Once thegirders are positioned, the decks are lifted and installed on thegirders. At that time, in the case of the prototype, the installationof the decks was performed on the girders equipped in advancewith the shear connectors. In such case, the assembling maybe delicate because of the lack of space due to the presence ofthe shear connectors and loop joints. As a measure to solve thisproblem, the shear connectors can be welded to the girder atthe shear pocket locations using a stud welder after installationof the decks on the girder and with the shear connectors notwelded to the girders. However, such methods require sufficientspace for the shear pockets, and the process may also be delicatewhen space cannot be secured. Especially, the enlargement ofthe shear pockets may generate adverse results in the deckbecause of the difficulty to arrange the reinforcing bars, whichmakes it difficult to enlarge indefinitely the space of the shearpockets. As another measure, shear connectors in the form ofscrew-bolt can be considered.On the other hand, once the installation of the decks iscompleted, the transverse reinforcement shall be arranged inthe overlapping sections of transverse loop joints (Fig. 2(a)and (b)). As illustrated in Fig. 2(b), six main reinforcementshave been arranged at the top and bottom of the overlappingsections of transverse loop joints for the prototype. After thearrangement of bars, concrete is cast in the shear pocketsand transverse joints, and curing is performed. As shown inFig. 2(a), the edge of the shear pockets has been rounded tominimize the effects of stress concentration. The fabrication ofthe two-girder continuous composite bridge is completed withthe end of curing (Fig. 2(c)).2.3. Loading and measurement locationsFor all the tests, loading was applied at both mid-spans(Fig. 3(a)) considering the contact surface of the wheel load(Fig. 3(b)). A location of loading with respect to the nearestjoints was about 600 mm. The first loading was applied up to360 kN and then fatigue loading test proceeded. Lastly, staticloading was applied up to 900 kN. The fatigue load of 1.5 Hzcycles was applied using a dynamic actuator with capacity of500 kN. Minimum and maximum load of the test were 3.9 kNand 360 kN, respectively.The deflection of the test bridge has been measuredby means of LVDT disposed at both loaded mid-spans asillustrated in Fig. 3(c). In the sectional direction, deflectionswere measured both in the bottom of the girder and in thedecks. After the development of cracks has been verified usingOmega gauges, the crack width was measured at the bottom ofthe decks in the loaded sections and at the top of the decks inthe internal support sections. Considering that the distributionof relative slip between the girder and the deck occurs anti-symmetrically with respect to the center of the internal support,measurements were performed only for the span of one girder Fig. 2. Fabrication of continuous composite bridge with loop joinprefabricated slabs.at locations (SL1–SL6) indicated in Fig. 3(c) [6]. The strainof the girders was measured by sticking strain gauges at midspan and at the web plate and top–bottom flanges of the internasupport girder (Fig. 3(c)). Furthermore, several strain gaugeswere installed on the top and bottom reinforcements in slabs.2.4.
Material propertiesTable 1 shows mix proportions of concrete. Material testswere performed to measure the strength of concrete, expansive concrete and non-shrinkage mortar, listed in Table 2. Theprecast deck was manufactured in advance using ordinaryconcrete and exhibited strength at 14 days as shown in Table 2.Since tests started one month after the manufacture of the deck,this strength was undoubtedly larger than measured. Expansiveconcrete was cast in the transverse joints between decks soas to minimize the effects due to initial drying shrinkage.Non-shrinkage mortar was cast as filling material in the shearpockets to combine the shear connectors and the deck. Thestrengths of expansive concrete and non-shrinkage mortar arelisted in Table 2 for the strength measured at the start of the teststhat is under loading. The plate girder applied for the specimen is SM490 steel, which is used as main member in real bridges.Steel reinforcement bars are SD40 with diameter of 16 mm andyield strength of 400 MPa (Table 3).3. Test results and analysis3.1. Monotonic service loadingFirstly, monotonic static load was applied up to 360 kN.This load was 1.5 times the design rear wheel load includingdynamic effects [9].Maximum slips were measured at 0.07 mm under the serviceloading (Fig. 4). It is considered that the experimental slipsmonitored during the tests were scattered due to the very lowvalues measured. From this result, it is considered that the testspecimen could be assumed as the full composite section underthe service loading.During the initial static loading up to 250 kN, cracks didnot develop in the decks and the prototype exhibited elasticbehaviour as an uncracked section. The software, MIDAS [11]was utilized for the analysis of the numerical model of theprototype. The elements applied in the model were 4-nodeshell elements for the concrete decks as well as for the girders. The connection of the concrete deck element with the upperflange element has been simulated using beam elements. Fig. 5illustrates the 3D model. Finite element analysis has beenperformed in the elastic range. In this model, prestressingtendon and loop joints were not modeled. Fig. 6 comparesthe deflections. The measured deflections correspond to theones recorded at the loaded sections (Fig. 3(c)), which are thedeflections of both girders (S4, S6) and the deflection at thebottom of the deck (S5). Since the load was applied at the centerof the transverse direction, the deflections of S4 and S6 shouldbe theoretically identical. However, it appeared that S4 waslarger than S6 by about 5%. This can be explained by the slighteccentricity of loading toward the girder where the deflection ofS4 was measured or by small difference in the stiffness of thegirders. Following this, it can be said that the behaviour in theelastic range is in good accuracy with the experimental results.Also, measurement of the deflection of S5 was larger than S4and S6, of which relative deflection obtained by the subtractionof the deflection S4 or S6 from S5 can be seen as the localdeflection of the deck.The initial crack load calculated with reference to the deckexhibiting maximum negative moment has been estimated tobe 450 kN. Since the precast deck could be efficiently used inlong span between girders particularly in two- or three-girderbridges, transverse pre-tension is usually needed. In this testbridge model, transverse pre-tension has also been introducedin the transverse direction to the deck, pre-tension has beendesigned so as to control the bending moment produced bythe dead load during the transport and the design bendingmoment due to design rear wheel load including the impact coefficient [9]. Following this, longitudinal crack at the bottomof the deck in the loaded sections is predicted to develop aftercracking in the maximum negative moment deck. However,in reality, transverse cracks began to develop near an appliedload of 250 kN in the deck at the loaded section. This revealsthat cracking of bottom slabs by the longitudinal bendingmoment should also be considered in addition to the transversebending moment considered in the design. 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(2):http://www.751com.cn/fanyi/lunwen_67647.html