Non-rectangular RC walls: A review of experimental investigations

When compared to tests on reinforced concrete (RC) walls with a rectangular or barbelled cross-section, only very few tests on RC walls with open cross-section exist. Most of these walls were subjected to unidirectional or bidirectional loading along one or both of the principal axes of the wall section. It was rare that tests were done using load paths that did not follow the principal axes. This article presents a review of experimental tests of non-rectangular RC walls with open cross-section. The summary comprises tests on individual non-rectangular walls with different cross-sections such as for example L-shaped or U-shaped walls which were tested under quasi-static or dynamic loads. The tests are described with their prototype structure, test objectives, investigated parameters, loading protocols and test conclusions. Emphasis is placed on observations that are specific to non-rectangular walls, which include out-of-plane buckling of the free end of wall segments and significant deviations from the hypothesis that plane sections remain plane. The importance of bidirectional loading for nonrectangular walls is stressed because the critical loading direction for design may be different from loading in the principal directions of the section and because loading along one direction reduces the stiffness of the wall for subsequent loading in the direction orthogonal to the previous one.

Seismic behaviour and analysis of U-shaped RC walls

Raluca-Tereza Constantin

In many countries of moderate to high seismicity, reinforced concrete (RC) core walls are used as lateral bracing systems in mid- to high-rise buildings where they typically accommodate lift shafts or stair cases. Unlike planar walls which provide horizontal strength and stiffness mainly in the in-plane direction of the wall, core walls provide bidirectional strength and stiffness. This feature complicates their inelastic behaviour which is not yet fully understood as experimental and numerical studies are still rather scarce when compared to planar walls. Therefore, current design codes are based on findings from planar walls while specific guidelines for the design and analysis of core walls are still lacking. In addition, simple analysis tools widely used by design engineers such as the plastic hinge model, have been derived and calibrated for columns, beams or planar walls and their suitability for core walls has been only marginally verified. For these reasons the current study focuses on: (1) improving the knowledge on the inelastic behaviour of U-shaped walls under bidirectional loading and (2) on extending easily applicable engineering type models, such as the plastic hinge model, to the analysis of U-shaped walls. The scope of the thesis is limited to U-shaped walls, which is the simplest type of core wall that still retains the key characteristics of such walls. The first part of the thesis focuses on understanding the behaviour of U-shaped walls under loading along the geometric diagonal of the section. This loading direction is not typically considered in the design process but it was found to determine the shear design of the flanges while the displacement capacity for this direction might be the lowest of all the loading directions. In order to address the behaviour under diagonal loading, two large-scale quasi-static cyclic tests on U-shaped walls were carried out. Failure mechanisms specific to diagonal loading and possible critical design aspects related to these failure modes were identified from the experimental results. In addition, the influence of the longitudinal reinforcement distribution on the wall behaviour was investigated. The second part of the thesis focuses on adapting equations used in plastic hinge model for the analysis of U-shaped walls. Plane section analyses and a shell element model validated against the experimental data were used to perform parametric studies on U-shaped walls. From the results of the parametric studies a new equation for the yield curvature for any direction of horizontal loading was proposed as well as a modified yield displacement equation that accounts for the partially cracked wall height at yield. With this equation, predictions of yield displacements were improved especially for very slender walls. Quantities that rely on the yield displacement, such as the effective stiffness of the wall were also better predicted when the cracked height was accounted for in the prediction. And finally, plastic hinge length equations were also modified to account for the variation with the different loading directions.

EPFL2016

Decay of Torsional Stiffness in RC U-Shaped Walls

Katrin Beyer, Ryan Hoult

Reinforced concrete (RC) U-shaped walls are a popular construction choice, commonly used to resist the lateral loads from wind and earthquakes. In many buildings, the center of stiffness of a floor is eccentric from the center of mass and the building will therefore undergo some twist. Often, U-shaped walls contribute significantly to the torsional stiffness of the building and the analysis of the structure therefore requires an estimate of the torsional stiffness. This research investigates the dependency of the torsional stiffness of U-shaped walls on the translational displacement demands using experimental evidence and advanced numerical models. The experimental results show that the torsional stiffness decreases with increasing translational displacement. Furthermore, the numerical and experimental results indicate that the torsional stiffness degrades at a similar rate as the translational stiffness of the walls. The rotational stiffness of the wall is smaller when it is centered than when it is subjected to a translation. This effect is attributed to the longitudinal stresses throughout the cross section that are already present due to the imposed flexural displacement. In particular, the compression zone resulting from the translational displacement imposed on the wall stiffens the wall with regard to the twisting. As a result, the effective torsional stiffness increases with increasing axial load ratio. Current codes do not provide reduction factors for the torsional stiffness, but the results show that the reduction factors for the shear stiffness can serve as estimate, albeit they do not capture the dependence on the axial load ratio. (C) 2020 American Society of Civil Engineers.

ASCE-AMER SOC CIVIL ENGINEERS2020

Seismic behaviour and analysis of U-shaped RC walls

Raluca-Tereza Constantin

EPFL2016