From the steelstrains measured by strain gauges installed on longitudinal andtransverse bottom reinforcements where loading was applied asshown in Fig. 7, the ratio of transverse strain to longitudinalstrain could be observed (Figs. 8 and 9). From the results,it can be attributed to the ratio of transverse moment andlongitudinal moment induced by applied load. The distributionof longitudinal strain is similar to that of transverse strain inthe direction of transverse length (Figs. 8(a) and 9(a)), butthe longitudinal strain in the direction of longitudinal length issuddenly decreased away from the loading point (Figs. 8(b) and9(b)). This reveals that the strain of the slab in the longitudinaldirection was mainly influenced by the global beam momentsexcept for the loading points locally, thus, the strain of the slabsbecame compressive at locations away from the loading point,away because the loaded section is the positive moment section(Figs. 8(b) and 9(b)). The longitudinal strain at the loadingpoint was about 70% of the transverse strain before cracking(110 kN) and about 80% after cracking of the bottom slabs atthe loading point. Since the longitudinal strain of reinforcementincreased due to cracking of the bottom slabs, the longitudinalstrain became equal to the transverse strain after cracking of thebottom slabs as shown in Fig. 9(a). Accordingly, it is consideredthat the longitudinal moment was about 70% of transversemoment in elastic range and over 80% in post-cracking ofbottom slabs in this bridge. In order to generalize the two way moment ratios in bridge slabs in both elastic and inelasticranges, more analytical studies will be performed.The initial crack load could not be sized-up visually inthe deck at internal support, while cast interface occurringgenerally at the transverse deck joint could be observed. Fig. 10sketches the shape of the cracks, where O-1–O-6 represent thecrack width gauges. These gauges were used to measure crackwidth with increase in the applied load, as plotted in Fig. 11.O-1 and O-2 correspond to the crack width gauges disposed atthe joint cast interface located at the top of the deck above theinternal support. O-5 is the gauge installed in the joint located atthe bottom of the deck in the loaded section. O-3, O-4 and O-5did not measure any particular crack width at initial monotonicloading up to 360 kN, while the progress of crack width couldbe observed mainly through O-1, O-2 and O-6, as shown inFig. 11. The size of the crack widths developed in the deckswas larger at the bottom of the deck in the loaded section (O-6)than at the transverse deck joints near the internal supports(O-1, O-2).In decks 7 (L7) and 8 (L8), rebar gauges were disposedlongitudinally in the central distribution reinforcement andconcrete gauges were installed at the same places at the topof the decks so as to observe the development of strains withincrease in the applied load (Fig. 12). If a cracking does not happen in slabs, strains of concrete slabs will be similar tostrains of reinforcements. Fig. 12 plots the strains measuredat loading of 250 kN by the reinforcement gauge (L8) andconcrete gauge (CL8) of deck 8. In Fig. 12, the value ofthe reinforcement strain observed at both extremities (L8),corresponding to the measurement at proximity of the deckjoint cast interface (Fig. 12; joint surfaces), is verified to belarger than the strain developed in the surface of the deck. Thisreveals that cracks concentrate at the cast interface in the jointsection. The initial crack spacing of the loop joint precast decksat the internal supports is assumed as the distance from thecast interface at the transverse joint to the cast interface at thetransverse joint in the opposite direction. Since the initial crackspacing was wider than in general RC slabs without joints,it is estimated that the crack width will be scaled up to anextent larger than generally proposed in the provisions [7]. Theexperimental curve plotted in Fig. 13 could be drawn fromthe relationship between the crack width (O-1 or O-2) andreinforcement strain measured in the joint cast interface. Also,the curves shown in Fig. 13 have been drawn using the formulaeproposed in foreign specifications [1–4] in order to control theexperimental results and crack widths. As expected, the crackwidth (O-1 or O-2) experimentally observed appears to havebeen scaled up.The strain on the reinforcement of the loop overlappingjoints has been measured by sticking reinforcement gaugesoriented longitudinally (Fig. 14; J7&8). Fig. 14 compares thevalues of the strain of the loop joint reinforcement (J7&8) withthe ones of the reinforcement located in the deck joint interface(L81). Comparison revealed that the reinforcement strain (L81)near the joint interface is significantly larger than that (J7&8) inthe overlapping section. This is due to the doubling of the steelratio in the loop joint section where reinforcement overlaps,while concrete crack concentrates in the joint cast interfacewithout overlapping of reinforcement. Accordingly, the sectionto which particular attention shall be paid for the control ofcrack width should be the joint cast interface.Through the experiments, moment–curvature relationship ofthe composite bridge was evaluated as shown in Fig. 15. Inthis figure, moment was calculated assuming uncracked sectionand curvature was recorded using girder strains measured inthe test.Moment curvature relationship of the composite bridgecan be evaluated according to Eurocode 4-2 [4] and comparedwith the test results. In spite of cracking of the bottom slabs atloading points, the stiffness of the maximum positive momentsection was nearly that of uncracked section. Also, the stiffnessof the maximum negative moment section was also nearly thatof uncracked section. It is considered that initial enlargementof crack width at transverse joints (O-1 and O-2) slightlyinfluenced the stiffness of the composite section. Up to theinitial static load, 360 kN before design cracking load estimatedby Eurocode 4-2, insignificant cracking was observed exceptfor enlargement of transverse joints of slabs on the interiorsupport. 3.2. Cyclic service loadingAfter verification of the elastic flexural behaviour and crackbehaviour, fatigue test has been performed by repeated loadingof 360 kN up to one million cycles. Fig. 16 plots the measureddeflection with the increase in the repeated loading cycles.The number of cycles is expressed in logarithmic scale and itcan be seen that the deflection increases gradually. Althoughfatigue load being of a level exceeding the service load, thatis 1.5 times of design rear wheel load [9], the results can beinterpreted as reflecting likely damage process that may occurin bridges due to repeated loading considering the frequencyof overloaded vehicles crossing domestic bridges during theservice life. It seems thus necessary to take into account thecumulative increases of deflection and crack width relative torepeated loads exceeding crack load, and to ensure that thesevalues do not exceed the design allowable values during thedesign of the bridge.Fig. 17 describes the propagation of crack width with thenumber of loading cycles. 预制板的连续组合双肋桥的裂缝控制英文文献和中文翻译(3):http://www.751com.cn/fanyi/lunwen_67647.html