Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/5326
Title: Study of torsional behavior of reinforced concrete walls
Authors: Peng, Xiaoning
Subjects: Concrete walls -- Testing.
Reinforced concrete.
Hong Kong Polytechnic University -- Dissertations
Issue Date: 2012
Publisher: The Hong Kong Polytechnic University
Abstract: As earthquake resistant design is not a requirement in Hong Kong construction, structural systems with high torsional eccentricity are not uncommon in Hong Kong buildings. The previous investigations on the seismic response of asymmetric-plan buildings are carried out based on the assumption that torsional resistance mainly comes from the base shear distributed among the lateral force resisting elements, while the torsional resistance of each individual element is neglected. Consequently, the interactions of torsion with flexure, shear and axial loads are not taken into account in the dynamic analysis. This PhD study conducted a dynamic analysis on a one-story bi-walls asymmetric system to investigate the effect of the torsional resistance of wall, with or without considering the flexure-torsion interaction, on the response of the asymmetric system. Aiming to simulate the torsional behavior of wall in the dynamic analysis, an experimental investigation was conducted on the behavior of eight half-scaled reinforced concrete (RC) cantilevered walls subjected to monotonic pure torsion. The test results including the failure mechanism, experimental torque-twist curve, torsional strength and torsional stiffness were discussed. Test results indicated that the whole cross sections of the test wall units, with the largest aspect ratio of 8, could be fully mobilized to resist the applied torque. In addition, various existing torsional models and the ACI Code were evaluated by comparing the predicted results with the experimental ones. Finally, the torsional skeleton curve of the wall is determined based on the softened membrane model for torsion (SMMT) and further adjusted by the conclusions drawn from the pure torsional tests. In order to incorporate the flexure-torsion interaction in the dynamic analysis, a combined loading test was conducted on three cantilevered RC walls. A series of simplified interaction curves was developed based on the test results. Furthermore, the effects of the torque to bending moment (T/M) ratio on the failure mechanisms, crack patterns, strength and stiffness were discussed. It was found that the flexural strength was reduced by the increasing torque. On the other hand, the wall under relatively large T/M ratio was capable of sustaining almost equal torsional strength as the wall under pure torsion. For the walls under combined loadings, the degradations of both pre-cracking torsional stiffness and lateral stiffness were observed. Besides, the test results were compared with the predictions by ACI provisions by evaluating the interaction diagrams between bending moment and torsion. It was found that the ACI provisions gave very conservative predictions especially for walls under pure torsion or dominant torsion.
Finally, the dynamic analysis was conducted to investigate the seismic response of asymmetric structural systems by taking into account the torsional resistance of wall element and its interaction with flexure. A numerical, step-by-step integration procedure (average acceleration method) was used to solve the equation of motion. The element forces (torsion and flexure) obtained at the previous time step were used to determine the flexural and torsional hysteretic relations of the current time step through interaction curves. Three real earthquake waves were selected and scaled to various PGA levels ranged between 0.1g and 1g with an interval of 0.1g. Three series of models with varying levels of plan-asymmetry, respectively characterized by large (erx/A=0.47), medium (erx/A=0.21) and small (erx/A=0.06) eccentricities, were adopted in the analysis. The analytical results showed that the torsional resistance of the wall element could be substantially mobilized by large rotational displacement under seismic excitation. Simultaneously, the flexural resistance of the wall could also be utilized. As a result, the torsional resistance of the wall and its interaction with flexure would have an influence on the seismic response of asymmetric buildings. It was found that the exclusion of the torsional resistance of the wall does not always lead to a conservative estimation of the system response. First, the flexure-torsion interaction effect did not play a significant role in the peak lateral displacement at CM, the peak slab rotation and the peak displacement of flexible wall. These parameters were underestimated to varying degrees under low-to-medium PGA levels, which ranged from 0.2g to 0.6g depending on the characteristics of the ground motion, if either the wall torsion or the interaction effect was ignored. On the other hand, the displacement of stiff wall was affected by the flexure-torsion interaction in a significant way. The displacement tended to be underestimated under most of the selected levels of ground motion if the interaction effect was neglected. In addition, as expected, the effect of wall torsion on the system response was more significant for the models with greater plan-asymmetry. However, even for the models with less plan-asymmetry, the system response could be underestimated by as much as 20% if the wall torsion was ignored.
Description: xxv, 189 leaves : ill. ; 30 cm.
PolyU Library Call No.: [THS] LG51 .H577P CSE 2012 Peng
Rights: All rights reserved.
Type: Thesis
URI: http://hdl.handle.net/10397/5326
Appears in Collections:CEE Theses
PolyU Electronic Theses

Files in This Item:
File Description SizeFormat 
b2507345x_link.htmFor PolyU Users162 BHTMLView/Open
b2507345x_ir.pdfFor All Users (Non-printable) 8.36 MBAdobe PDFView/Open


All items in the PolyU Institutional Repository are protected by copyright, with all rights reserved, unless otherwise indicated. No item in the PolyU IR may be reproduced for commercial or resale purposes.