• Steel and Composite Structures
• /
• v.9 no.5
• /
• pp.457-471
• /
• 2009

This study investigates the role of accidental torsion in seismic reliability assessment. The analyzed structures are regular 6-story and 20-story steel office buildings. The eccentricity in a floor plan was simulated by shifting the mass from the centroid by 5% of the dimension normal to earthquake shaking. The eccentricity along building heights was replicated by Latin hypercube sampling. The fragilities for immediate occupancy and life safety were evaluated using 0.7% and 2.5% inter-story drift limits. Two limit-state probabilities and the corresponding earthquake intensities were compared. The effect of ignoring accidental torsion and the use of code accidental eccentricity were also assessed. The results show that accidental torsion may influence differently the structural reliability and limit-state PGAs. In terms of structural reliability, significant differences in the probability of failure are obtained depending on whether accidental torsion is considered or not. In terms of limit-state PGAs, accidental torsion does not have a significant effect. In detail, ignoring accidental torsion leads to underestimates in low-rise buildings and at small drift limits. On the other hand, the use of code accidental eccentricity gives conservative estimates, especially in high-rise buildings at small drift limits.

• Structural Engineering and Mechanics
• /
• v.83 no.1
• /
• pp.45-51
• /
• 2022

The need to account for accidental torsion in seismic design is no longer debatable, however, the seismic codes' requirement for accidental eccentricity has recently faced criticism. In order to get as close to real conditions as possible, this study investigated the impact of accidental torsion in symmetric RC multistory buildings caused by one of its many sources, the torsional earthquake component, and compared the results to those obtained by using the accidental eccentricity recommended by the codes (shifting the center of mass). To cover a wide range of frequencies and site conditions, two types of torsion seismic components were used: a recorded torsion accelerogram and five others generated using translation accelerograms. The main parameters that govern seismic responses, such as the number of stories (to account for the influence of all modes of vibration) and the frequency ratio (Ω) variation, were studied in terms of inter-story drift and displacement responses, as well as torsional moment. The results show that the eccentricity ratio of 5% required by most codes for accidental torsion should be reexamined and that it is prudent for computer analysis to use the static moment approach to implement the accidental eccentricity while waiting for new seismic code recommendations on the subject.

• Earthquakes and Structures
• /
• v.8 no.2
• /
• pp.443-462
• /
• 2015

Some model building codes stipulate that the design displacement of a building can be computed using the elastic static analysis results multiplied by the deflection amplification factor, $C_d$. This approach for estimating the design displacement is essential and appealing in structural engineering practice when nonlinear response history analysis (NRHA) is not required. Furthermore, building codes stipulate the consideration of accidental torsion effects using accidental eccentricity, whether the buildings are symmetric-plan, or asymmetric-plan. In some model building codes, the accidental eccentricity is further amplified by the torsional amplification factor $A_x$ in order to minimize the discrepancy between statically and dynamically estimated responses. Therefore, this warrants exploration of the reliability of statically estimated design displacements in accordance with the building code requirements. This study uses the discrepancy curves as a way of assessing the reliability of the design displacement estimates resulting from the factors $C_d$ and $A_x$. The discrepancy curves show the exceedance probabilities of the differences between the statically estimated design displacements and NRHA results. The discrepancy curves of 3-story, 9-story, and 20-story example buildings are investigated in this study. The example buildings are steel special moment frames with frequency ratios equal to 0.7, 1.0, 1.3, and 1.6, as well as existing eccentricity ratios ranging from 0% to 30%.

• Wind and Structures
• /
• v.6 no.1
• /
• pp.23-40
• /
• 2003

The steady state response of a torsionally coupled system with tuned mass dampers (TMDs) to external wind-induced harmonic excitation is presented. The torsionally coupled system is considered as one-way eccentric system. The eccentricity considered in the system is accidental eccentricity only. The performance of single tuned mass damper (TMD) optimally designed without considering the torsion is investigated for the torsionally coupled system and found that the effectiveness of a single TMD is significantly reduced due to torsion in the system. However, the design of TMD system without considering the torsion is only justified for torsionally stiff systems. Further, the optimum parameters of a single TMD considering the accidental eccentricity are obtained using numerical searching technique for different values of uncoupled torsional to lateral frequency ratio and aspect ratio of the system. The optimally designed single TMD system is found to be less effective for torsionally coupled system in comparison to uncoupled system. This is due to the fact that a torsionally coupled system has two natural frequencies of vibration, as a result, at least two TMDs are required which can control both lateral and torsional response of the system. The optimum damper parameters of different alternate arrangements such as (i) two identical TMDs placed at opposite corners, (ii) two independent TMDs and (iii) four TMDs are evaluated for minimum response of the system. The comparative performance of the above TMDs arrangements is also studied for both torsionally coupled and uncoupled systems. It is found that four TMDs arrangement is quite effective solution for vibration control of torsionally coupled system.

• Structural Engineering and Mechanics
• /
• v.41 no.4
• /
• pp.527-538
• /
• 2012

In seismic analyses of structures, additional eccentricity is introduced to take account for oscillations of random and unknown origins. In many codes of practice, the torsion about the vertical axis is considered through empirical accidental eccentricity formulation. Due to the random nature of structural systems, it is very difficult to evaluate the accidental eccentricity in a deterministic way and to specify its effect on the overall seismic response of structures. The aim of this study is to develop a procedure for the evaluation of the accidental eccentricity induced by uncertainties in stiffness and mass of structural members, using the neural network techniques coupled with Monte Carlo simulations. This method gives very interesting results for single story structures. For real structures, this method can be used as a tool to determine the accidental eccentricity in the seismic vulnerability studies of buildings.

• Structural Engineering and Mechanics
• /
• v.67 no.6
• /
• pp.671-682
• /
• 2018

The effect of earthquake induced torsion, due to mass eccentricities, is investigated with the objective of providing practical design guidelines for minimizing the torsional response of building structures. Current code provisions recommend performing three dimensional static or dynamic analyses, which involve shifting the centers of the floor masses from their nominal positions to what is called an accidental eccentricity. This procedure however may significantly increase the design cost of multistory buildings, due to the numerous possible spatial combinations of mass eccentricities and it is doubtful whether such a cost would be justifiable. This paper addresses this issue on a theoretical basis and investigates the torsional response of asymmetric multistory buildings in relation to their behavior when all floor masses lie on the same vertical line. This approach provides an insight on the overall seismic response of buildings and reveals how the torsional response of a structure is influenced by an arbitrary spatial combination of mass eccentricities. It also provides practical guidelines of how a structural configuration may be designed to sustain minor torsion, which is the main objective of any practicing engineer. A parametric study is presented on 9-story common building types having a mixed-type lateral load resisting system (frames, walls, coupled wall bents) and representative heightwise variations of accidental eccentricities.

• Earthquakes and Structures
• /
• v.8 no.2
• /
• pp.463-484
• /
• 2015

Modern codes for earthquake resistant building design require consideration of the so-called accidental design eccentricity, to account for torsional response caused by several factors not explicitly considered in design. This provision requires that the mass centres in the building floor be moved a certain percentage of the building's dimension (usually 5%) along both the x and y axes and in both positive and negative directions. If one considers also the spatial combinations of the two component motion in a dynamic analysis of the building, the number of required analyses and combinations increases substantially, causing a corresponding work load increase for practicing structural engineers. Another shortcoming of this code provision is that its introduction has been based primarily on elastic results from investigations of oversimplified, hence questionable, one story building models. This problem is addressed in the present paper using four groups of eccentric braced steel buildings, designed in accordance with Eurocodes 3 (steel) and 8 (earthquake design), with and without accidental eccentricities considered. The results indicate that although accidental design eccentricities can lead to somewhat reduced inelastic response demands, the benefit is not significant from a practical point of view. This leads to suggestions that accidental design eccentricities should probably be abolished or perhaps replaced by a simpler and more effective design provision, at least for torsionally stiff buildings that constitute the vast majority of buildings encountered in practice.

• Earthquakes and Structures
• /
• v.19 no.1
• /
• pp.59-77
• /
• 2020

The code static eccentricity model for elastic torsional design of structures has two critical shortcomings: (1) the negation of the inertial torsional moment at the center of mass (CM), particularly for torsionally-unbalanced (TU) building structures, and (2) the confusion caused by the discrepancy in the definition of the design eccentricity in codes and the resistance eccentricity commonly used by engineers such as in FEMA454. To overcome these shortcomings, using the resistance eccentricity model that can accommodate the inertial torsional moment at the CM, interactive relations between shear and torsion are proposed as follows: (1) elastic responses of structures at instants of peak edge-frame drifts are given as functions of resistance eccentricity, and (2) elastic hysteretic relationships between shear and torsion in forces and deformations are bounded by ellipsoids constructed using two adjacent dominant modes. Comparison of demands estimated using these two interactive relations with those from shake-table tests of two TU building structures (a 1:5-scale five-story reinforced concrete (RC) building model and a 1:12-scale 17-story RC building model) under the service level earthquake (SLE) show that these relations match experimental results of models reasonably well. Concepts proposed in this study enable engineers to not only visualize the overall picture of torsional behavior including the relationship between shear and torsion with the range of forces and deformations, but also pinpoint easily the information about critical responses of structures such as the maximum edge-frame drifts and the corresponding shear force and torsion moment with the eccentricity.

• Structural Engineering and Mechanics
• /
• v.75 no.2
• /
• pp.145-155
• /
• 2020

This paper investigates the influence of spatial varations of accidental mass eccentricities on the torsional response of inelastic multistorey reinforced concrete buildings. It complements recent studies on the elastic response of structural buildings and extends the investigation into the inelastic range, with the aim of providing guidelines for minimising the torsional response of structural buildings. Four spatial mass eccentricity configurations of common nine story buildings, along with their reversed mass eccentricities subjected to the Erzincan-1992 and Kobe-1995 ground motions were investigated, and the results are discussed in the context of the structural response of the no eccentricity models. It is demonstrated that when the initial linear response is practically translational, it is maintained into the inelastic phase of deformation as long as the strength assignment of the lateral resisting bents is based on a planar static analysis where the applied lateral loads simulate the first mode of vibration of the uncoupled structure.

• Earthquakes and Structures
• /
• v.13 no.1
• /
• pp.9-15
• /
• 2017

In this paper, the influence of the rotational component, about a vertical axis, of earthquake ground motion on the response of building structures subjected to seismic action is considered. The torsional component of ground motion is generated from the records of translational components. Torsional component of ground motion is then, together with translational components, applied in numerical linear dynamic analysis of different reinforced concrete framed structure of three stories buildings. In total, more than 40 numerical models were created and analyzed. The obtained results show clearly the dependence of the effects of the torsional seismic component on structural system and soil properties. Thus, the current approach in seismic codes of accounting for the effects of accidental torsion due to the torsional ground motion, by shifting the center of mass, should be reevaluated.