The response spectrum (RS) analysis technique is a valuable tool for finite element seismic analysis; however, its application to structures and foundations requires caution. The method's inherent technique of disregarding the sign of resultants can lead to oversimplifications and potentially inaccurate outcomes. This is especially critical when dealing with the unique characteristics of structural systems, including shear walls, floor slabs, mat foundations, and design mechanism including buckling, failure modes, and anchorage design during seismic events. In this session, practitioners and researchers present different limitations of the Response Spectrum (RS) analysis and provide practical resolutions and/or work around tips.
Learning Objectives:
(1) Explain different classes and the nature of RSA simplifications' limitations;
(2) List RSA limitations on analysis, design, combinations, failure mechanisms, and stability assessments;
(3) Discuss practical solutions for FE approximation of RSA techniques;
(4) Review alternative examples and approaches in FE dynamic analysis for practitioners.
This session has been approved by AIA and ICC for 2 PDHs (0.2 CEUs). Please note: You must attend the live session for the entire duration to receive credit. On-demand sessions do not qualify for PDH/CEU credit.
Design of Nonplanar Reinforced Concrete Ductile Coupled Walls Using Modal Analysis
Presented By: Jeff Dragovich
Affiliation: Desimone Consulting Engineering
Description: Challenges in the design of non-planar reinforced concrete walls based on linear dynamic analysis using the modal response spectrum procedure are presented. For coupled wall systems, the general case of biaxial bending with varying axial loads requires consideration of multiple load cases prescribed by the governing building code. The load cases relying on response spectrum modal analysis, lead to biaxial direction combinations inducing biaxial bending about neutral axes that are not aligned with the wall cross-section principal axes. Because traditional SRSS or CQC modal combinations result in design forces with an absolute magnitude only, the kinematics of wall cross section deformation are lost, resulting in potentially overly conservative wall designs with the signs of the applied loads at the discretion of the designer.
This paper reviews practical approaches to address multicomponent design demands, for the code-based orthogonal combinations of loading, using response spectrum analysis. A simplified approach for development of signed multicomponent design demands is presented, and procedures for determining required boundary zone extents for the general case of biaxial bending and axial loads.
Limitation of Modal Response Spectrum Analysis for Two-Step Seismic Analysis of Structures
Presented By: Mukti Das
Affiliation: Das Consulting
Description: For flexible structural systems supported by relatively rigid structural systems, two-step method of seismic analysis using decoupled models is typically considered acceptable. ASCE 7 and ASCE 4 allow the use of two-step analysis procedure for equivalent static and dynamic analysis methods, respectively, when certain conditions related to stiffness and mass distribution for the two systems are met. For an accurate representation of the overall dynamic behavior using the decoupled analysis, it is important that the forces transferred from the supported structure to the supporting structure are accurately captured. When the seismic analysis of the decoupled supported system is conducted using the modal response spectrum analysis (MRSA) where modal combination is performed using square root of sum of squares (SRSS) or complete quadratic combination (CQC) methods, although the peak responses of interest are captured at the discrete support locations, the signs are lost. So, the application of these seismic analysis reactions from the decoupled supported structure on the supporting structure model can result in inaccurate seismic analysis results for the design of the supporting structure and global evaluations. This paper examines this limitation of the MRSA using an example structure with a steel structure supported on a concrete shear-wall structure and provides resolution strategies based on comparisons from the seismic analyses of decoupled and coupled analytical models.
Inadequacies of the Approximate Response Spectrum Technique and Practical Solutions
Presented By: Abbas Mokhtar-zadeh
Affiliation: M3 Engineering & Technology Corporation
Description: Response spectrum dynamic analysis technique has been developed since more than eighty years ago with the aim of improving the accuracy of earthquake loads. This technique approximates linear method for the seismic loading and to provide simple guidelines for its practical uses in modal considerations. Today, due to the advanced technological procedures and tools we can have more accurate, flexible and simple time-history earthquake response analysis approaches for both linear and nonlinear analysis of simple or complex structures. In our contemporary era, it appears that the use of more accurate and precise analysis methods has not occurred. In fact, the use of the approximate response spectrum methods and other simplified static methods seems to have been expanded and the use of the more accurate time-history analysis methods has been reduced. This article outlines basic historical developments and simple approximations of the response spectrum versus actual engineering principles and demands. Then, it illustrates a list of important limitations of this approximate response spectrum techniques and tool in structural analysis and design for the strength and performance assessments and stability considerations. Finally, simple examples will be given to highlight the nature of the inaccuracies and provided some practical resolutions in finite element analysis of structures and foundations.
Alternative Framework for Modeling Damping in Nonlinear Analysis
Presented By: Xianjue Deng
Affiliation: New Mexico State University
Description: Nonlinear analysis using inelastic material laws inherently accounts for the relevant energy dissipation. Still, there is a need to model damping in the system, assumed to correspond to sources of energy dissipation not inherently captured by the model, due to, e.g., small-strain material hysteresis or even dissipation from nonstructural components. Damping in nonlinear dynamic analysis for seismic loads is typically modeled as viscous, i.e. a force proportional to velocity (or a stress proportional to strain rate). While effective and advantageous for linear systems, the use of viscous damping to nonlinear structural response introduces significant challenges, including frequency-dependence of damping ratios, artificial energy dissipation, and limited correlation with physical reality.
This presentation suggests the adoption of a hysteretic damping approach to model dissipation mechanisms not directly described by nonlinear structural models. The approach is implemented by adding an extra, hysteretic component to the actual constitutive law of each material point in the model. The hysteresis of the additional component is described through an elastoplastic, Armstrong-Frederick hardening model with vanishing yield strength, yielding an algorithm that is applicable to both uniaxial and multiaxial material models alike. Unlike Rayleigh or classical viscous damping, this formulation offers physically meaningful parameters that can be calibrated against experimental data, such as small-strain hysteresis, friction at connections, and impact of nonstructural components.