Search

The curvature causes a variation in the deflection of the outer and inner sides. The effect of curvature was investigated by casting and testing two specimens with the same section – one straight and the other horizontally curved continuous box girder. ABAQUS software was used to numerically model the box girder in order to verify the model and investigate additional parameters. Numerical modeling is successful with less effort, cost, and time because good results are obtained. The effect of the span-to-depth ratio, the compressive strength of concrete, and the percentage of stirrup steel reinforcement was studied numerically. Increasing the height, compressive strength, and percentage of stirrup steel led to a significant increase in load capacity and stiffness. The load capacity in the curved specimen decreased by 11% compared to the straight one due to the effect of torsional moments. A mathematical model was proposed based on the theory of strut-and-tie modeling (STM), where the span was divided into several panels, the effect of torsion was added, and then the results were compared with the traditional sectional method according to ACI and CEB-FIB. For the straight specimen, the sectional ACI, CEB-FIB, and STM methods were used, which gave theoretical results less than the experimental by 31, 48 and 13%, respectively. For the curved specimen, to get closer to reality, the sectional and STM methods were modified by adding the effect of torsion, and the results were less than the experimental tests by 43, 61 and 22%, respectively.

Reducing the ratio of box girders' span to their height reduces bending moments and makes them deep members that are controlled by shear behavior. The presence of horizontal curvature generates torsional moments and different deflection between the outer and inner web as a result of twisting. Two experimental specimens with the same dimensions were cast and tested, one straight and the other horizontally curved. The two specimens were numerically modeled using the ABAQUS software for the purpose of verifying the numerical model to study more parameters. The finite element analysis showed a good agreement with the experimental values with a difference of (98-99) %, (94-97)% and 103% of the experiment for ultimate loading, deflection and twisting angle, respectively. It also showed stress paths that match the experimental and theoretical results that explain the behavior of deep beams, such as the Strut and Tie method (STM). The effect of compressive strength, whole width, and the width of the bearing and supporting plates was studied. Results showed that increasing the concrete's compressive strength by about 40-120%, increased the load capacity and decreased its deflection by approximately 6-14% and 3-15%, respectively. The deep box section torsional resistance increased when its width was increased by approximately 17–50%, although this had no significantly affect on the load capacity level. A reduction of 17-33% in the width of the loading and supporting bearing plates caused a 6-16% drop in load capacity besides a notable decrease in stiffness.

Shear lag effect of curved box girder under multi-dimensional seismic excitation is studied in this paper. Firstly, spatial finite element model is established based on ANSYS, and a seismic wave, which is recorded in second site, is chosen as ground acceleration time history. Secondly, elastic dynamic time-history analysis focused on shear lag effect is carried out, where 4 working conditions, 3-D seismic, longitudinal-vertical seismic, vertical seismic and transverse seismic only, are considered. Thirdly, critical angle of seismic waves is investigated, it is seen that under seismic excitation, there is a prominent shear lag effect on upper flange at mid-span of the curved box girder, and there are also various shear lag effect modes under the different working conditions of seismic excitation. The shear lag under 3-D seismic is severest, normal stress is concentrated on inside upper flange, then that under longitudinal-vertical seismic is less serious, in which case, the stress is appearing within a regional proximity to the junction between webs and flange, the next is under vertical seismic, and the shear lag effect under transverse seismic is most non-prominent. Finally, the numeric results are compared with the experimental results from a vibration table testing, which shows great consistencies. [ABSTRACT FROM AUTHOR]