicit finite differen
28、ce program model (FLAC), (3) Finite element program model (ANSYS, ATENA). The first and second models are developed to simulate an overall debris flow impact scenario, whereas the third model is used to investigate details, such as single force against a spandrel wall, single force against parapets,
29、 friction at the arch, single impact force against the arch. Results from the impact simulation against the superstructure should give an answer, whether the complete process can be separated into forces acting on the bridge. This reference force (force-time-function) will then be applied on the fin
30、ite element models. The numerical modelling will be accompanied by testing to permit validation of the models. The tests will be carried out as miniaturized tests (scale about 1:2050). Already miniaturized tests of the impact of debris flows against debris flow barriers were already carried out at t
31、he Institute of Mountain Risk Engineering (Proske et al. 2008, Hbl & Holzinger 2003,Fig.3). Based on this experience, miniaturized arch bridges (span about 40 to 50 cm) will be constructed and investigated. Also single parts of the arch structure will be investigated in testing machines, such as beh
32、aviour of a pure arch under a horizontal load. Since the machine cannot be turned, force redirection mechanisms will be used to allow the application of a standard compression test machine from the University of Natural Resources and Applied Life Sciences, Vienna. 402 ARCH10 6th International Confer
33、ence on Arch Bridges Figure 3 : Side view and view from above of the used debris flow impact measurement test set-up (Hbl & Holzinger 2003) 3 CALCULATIONS 3.1 Discrete element methods Discrete element modeling can be done by usingPFC3D (Particle Flow Code 3D) which is used in analysis, testing and r
34、esearch in any field where the interaction of many discrete objects exhibiting large-strain and/or fracturing is required. By using the program PFC3D, materials can be modeled as either bonded (cemented) or granular assemblies of particles. 3.2 Finite element methods The finite element method (FEM)
35、is one of the most powerful computer methods for solving partial differential equations applied on complex shapes and with complex boundary conditions. A mesh made of a complex system of points is programmed containing material and structural properties defining the reaction of the structure to cert
36、ain loading conditions. Nodes are assigned at a certain density throughout the material depending on the anticipated stress levels of a certain area. Two types of analysis are commonly used: 2-D modelling and 3-D modelling. 2-D modelling allows the analysis to be run on a normal computer but tends t
37、o yield less accurate results whereas 3-D modelling shows more accurate results. For this investigation two FEM programs are used: (1) ANSYS (2) ATENA ANSYS is the leading finite element analysis package for numerically solving a wide variety of mechanical problems in 2D and 3D. By using ANSYS, the
38、analysis can be done linear and non-linear, is applicable to static and dynamic structural analysis, heat transfer and fluid problems as well as acoustic and electromagnetic problems. The ATENA program is determined for nonlinear finite element analysis of structures, offers tools specially designed
39、 for computer simulation of concrete and reinforced concrete structural behaviour. Moreover, structures from other materials, such as soils, metals etc. can be treated as well. In the first step finite element methods are used to simulate the behaviour of historical natural stone arch bridges under an impact. Required data for the debris flow models are taken from the database of the Institute of Mountain Risk Engineering as well from the Austrian Railway
