Performance Based Seismic Design Of Tall Buildings: Risks And Responsibilities

Dr. Ashok K. Jain
Professor of Civil Engineering (Retd.), IIT Roorkee

After the introduction of the limit state design philosophy, there has been a serious debate on the merits of prescriptive Code versus performance-based Code. The stakeholders including the owner, designers, contractors, and the users of the facility, all want a more efficient, safer, economical and durable building. The prescriptive Codes are too restrictive and discourage innovation. The performance-based Codes are “open ended”, and the entire risk, liability and responsibility lies with the owners and the designers. Of course, this approach should be acceptable to the Authority having jurisdiction. The Indian earthquake Codes are too restrictive and sometimes arbitrary. Moreover, in the absence of any commentary, it becomes very difficult to appreciate the need of a given specification. This paper attempts to bring out salient features of the PBSD as developed by PEER, Berkeley, as well as those of the Chinese Codes.

CECR

Introduction
Building Codes can be classified into three main categories: material Codes, loading Codes and design Codes. There are numerous Codes on materials: cements of different types and grades, different types of pozzolanas, aggregates, admixtures, reinforcement steel bars, and couplers etc. There are two main loading Codes in India: IS 875 (parts 1 to 5) and IS 1893 (parts 1 to 5). Similarly, there are three basic design Codes: IS 456, IS 1343 for concrete and IS 800 for steel. Depending upon the type of structure, there are numerous specialized design Codes: IS 3370 (parts 1 to 3) on water tanks, IS 13920 on seismic detailing for buildings, and so on. In India, the process of revising a code takes decades, and even then, the content of the Codes is not necessarily up to the expectations of the Indian designers or up to the international standards. It is a matter of great concern that our Codes are being revised using copy, paste, and edit technology from various sources without keeping continuity and implications in mind. For example, IS 875-1987 on wind loads was revised in 2015; the dynamic wind loads are based on the Australian Code AS 1170.2 of 1989. The values of cross-wind force spectrum coefficient as given in Fig. 10 are very confusing and unusable. No equations are provided to compute these values in a worksheet. There is no reason as to why it could not be based on the latest AS 1170.2- 2011 that takes care of these issues.
Generic placeholder image Dr. Ashok K. Jain
Professor of Civil Engineering (Retd.)
IIT, Roorkee
IS 1893-2002 part 1 has been revised in 2016 but the design philosophy, in vogue since the first edition in 1962, has been omitted without any justification. There are generally three levels of earthquakes: minor earthquake, moderate earthquake and severe earthquake, in one form or the other, in all international codes. The design philosophy clearly used to state as to how the buildings designed using this code is expected to perform during these earthquakes. No reason has been given in the preface to the code or elsewhere as to why it has been deleted. It has a direct impact on the performance based seismic design of buildings. IS 1893-2002 part 1 had introduced the concept of the limit state design through the terms maximum considered earthquake (MCE), and design basis earthquake (DBE), and checking the drifts under serviceability conditions. The design response spectra and base shear formula were derived accordingly. Now, in 2016 edition, the terms MCE and DBE have been dropped, whereas, the design response spectra and the base shear equations have been retained. This is baffling and a very regressive step. Further, a new clause has been added, which forces the designer to design a building in any seismic zone for a minimum seismic force, in case the computed base shear is less than this minimum value. There is no clarification anywhere on how this minimum base shear was obtained, what was the need for such a clause. What is the problem, if wind governs the design? It is interesting to know that accelerograms of thousands of earthquakes in various parts of India have been recorded over the past few decades, but these are essentially minor earthquakes. There is no accelerogram of any of the strong earthquakes, such as the Bhuj earthquake of 2001and Nepal earthquake of 2015. Moreover, Bhuj 2001 is the only earthquake that has struck an urban city, Ahmedabad, about 300 km away from the epicentre over the past century. This raises a more fundamental question – is there any need to update the seismic zones based on more realistic data? Should there be a map showing the contours of peak ground accelerations at close interval, for the entire country, taking into account seismological and geo-technical features? This is also called micro-zonation. Such an approach will create more confidence in the seismic map, and will make the buildings more economical. Similarly, there are many issues with IS 13920-2016 (Jain 2017). Recently, a new prescriptive Code on tall concrete buildings (IS 16700) has been published, which is applicable for buildings of heights between 50 m and 250 m. It is biased towards seismic loads over wind loads. It limits the heights of various structural systems in different seismic zones and specifies minimum base shear in each zone (Jain 2016b). For a building whose height is in between 50 m and 250 m, it forces a site-specific design spectrum, for use in zones IV and V. If it is so, then what is the point in specifying a design response spectrum in IS 1893-part 1? What is the reliability of this response spectrum? How many agencies can carry out site-specific seismic studies in India? Do they have all the necessary data, and man-power to cater to the need of the country? What happens to the already built buildings in the range of 50 m to 150 m in zones IV and V? Here, it is pertinent to point out that ASCE 7, PEER (2017) and EC8-part-1 permit use of design response spectrum, design response spectrum compatible time-history and artificial earthquakes. The most critical clause of IS 16700 is clause 1.7. It states: “For buildings that do not conform to the prescriptive requirements of this Code, a more rigorous procedure is necessary for design and review. The general procedure to be adopted is given in Annex A to proportion, analyse, design, detail, gain approval and construct such buildings. The performance objectives or procedures more stringent than those specified in Annex A may be specified by the client/ owner of the building or by the Tall Building Committee appointed by the local authority administering the building project. It introduces, vide clause A-5.2, linear time-history analysis for moderate earthquake (DBE with 10% probability of occurrence in 50 years), and non-linear time-history analysis for rare earthquakes (MCE with 1% probability in 50 years) for buildings including code-exceeding buildings. Is it not ironic that these two terms, DBE and MCE, were deleted in IS 1893-1 in its 2016 edition, without any justification? Anyway, the essence of clause 1.7 of IS 16700 is that it permits a designer to adopt the latest Performance-Based Design for a tall building. This freedom comes with tremendous risk and responsibility. The last line of clause 1.7 is difficult to understand. It says, “Tall Building Committee appointed by the local authority administering the building project”. Does it mean such a committee will be appointed by the likes of Mumbai Metropolitan Region Development Authority (MMRDA), Delhi Development Authority (DDA), or Ghaziabad Development Authority (GDA)? Do they have the expertise or clue of the issues involved in PBSD?
Performance Based Design
A considerable research has been done on various aspects of performance-based design of buildings during the past few decades. Most advanced Codes such as ASCE 7, Eurocode 8, New Zealand Code AS/NZ 1170.5, and Chinese Code permit PBD. The refreshing part is that each one of these Codes uses an independent thinking based on their own research, experience and judgments. Let us first understand what performance-based design is and how it is different than the Codes in vogue (Jain 2016a). So far, all the Indian Codes, mentioned earlier are prescriptive Codes. Each Code tells the designer – what to do and not to do. The designer is forced to follow each clause religiously. The product may be safe or even unsafe. It may or may not perform well during its lifetime. There is no guarantee. On the other hand, the performance-based design procedure unshackles the designer – no doubt with responsibility. Building must respond to MCE shaking without excessive lateral drift or global structural instability. It may require additional analytical or experimental studies. This procedure is intended to provide a reliable basis for the seismic design of tall buildings based on the present state-of-knowledge, laboratory and analytical research, and the engineering judgment of persons with substantial knowledge in the design and seismic behaviour of tall buildings. When properly implemented, it should permit the design of tall buildings that are capable of seismic performance equivalent or superior to that attainable by design in accordance with present prescriptive building code provisions.
The Essence of PBSD
- Under the frequent earthquake, that is, Service-Level Earthquake (SLE), the building must demonstrate an elastic response, and negligible damage to non-structural components. In other words, the building should remain fully operational immediately after SLE shaking. Such performance is achievable if minor structural damage occurs that does not affect either immediate or long-term performance of the building and, therefore, does not compromise safety associated with continued building use. Repair, if required, should generally be of a nature and extent that it can be conducted while the building remains occupied and in use, though some local disruption of occupancy around the areas of repair may be necessary during repair activities.
- Under the design wind force, both static and dynamic, the building must demonstrate an elastic response, and negligible damage to non-structural components.
- Demonstrate that the building will respond to MCE shaking without loss of gravity-load-carrying capacity; without inelastic straining of important lateral force-resisting elements to a level that will severely degrade their strength; and without experiencing excessive permanent lateral drift or development of global structural instability. In other words, it may be necessary to classify certain members as force-controlled, while, other members as deformation controlled.
- Detail all elements of the structure for compatibility with the anticipated deformations of the seismic-force-resisting system under MCE shaking; and,
- Anchor and brace all non-structural components and systems, such that, they do not create falling hazards. In other words, it will require an appropriate limit on the storey drift, both under the SLE and MCE conditions.
Return Periods
- The SLE may have a return period of 43 years, that is, 50% probability of exceedance in 30 years, or any other suitable definition.
- The DBE may have a return period of 475 years, that is, 10% probability of exceedance in 50 years, or any other suitable definition.
- The MCE may have a return period ranging between 1000 years and 2500 years, that is, 10% probability of exceedance in 100 years, or any other suitable definition.
- The wind force may have a return period of 100 years to 1000 years as appropriate.
Prerequisites
- Prior to initiating a design using PBSD, ascertain that this approach will be acceptable to the Authority having jurisdiction.
- Inform the Project Developer of the risks associated with the use of alternative procedures for design.
- Ensure that the design team has the requisite knowledge and experience in subjects of ground shaking hazards, selection of structural systems for resistance to earthquake shaking and wind loads, material behaviour, nonlinear dynamic structural response and analysis, and structural proportioning and detailing necessary to achieve intended performance.
- The design must consider all possible loads as well as soil-structure interaction.
- Do not forget to carry out a detailed wind tunnel test by suitably modelling the building and the surrounding environment.
- Confirm that the development team is aware of, and accepts, the risks associated with the use of alternative design procedures, that is, PBSD, as well as challenges.
- Specify sufficient construction quality assurance to ensure that construction conforms to the requirements of the design.
- Provide peer review by qualified experts as part of the design process. This may require hiring an international team of experts including the designers.
A summary of the design process for PBSD of tall buildings is presented here.
Performance-Based Design Process – PEER 2017
1.0 Design Process
- Confirm Design Process. Confirm that the Authority Having Jurisdiction is amenable to performance-based design alternatives - Establish the Risk Category of the Building
- Establish Performance Objectives
1.1 Seismic Input
- Determine response spectra for SLE shaking and MCE shaking, and select and modify earthquake ground motion-time series for use in nonlinear time history analysis.
1.2 Conceptual Design
- Select the structural systems and materials; their approximate configuration, proportions and strengths; and the intended primary mechanisms of inelastic behaviour.
- Apply capacity design principles to establish the target inelastic mechanisms.
- For all members of the structural system, define deformationcontrolled actions and force-controlled actions.
1.3 Basis of Design
Prepare a formal Basis of Design document that describes
- the building configuration
- the structural systems and materials of construction
- the anticipated mechanisms of inelastic response and behaviour;
- the design performance objectives;
- the specific design and analysis measures to be conducted to demonstrate acceptable performance capability;
- Deformation controlled and force-controlled actions.
1.4 Preliminary Design
2.0 Ground Motion Characterization
- Hazard Analysis: Determine acceleration response spectra for Service-Level Earthquake (SLE) and Risk-Targeted Maximum Considered Earthquake (MCE)
- Target Spectra
- Selection and modification of Ground Motion Records
3.0 Modelling for Advanced Analysis
- Conduct analyses using a three-dimensional structural model that represents the spatial distribution of mass and stiffness to an extent adequate for calculation of the significant features of the dynamic response of the building. Include the intended lateral-force-resisting system as well as any vertical-load-bearing elements and nonstructural components expected to contribute significant lateral stiffness and strength under the anticipated deformations.
- Construct the model to represent reasonably the geometry and finite size of members and components, including areas of overlap between members (e.g., panel zone regions of beam–column connections, wall–coupling beam geometries, braced-frame connections, uplift effects due to flexural elongation of walls, etc.).
- Define all modelling parameters to represent the expected, or best estimate, stiffness and strength of the components, using mean or median values from structural material or component tests.
3.1 Drift and Drift Ratio Demands
- Track peak transient and residual drifts and storey drift ratios recorded along the two orthogonal axes of the building plan, up the building height
- Monitor drifts at multiple points in the floor plan to identify building twist.
3.2 Component Force and Deformation Demands
- Track force and deformation demands in structural members and components. Report the average and peak values for comparison against specified acceptance criteriaand check maximum peak values to confirm that the demands are within the permissible modelling range. Additional demand measures, such as cumulative deformations or displacement velocities, may also be required to evaluate certain components.
3.3 Floor Diaphragms
- Model floor diaphragms to simulate the distribution of inertial forces to the vertical elements of the seismic-force-resisting system as well as transfer forces acting between these elements.
3.4 Seismic Mass, Torsion, and Expected Gravity Loads
3.5 Load Combinations
3.6 Equivalent Viscous Damping
3.7 P-Delta Effects
3.8 Vertical Ground Motion Effects
4.0 Linear Analysis (Service Level)
- Response spectrum analysis
- Linear response history analysis may also be used.
4.1 Nonlinear Analysis (Service or MCE Level)
- Nonlinear response history analysis may be used for SLE evaluation.
- Nonlinear response history analysis is necessary to evaluate earthquake demands when the structure deforms significantly beyond the elastic range. Nonlinear response history analysis is required for MCE evaluation.
- Model foundations and soil–foundation–structure interaction effects.
- Introduce ground motions at the base mat or top of pile caps, or through soil springs.
- Use hysteretic models that adequately account for all important phenomena affecting response and demand simulation at response amplitudes for the hazard level of interest.
- Methods for Establishing Component Properties: Establish component monotonic backbone curve and cyclic deterioration characteristics from a combination of physical test data and analytical approaches that have been benchmarked to physical test data.
- Component Analytical Models
- Residual Drift Demands
- Ground Motion Duration
4.2 Foundation Modelling and Soil–Structure Interaction
4.3 Structural Modelling Parameters
- Expected Material Strengths
- Expected Component Strengths
- Effective Member Stiffness
- Structural Steel Components
- Reinforced Concrete Components
- Analysis models for overall structural system response can range fromconcentrated hinge or spring models, to fibre-type beam– column or wall models, to detailed continuum finite-element models. 4.4 Response Modification Devices
- Model properties of response modification devices (such as seismic isolation, damping, and energy-dissipation devices) based on data from physical tests representing the conditions anticipated in MCE shaking.
5.1 Service-Level Evaluation
- Evaluation Criteria
5.2 Global Acceptance Criteria
- Storey Drift Limit: At the service level, storey drift in any storey is not permitted to exceed 0.5% to provide some protection to nonstructural components
5.3 Component Acceptance Criteria–Linear Analysis
Deformation-Controlled Actions:
- In members and components having deformation-controlled actions, report strains, axial or shear deformations, and rotations, as applicable.
- Document methods and assumptions used to determine the reported strains or deformations for evaluation of consistency with the acceptance criteria.
- When response spectrum or linear response history analysis is used for the SLE evaluation, calculated demand-to-capacity ratios for deformation-controlled actions shall not exceed 1.5.
Force-Controlled Actions:
- In members and components having force-controlled actions, report stress resultants including axial force, shear force, and moments, as applicable.
- Document the methods and assumptions used to determine the reported forces for evaluation of consistency with the acceptance criteria.
- Calculated demand-to-capacity ratios for force-controlled actions shall not exceed 1.0.
5.4 Component Acceptance Criteria–Nonlinear Analysis
Deformation-Controlled Actions
- In members and components having deformation-controlled actions, report strains, axial or shear deformations, and rotations, as applicable.
- Document methods and assumptions used to determine the reported strains or deformations for evaluation of consistency with the acceptance criteria.
- Strain in reinforcing bars may be limited to 2% in compression to check the possibility of bar buckling.
Force-Controlled Actions
- In members and components having force-controlled actions, report stress resultants including axial force, shear force, and moments, as applicable.
- Document the methods and assumptions used to determine the reported forces for evaluation of consistency with the acceptance criteria.
6.0 MCE Evaluation
Evaluation Criteria
- Global Acceptance Criteria & Component Acceptance Criteria as above
- Peak storey transient drift from nonlinear analysis is not permitted to exceed 3%.
- In any single storey, the deformation imposed is not permitted to result in a loss of total storey strength that exceeds 20% of the initial strength.
6.1 Unacceptable Response
- Analytical solution fails to converge
- Demands on deformation-controlled elements exceed the valid range of modelling
- Demands on critical or ordinary force-controlled elements exceed the element capacity
- Deformation demands on elements not explicitly modelled exceed the deformation limits at which the members are no longer able to carry their gravity loads
- Peak transient storey drift ratio in any storey exceeds 0.045 and Residual storey drift ratio in any storey exceeds 0.015.
- Analytical solution fails to converge
7.0 Proportioning and Detailing
8.0 Presentation of Results & Interpretation
Chinese Code on Performance-Based Design

There are two main codes in China: GB 50011-2010 on seismic design of buildings and JGJ3-2010 for PBSD of tall buildings. The basic design philosophy is like that in the PEER document. However, the details are different. The seismic intensity scale of China specifies the peak ground acceleration and velocity in addition to the effect as felt by people and extent of damage to buildings. The Modified Mercalli Intensity scale (MMI) and MSK 64 scale do not attempt to correlate the peak ground acceleration or peak ground velocity with the scale of damage.
CECR

There are three levels of earthquake motions - minor earthquakes (50 years return period), moderate earthquakes (500 years return period), and severe earthquakes (1000 to 2500 years return period). Under each earthquake, there are four levels of performance objectives. For each performance objective, and for each ground motion, they have specified extent of damage to the structure, as well as inter-storey drift demand. Code JGJ3-2010 specifies anticipated post-earthquake performances of different members of a tall building. It specifies five levels of seismic performance along with the corresponding load combinations and partial safety factors to compute flexure and shear capacities of different members. It requires non-linear analysis for performance levels 3, 4 and 5. The code requires two independent structural analysis software, for linear as well as non-linear each, to analyse a building. Out-of-code high-rise buildings are divided in two types: - Building height exceeding limits – treatment covered in one table - Structure regularity exceeding limits – treatment covered in three tables Under certain conditions, additional experiments are specified – new materials, new seismic form, and new seismic or construction technique. The test may be static, cyclic loading test of key members, or shake table test, and wind tunnel test.
A Word of Caution
There are many designers, who feel that they can easily carryout non-linear analysis of a tall building due to the availability of commercial software without really going through a rigorous training. This is simply not true. By carefully choosing various parameters, it is possible to predict a pre-desired result out of a non-linear analysis. The main problem with a non-linear analysis is that there is no “static check” on the calculations. The results must be accepted on their face value if they look “kind of OK”. The sectional properties, material properties and confining pressure have a significant influence on the moment-curvature relations that, in turn, influence the non-linear behaviour of members. Different softwares give different moment-curvature curves for the same data due to various assumptions. Selection of more realistic hysteresis model or a stress-strain curve in the strain-hardening region is a serious issue (Jain 2016a). Selection of appropriate earthquake ground motions for analysis is another headache in the absence of any strong earthquake motion ever recorded in India. Sometimes, the software converges to give wrong results or will not converge at all. This requires a serious look at all the intermediate results as well as formation of any plastic mechanism, local or global. The commercial software is to be used as a black box. Many underlying assumptions and algorithms are simply not explained in the user manual. The software should be thoroughly tested with carefully selected structural configurations with different inputs to generate a satisfactory level of confidence. They should be able to predict the expected plastic mechanism and local or global instability. The Chinese code is right in requiring the use of two independent softwares for the same building. The owners and the approving Authority, besides the designers, need to be aware of these issues and risks.
Conclusion
This paper highlights some of the shortcomings in the Indian Codes. It is a matter of great concern that these Codes are now being revised using the copy, paste and edit technology, without keeping continuity and implications in mind. The salient features of performance-based seismic design of tall buildings as elaborated in PEER (2017) are presented in brief. The Chinese Codes are also based on the PBSD design philosophy, but the details are quite different. IS16700-2017 does permit use of PBSD approach. However, the designers and the owners must appreciate the risk, liability and responsibilities involved in the PBSD.
References
1. ASCE 7 (2010) “Minimum Design Loads for Buildings and Other Structures”, ASCE/ SEI/7-10, Reston, VA.
2. Eurocode 8 (2004) “Design of Structures for Earthquake Resistance—Part 1: General rules, seismic actions and rules for buildings”,European Committee for Standardization,Brussels.
3. Jain, A.K. (2014) “The State of Codes on Structural Engineering in India”, The Bridge and Structural Engineer, Vol 44. No. 1, March, pp. 1-7
4. Jain, A.K. (2016a) “Dynamics of Structures with MATLAB Applications”, First Ed., Pearson Education (India), Noida.
5. Jain, A.K. (2016b) “Challenges Posed by Tall Buildings to Indian Codes”, TheMasterbuilder, Oct., pp.72-78.
6. Jain, A.K. (2017) “A Critical Review of IS13920:2016”, Indian Concrete Journal, Vol. 91, no. 9, pp. 18-24.
7. IS:1893-Part 1 (2016) “Criteria For Earthquake Resistant Design of Structures, Part 1 General Provisions and Buildings”, BIS, New Delhi.
8. IS 16700 (2017) “Criteria for Structural Safety of Tall Concrete Buildings”, B.I.S., New Delhi.
9. PEER (2017) “Guidelines for Performance-Based Seismic Design of Tall Buildings”, Pacific Earthquake Engineering Research Center, Tall Building Initiative, Report no. 6, Ver. 2.02, May.
10. National Codes of P.R.C. (2010) “Codes for Seismic Design of Buildings” (GB50011- 2010), China Building Industry Press, Beijing.
11. Industry Standard of P.R.C. Codes (2010) “Technical Specifications for Concrete Structures of Tall Buildings” (JGJ3-2010), China Building Industry Press, Beijing.
12. Li, Guo-Qiang; Xu, Yan-Bin, and Sun, Fei-Fei (2012) “Overview of Performance- Based Seismic Design of Building Structures in China”, International Journal of High-Rise Buildings, Vol. 1, No. 3, Sept., pp. 169-179.