Building regularity for simplified modelling

Technical Abstract

As part of structural design, members in buildings are selected and detailed such that the expected demands, such as forces or displacements, on a structure are less than the capacity of the structure to resist those forces and displacements. However, to obtain these forces or displacements, structural analysis is required considering the loading applied to the building from its weight, its use, and other factors such as wind, or shaking of the ground in the case of earthquake, which is considered in this report.

Many different analysis methods are available for earthquake. Some give a realistic understanding of the behaviour of a structure in a particular earthquake, but are too complex for design. Many simple structures in New Zealand are designed using the NZ Structural Design code Equivalent Static Procedure. This procedure was developed for structures which have relatively regular configurations. For those structures with discontinuities, significant changes in stiffness, strength and mass over the height, or structures which have irregular plans or flexible diaphragms, it is possible that the Equivalent Static Procedure may underestimate the actual demands, which will result in unsafe structures. For this reason design codes have limitations on the amount of irregularity of structures designed according to the Equivalent Static Procedure. These limitations are based on engineering judgement, rather than on quantitative analysis, and it is not clear how much the demands are likely to change with different amounts of irregularity.

This project was initiated to quantify the effect of different degrees of irregularity on structures designed for earthquake using simplified analysis. The types of irregularity considered were:

1. Vertical irregularity
   1. mass
   2. stiffness - strength
2. Horizontal (Plan) Irregularity
   1. Torsional
   2. Diaphragm

Code compliant structures expected to exhibit sensitivity to these types of analysis were selected. These were chosen so that they could represent a number of frame types. Simplified modelling was used in the analyses, so that many analyses could be conducted in a relatively short time. The structures were designed according to NZS1170.5 for regions of high, medium and low seismicity firstly as regular structures, and then they were redesigned as irregular structures. By subjecting the structures to a suite of records which represented design level shaking, the differences between the actual and predicted responses could be compared for structures with different levels of irregularity.

Relationships between the degree of irregularity and the change in behaviour could therefore be developed which allows guidance as to:
a) when the effect of structural irregularity can be ignored, and
b) the change in demands for different degrees of structural irregularity.

The key findings in the study are:

• A simple method was developed to quantify irregularity, based on the engineering demand parameter of interstorey drift, was more robust than that used in previous literature.
• Relationships were developed to describe the change in response as a function of mass irregularity.
• For stiffness and strength irregularity, relationships between stiffness and strength were developed for realistic structures. Only these realistic relationships were considered when evaluating stiffness and strength irregularity for a constant storey height. Relationships to describe the change in response as a function of stiffness-strength irregularity were developed.
• A second type of stiffness-strength irregularity can occur due to changes in interstorey height. Relationships to describe the change in response as a function of stiffness-strength irregularity were developed.
• It was shown that torsional effects resulting from earthquake shaking can be modelled well using impulsive loading. The rotational mass inertia effect is significant with systems with very little rotational restraint, but not for systems with significant torsional restraint. A relationship for the amount of torsional restraint required for a specified amount of torsional response was described which can be used in simplified analysis.
• It was shown that diaphragm flexibility is unlikely to increase the lateral forces in most structures due to the increase in period, and the decrease in spectral acceleration. However, structural displacement may increase as a result of the increase in spectral displacement. Conservative methods to assess the likely increase in displacement were developed.

While this study has developed some simple tools for structural engineers to assess how building irregularity is likely to affect the design process, its most significant contributions to decreasing structural damage during an earthquake are:

• Helping designers qualitatively understand how irregularity affects building performance so they can mitigate its effects during the preliminary design stage.
• Providing a rigorous technical basis for revision of the regularity provisions in the Structural Actions Standard, NZS 1170.5.
• Providing a basis for defining acceptable irregularity limitations for structures.

The funding provided by the Earthquake Commission for this work was used primarily to support Vinod Kota Sadashiva during his doctoral research at the University of Canterbury. Additional work was carried out by others. In particular, Eu Ving Au studied torsional effects in structures as part of his undergraduate research project in 2007. Matt Spooner undertook an undergraduate project at the University of Canterbury in 2008 related to floor flexibility. Other students and staff also contributed to this work.