Seismic Analysis of Structures: ELF vs Response Spectrum Analysis vs Time History Analysis

Equivalent Lateral Force (ELF), Response Spectrum Analysis (RSA), and Time History Analysis (THA) are three fundamentally different approaches for estimating the seismic response of a structure. The correct choice is not merely a question of which method is "more advanced"; it depends on the structural regularity, height, dynamic characteristics, seismic demand, soil conditions, nonlinear behaviour, importance of the structure, and the requirements of the governing design standard.

Indian-code note: The commonly circulated comparison of ELF, RSA and THA is often based on IS 1893 (Part 1):2016. BIS has subsequently developed the seventh-revision framework, including IS 1893 (Part 1):2025 for general provisions and a separate buildings part. Therefore, for a live project, the designer should verify the edition, amendments, project specifications and statutory adoption applicable to that project before finalizing seismic parameters.

1. What Happens to a Building During an Earthquake?

An earthquake does not apply a conventional horizontal force to a building in the same way as wind. The ground beneath the structure accelerates, while the mass of the building tends to resist this change in motion because of inertia.

This produces inertial forces throughout the structure. The magnitude and distribution of these forces depend on the mass, stiffness, damping, natural periods, mode shapes and foundation/soil characteristics of the structure.

Inertial force = Mass × Ground/structural accelerationF = m × a

The fundamental problem of seismic analysis is therefore to determine how the structural mass responds to ground motion and how that response is transferred through floors, diaphragms, beams, columns, walls, foundations and ultimately into the soil.

ELF, RSA and THA solve this problem at progressively different levels of dynamic representation.

2. The Three Methods at a Glance

Parameter	ELF	Response Spectrum Analysis	Time History Analysis
Basic nature	Static equivalent representation	Modal dynamic analysis	Direct dynamic response to ground acceleration
Earthquake representation	Equivalent lateral forces	Design response spectrum	Acceleration versus time records
Modes	Primarily fundamental-mode based	Multiple modes	All dynamically participating modes in the numerical model
Phase information	Not represented	Lost in conventional response-spectrum analysis	Retained
Higher-mode effects	Limited	Captured	Captured
Nonlinear analysis	No	Normally linear elastic	Can be linear or nonlinear
Ground motion record	No actual record required	No actual acceleration time record required	Actual/synthetic/modified record required
Computational demand	Low	Moderate	High
Typical use	Regular structures where permitted	Medium/high-rise, irregular or dynamically significant structures	Important, complex, nonlinear or performance-based studies

3. Equivalent Lateral Force Method (ELF)

3.1 Basic Concept

The Equivalent Lateral Force method converts the dynamic earthquake action into a set of static horizontal forces acting at different floor levels.

Instead of following the complete earthquake acceleration record, the method estimates a design base shear and distributes that force vertically over the height of the structure.

V_B = A_h WWhere: V_B = design seismic base shear A_h = design horizontal seismic acceleration coefficient W = effective seismic weight of the structure

This relationship is one of the most important equations in conventional seismic design. The quality of an ELF analysis therefore depends heavily on correctly determining the seismic weight, seismic coefficient, structural period, soil/site condition, damping and applicable code parameters.

3.2 Effective Seismic Weight

The effective seismic weight is not necessarily equal to the total dead load of the building. It generally includes the appropriate portion of permanent loads and the code-specified portion of imposed loads that can reasonably be expected to be present during an earthquake.

A common modelling error is to define an incorrect mass source. If the structural model contains only self-weight while floor finishes, walls, equipment or applicable imposed-load portions are omitted from seismic mass, the calculated seismic demand can be seriously underestimated.

3.3 Distribution of Base Shear

The total base shear is distributed to the various floor levels according to the governing code relationship. Under the conventional IS 1893:2016 approach, the distribution is related to floor seismic weight and height.

Q_i = V_B × W_i h_i² / Σ(W_j h_j²)

Thus, upper floors generally attract a larger proportion of the lateral force because their elevation above the base is greater.

3.4 Advantages of ELF

• Simple to understand and implement.
• Computationally economical.
• Useful for preliminary structural sizing.
• Provides a straightforward estimate of design base shear.
• Very useful as a benchmark against dynamic-analysis results.
• Suitable for regular structures when permitted by the governing code.

3.5 Limitations of ELF

• Does not explicitly capture higher-mode response.
• Does not reproduce the actual dynamic response of the structure.
• Can become inadequate for tall, slender or highly irregular structures.
• Does not directly provide modal participation information.
• Cannot represent nonlinear cyclic behaviour.
• May not adequately represent torsional and higher-mode effects in complicated structures.

Structural engineer's interpretation: ELF should not be regarded as a "poor" analysis method. For a regular low-rise structure, a properly executed ELF analysis can be entirely appropriate. Its weakness appears when the structure's dynamic behaviour becomes too complicated to be represented by a predominantly first-mode equivalent-force pattern.

4. Response Spectrum Analysis (RSA)

4.1 What Is a Response Spectrum?

A response spectrum is a plot showing the maximum response of a family of idealized single-degree-of-freedom systems having different natural periods when subjected to a specified ground motion or design spectrum.

For structural engineering, the most commonly used spectrum is expressed in terms of spectral acceleration, although displacement and velocity spectra are also possible.

The spectrum therefore answers an important question:

"If a structure has a particular natural period T, what level of maximum seismic response should be expected under the specified earthquake demand?"

4.2 Why RSA Is More Powerful Than ELF

A real building does not vibrate in only one shape. It possesses many natural modes of vibration. The first mode usually dominates the response of a regular low-rise structure, but higher modes can become increasingly important as the structure becomes taller, more irregular or stiffer in certain directions.

RSA performs an eigenvalue or modal analysis and obtains:

• Natural periods.
• Natural frequencies.
• Mode shapes.
• Modal participation factors.
• Effective modal masses.
• Modal seismic forces and displacements.

Each mode is then subjected to the appropriate spectral acceleration corresponding to its natural period.

For mode k:A_k = S_a(T_k) where T_k is the natural period of mode k.

4.3 Modal Participation

Not every mode contributes equally to earthquake response. A mode with a large effective modal mass in the direction being considered can make a significant contribution, while another mode may contribute very little.

For this reason, a designer should never simply assume that "10 modes are enough" or that "20 modes are enough." The appropriate number should be determined by modal mass participation and the requirements of the governing code.

For IS 1893:2016, the commonly applied requirement is that the modes considered should capture at least 90 percent of the total seismic mass in the direction under consideration.

5. SRSS and CQC Modal Combination

Once individual modal responses have been obtained, they cannot simply be added algebraically because the maximum response in each mode does not necessarily occur at the same instant. Modal responses must therefore be combined using an appropriate statistical/dynamic combination procedure.

5.1 SRSS

R = √(R₁² + R₂² + ... + R_n²)

SRSS stands for Square Root of the Sum of Squares. It is generally appropriate when modal frequencies are sufficiently separated.

5.2 CQC

CQC means Complete Quadratic Combination. Unlike simple SRSS, CQC accounts for correlation between modal responses.

This becomes particularly important when modes are closely spaced, which may occur in buildings with torsional behaviour, plan irregularity or other dynamic characteristics.

The Indian seismic provisions include procedures for treatment of closely spaced modes and modal combination.

6. Time History Analysis (THA)

6.1 The Fundamental Difference

Time History Analysis is fundamentally different from response spectrum analysis. Instead of providing only a maximum spectral ordinate for each period, THA applies an acceleration record as a function of time.

Ground acceleration:a_g(t) The structural response is calculated continuously with respect to time.

The analysis therefore produces response histories such as:

• Displacement versus time.
• Velocity versus time.
• Acceleration versus time.
• Member force versus time.
• Base shear versus time.
• Support reaction versus time.
• Plastic hinge formation versus time in nonlinear analysis.

6.2 Basic THA Procedure

1. Select appropriate earthquake ground-motion records.
2. Establish the target design spectrum.
3. Scale or otherwise modify the records as permitted by the governing procedure.
4. Define the structural model and appropriate damping.
5. Apply the acceleration record at the foundation/base.
6. Integrate the dynamic equations of motion.
7. Obtain structural response as a function of time.
8. Extract maximum and relevant response quantities.
9. Compare the results against design and performance criteria.

7. Linear vs Nonlinear Time History Analysis

Time history analysis should not automatically be equated with nonlinear analysis. There are two broad possibilities.

7.1 Linear Time History Analysis

The stiffness and material behaviour remain essentially linear elastic. The analysis provides a more detailed dynamic response than RSA but does not explicitly model yielding and permanent deformation.

7.2 Nonlinear Time History Analysis

Nonlinear THA allows the model to represent phenomena such as:

• Yielding of reinforcement.
• Plastic hinges.
• Material nonlinearity.
• Stiffness degradation.
• Strength deterioration.
• Energy dissipation.
• Residual deformation.
• Pinching or cyclic effects where appropriate.

This makes nonlinear THA one of the most powerful tools for performance-based seismic assessment, but also one of the most demanding methods in terms of modelling assumptions, numerical stability, record selection and interpretation.

8. Equation of Motion Behind Dynamic Analysis

The mathematical foundation of dynamic structural analysis can be represented by the familiar matrix equation:

[M]{ü} + [C]{u̇} + [K]{u} = -[M]{r}a_g(t)

where:

• [M] = mass matrix
• [C] = damping matrix
• [K] = stiffness matrix
• {u} = structural displacement vector
• {u̇} = structural velocity vector
• {ü} = structural acceleration vector
• a_g(t) = ground acceleration time history
• {r} = influence vector

ELF simplifies this dynamic problem into an equivalent static problem. RSA transforms the system into modal coordinates and uses maximum spectral response. THA solves the dynamic response as a function of time.

9. Why RSA Does Not Give the Actual Earthquake Response History

This is an important distinction that is frequently misunderstood.

A response spectrum gives the maximum response of an idealized oscillator for each period. It does not retain the complete phase information of the original earthquake record. Therefore, conventional RSA provides peak modal responses rather than a unique displacement, force or acceleration history.

THA, on the other hand, retains the sequence of the ground motion and calculates the response at each time step.

10. Earthquake Record Selection for THA

The quality of a time-history analysis depends strongly on the earthquake records used. Choosing a famous earthquake record simply because it has a high peak ground acceleration is not sufficient.

The record should be appropriate for the seismic environment and compatible with the target design spectrum over the structural period range that materially affects the response.

IS 1893:2016 states that time-history analysis should use an appropriate ground motion, preferably compatible with the design acceleration spectrum in the desired period range.

11. What Does "Spectrum Compatible" Actually Mean?

Suppose the design spectrum requires a particular spectral acceleration around the fundamental period of the structure. If the selected earthquake record produces substantially lower demand in that period range, the resulting analysis may underestimate structural demand.

Conversely, a record that is excessively severe throughout the entire period range may produce unrealistic overestimation.

Therefore, record selection and scaling should be treated as an engineering task rather than a software-input exercise.

12. The Role of Damping

Damping represents the mechanisms through which vibrational energy is dissipated.

For conventional elastic response-spectrum analysis, a 5 percent damping spectrum is commonly used for ordinary building structures unless a different value is justified.

In nonlinear time-history analysis, however, the treatment of damping requires much greater care. Using excessive Rayleigh damping can artificially suppress response, particularly at higher frequencies, depending on the formulation and frequency range selected.

Important: Damping should never be selected merely because it is the default value in the analysis software. The designer should understand the physical and numerical meaning of the selected damping model.

13. ELF vs RSA vs THA — Engineering Comparison

Aspect	ELF	RSA	THA
Dynamic representation	Low	High	Very high
Higher modes	Limited	Yes	Yes
Actual earthquake record	No	No	Yes
Phase information	No	No	Yes
Computational effort	Low	Moderate	High
Linear analysis	Yes	Yes	Yes
Nonlinear capability	No	Generally no	Yes
Record-selection sensitivity	None	Low	High
Modelling sensitivity	Moderate	High	Very high
Best use	Regular/simple structures	General dynamic design	Advanced/performance-based analysis

14. When Should a Structural Engineer Use ELF?

ELF is attractive when the structural system is regular and its response is dominated by the fundamental mode.

Typical applications include:

• Regular low-rise RCC buildings.
• Regular steel frames.
• Preliminary seismic design.
• Conceptual structural sizing.
• Independent verification of a dynamic model.

The applicability limits, however, must always be checked against the current governing seismic standard and the specific structural system.

15. When Should RSA Be Preferred?

RSA becomes particularly useful when the building's response cannot be adequately represented by a single lateral-force pattern.

Typical examples include:

• Medium- and high-rise buildings.
• Buildings with significant higher-mode effects.
• Buildings with plan irregularity.
• Buildings with vertical irregularity.
• Structures with significant torsional response.
• Buildings where dynamic analysis is required by code.
• Structures where reliable modal force and displacement distribution is required.

The method is widely implemented in structural analysis programs, and Indian seismic-analysis software implementations explicitly provide both equivalent-static and response-spectrum procedures based on IS 1893 parameters.

16. When Is Time History Analysis Justified?

THA is appropriate when the structural response is sufficiently complex that a spectrum-based analysis does not provide enough information, or when the project specifically requires performance-based dynamic evaluation.

Potential applications include:

• Very tall buildings.
• Highly irregular structures.
• Important or critical structures.
• Structures with significant nonlinear behaviour.
• Seismic retrofit studies.
• Performance-based design.
• Structures with special damping or isolation systems.
• Structures where actual earthquake response history is important.

17. A Crucial Point: More Advanced Does Not Always Mean More Accurate

A nonlinear time-history analysis performed with a poor structural model can be less reliable than a carefully prepared response-spectrum analysis.

For example, an advanced THA can be compromised by:

• Incorrect member stiffness.
• Incorrect cracked-section properties.
• Incorrect mass source.
• Unrealistic damping.
• Incorrect soil stiffness.
• Poorly selected ground motions.
• Incorrect hysteretic parameters.
• Unrealistic plastic hinge properties.
• Numerical instability.

Therefore:

Analysis sophistication ≠ Analysis quality

The most sophisticated method is useful only when the input assumptions are equally rigorous.

18. Structural Modelling Is More Important Than Software

Whether the engineer uses ETABS, STAAD, SAP2000, MIDAS, Robot or another structural-analysis platform, the software merely solves the mathematical model supplied by the engineer.

A seismic model should therefore be carefully reviewed for:

• Mass source.
• Material properties.
• Member stiffness.
• Cracked-section properties where applicable.
• Diaphragm behaviour.
• Rigid versus semi-rigid diaphragm assumptions.
• Support conditions.
• Foundation flexibility.
• Wall stiffness.
• Load combinations.
• Accidental eccentricity/torsion requirements.
• Modal participation.
• Fundamental period.

19. Common Errors in RSA Models

Error 1 — Looking Only at the First Mode

A designer may see a first-mode mass participation of 60–70 percent and assume the analysis is complete. It is not. The total participating mass of the modes considered must satisfy the applicable code requirement.

Error 2 — Blindly Using SRSS

SRSS is not universally superior to CQC. Closely spaced modes require appropriate treatment. The modal correlation characteristics must be considered.

Error 3 — Ignoring Torsional Modes

In an asymmetric building, translational and torsional modes can interact significantly. Ignoring them can lead to an incorrect understanding of diaphragm and member demand.

Error 4 — Incorrect Mass Source

An incorrect mass source can affect natural periods, modal participation, base shear and member forces simultaneously.

20. Common Errors in Time History Analysis

• Using arbitrary earthquake records.
• Ignoring spectrum compatibility.
• Applying incorrect scale factors.
• Using unrealistic damping.
• Using an inappropriate time step.
• Ignoring baseline correction or record-processing requirements.
• Using insufficient record duration.
• Not checking numerical convergence.
• Using nonlinear hinges without calibration or engineering justification.
• Reporting only the maximum result without understanding when and why it occurs.

21. How a Senior Structural Engineer Should Review an RSA Model

1. Check seismic zone/site parameters.
2. Check soil/site classification.
3. Check importance and structural system factors.
4. Check seismic mass.
5. Check stiffness assumptions.
6. Check fundamental periods.
7. Check mode shapes visually.
8. Check modal mass participation.
9. Check torsional response.
10. Check base shear.
11. Compare dynamic base shear with applicable code requirements.
12. Check storey drift.
13. Check storey displacement.
14. Check member forces.
15. Check foundation reactions.
16. Check governing load combinations.

22. Why Base Shear Comparison Is Important

One of the most useful engineering checks is to compare the seismic base shear obtained from dynamic analysis with the corresponding code-based minimum or equivalent-static design force, where the governing standard requires such comparison or scaling.

This prevents a dynamic model from unintentionally producing a design demand that is lower than the minimum code-level seismic demand.

However, the exact scaling procedure should be taken directly from the governing edition of IS 1893 and the applicable project requirements rather than applying an old rule mechanically.

23. ELF as a Benchmark for RSA

Even when RSA is the primary design method, ELF remains extremely useful as an independent engineering benchmark.

For example, suppose:

ELF base shear = 10,000 kNRSA base shear = 6,500 kN

The engineer should not simply conclude that RSA is "more accurate" and proceed. The difference must be investigated.

Possible reasons include:

• Insufficient modal mass participation.
• Incorrect mass source.
• Incorrect spectrum definition.
• Incorrect period calculation.
• Incorrect units.
• Incorrect damping.
• Incorrect direction of seismic input.
• Modelling assumptions.

This is why independent hand calculations remain valuable even when advanced software is used.

24. Example Conceptual Comparison

Consider a hypothetical 15-storey RCC building.

Method	What the engineer obtains
ELF	One equivalent lateral-force distribution over the height.
RSA	Multiple modal responses combined to estimate peak structural demand.
Linear THA	Structural response at every time step for each selected ground motion.
Nonlinear THA	Time-dependent response including yielding and other explicitly modelled nonlinear effects.

The computational sophistication increases from ELF to RSA to THA, but the engineering effort required to define and validate the model also increases.

25. A Practical Decision Tree

Step 1 — Is the structure regular and within the applicable limits for equivalent-static analysis?

Yes → ELF may be appropriate.

No → Continue to dynamic analysis.

Step 2 — Is a linear modal dynamic assessment sufficient?

Yes → Use RSA where applicable.

No / special performance requirements → Consider THA.

Step 3 — Is nonlinear behaviour important?

Yes → Consider nonlinear THA, subject to appropriate modelling, record selection, damping, convergence and performance criteria.

26. The Engineer's Perspective: Which Method Is Best?

There is no universally "best" seismic analysis method.

A better question is:

Which analysis method provides an adequate representation of the structural behaviour for the level of seismic performance required by the project?

Situation	Preferred starting point
Simple regular low-rise building	ELF, where permitted
Regular medium/high-rise building	RSA
Plan irregular building	RSA, subject to code applicability
Vertical irregularity	Dynamic analysis is often more appropriate
Highly irregular/tall structure	RSA and potentially THA
Performance-based design	Nonlinear analysis, often THA
Critical/special structure	Project-specific advanced dynamic assessment

27. Relationship Between the Three Methods

The three methods can be understood as three levels of dynamic idealization:

ELF ↓ Dynamic behaviour represented by an equivalent static force patternRSA ↓ Dynamic behaviour represented by multiple vibration modes THA ↓ Dynamic behaviour represented directly as a function of time

This is the central idea behind the comparison shown in the original illustration.

28. What the Original Image Gets Right — and What Needs More Explanation

The illustration correctly communicates the fundamental distinction: ELF converts earthquake action into equivalent static forces; RSA considers multiple modes using a design spectrum; and THA applies earthquake ground-motion records to obtain response over time.

However, an expert structural designer should add several important qualifications.

• ELF is not simply a "low accuracy" method; it is a code-defined idealization suitable for specified structural classes.
• RSA is still fundamentally a linear elastic modal method in conventional building design.
• RSA does not retain earthquake phase information.
• THA can be linear or nonlinear.
• Nonlinear THA requires significantly more modelling judgement than simply selecting an earthquake record.
• Ground-motion selection is a major engineering issue in THA.
• Modal mass participation must be checked in RSA.
• Closely spaced modes require appropriate modal combination procedures.
• Code applicability must be checked against the current edition and project requirements.

29. Indian Seismic Design Context

For Indian projects, the seismic analysis procedure should not be selected solely from software defaults. The designer must establish the applicable Indian Standard, project specifications, seismic hazard parameters, site conditions, structural system and analysis requirements.

IS 1893 has historically provided the central framework for earthquake-resistant design in India, with different parts addressing different categories of structures. BIS documentation identifies IS 1893 as the principal Indian seismic-design series.

The 2016 edition explicitly recognized equivalent static, response spectrum and time-history methods.

BIS has since moved to the seventh-revision framework, with IS 1893 (Part 1):2025 covering general provisions and a separate buildings part. Consequently, engineers preparing new designs should confirm the current adopted provisions rather than automatically carrying forward parameters from older 2016-based design spreadsheets or software templates.

30. Final Engineering Takeaway

ELF asks: "What equivalent static forces will represent the earthquake demand?"

RSA asks: "How does the structure respond through its different natural modes to the prescribed design spectrum?"

THA asks: "How does the structure respond at every instant when subjected to a specified earthquake ground-motion history?"

The progression is therefore:

Equivalent static demand → Modal dynamic demand → Time-dependent dynamic demand

For the practising structural engineer, the real objective is not to use the most complicated analysis available. The objective is to use a method that adequately captures the behaviour that actually governs the structure, while maintaining defensible assumptions for mass, stiffness, damping, soil interaction, seismic demand and structural nonlinearity.

Professional caution: Seismic analysis results should not be accepted solely because the software reports "analysis completed successfully." A structurally meaningful seismic design requires independent checks of the model, mass source, periods, modal participation, base shear, drift, torsion, load combinations, member forces, foundations and code compliance. For final design, the current applicable Indian Standards, amendments, project specifications and statutory requirements must govern.

31. Quick Reference Summary

ELF	RSA	THA
Static equivalent earthquake forces	Multiple vibration modes	Ground acceleration versus time
Simple and fast	More representative dynamic behaviour	Most detailed response information
Best for regular structures where permitted	Best general-purpose dynamic method	Best for advanced/performance-based studies when justified
Limited higher-mode representation	Captures higher modes	Captures time-dependent modal interaction
No actual earthquake record	Uses response spectrum	Uses selected ground-motion records
Linear static idealization	Normally linear elastic modal analysis	Linear or nonlinear dynamic analysis

References and standards to verify for project use: IS 1893 series — Criteria/Design Earthquake Hazard and Criteria for Earthquake-Resistant Design of Structures; applicable structural material standards; IS 13920 where applicable for seismic detailing; and the latest BIS-published/amended provisions applicable to the project.

This article is intended for engineering education and design-review understanding. It does not replace the governing Indian Standard, project-specific seismic hazard study, geotechnical report, or review by the responsible structural engineer.