Highway bridges are among the most critical pieces of infrastructure in any nation. When an earthquake strikes, a collapsed bridge doesn't just destroy property — it cuts off evacuation routes, blocks emergency responders, and can isolate entire communities for weeks. The 1994 Northridge earthquake alone caused 7 major bridge collapses on the freeway network around Los Angeles, creating traffic disruptions that cost an estimated $1.5 billion in daily economic losses.
Bridge seismic design is fundamentally different from building design. Bridges are long, irregular structures with complex dynamic behaviour, soil-structure interaction at every pier, and failure modes — such as unseating, shear key failure, and abutment displacement — that have no direct equivalent in building design. This guide walks through the complete process using the AASHTO LRFD Seismic Design framework (based on MCEER/ATC-49, the definitive NCHRP 12-49 provisions) and the AASHTO LRFD Bridge Design Specifications.
🏛️ KEY DIFFERENCES: BRIDGE vs BUILDING SEISMIC DESIGN
- 2 earthquake levels (MCE + Frequent)
- SDAP classification (not SDC)
- Displacement-based design (pushover)
- Soil-structure interaction critical
- Abutment passive resistance counted
- Unseating & span continuity checks
- RB (not R) modification factor
- Collateral hazards: liquefaction, lateral spreading
- 1 primary earthquake level (MCER)
- SDC classification (A through F)
- Force-based with drift checks
- Often rigid base assumption
- No abutment passive resistance
- Floor slab diaphragm action
- R (ASCE 7) modification factor
- Liquefaction treated via foundation requirements
1. Bridge Seismic Design vs. Building Design
Bridge seismic design in the United States is governed primarily by:
- AASHTO LRFD Bridge Design Specifications, 9th Edition (2020) — Section 3.10 for seismic provisions
- MCEER/ATC-49 (2003) — Recommended LRFD Guidelines for the Seismic Design of Highway Bridges (NCHRP Project 12-49), which forms the technical basis of modern AASHTO seismic provisions
- AASHTO Guide Specifications for LRFD Seismic Bridge Design, 2nd Edition (2011) — the performance-based alternative
- Caltrans SDC (Seismic Design Criteria), Version 2.0 — California-specific provisions, widely referenced nationally
The NCHRP 12-49 provisions introduced a fundamentally new approach to bridge seismic design, moving away from the older single-level force-based AASHTO Division I-A approach toward a two-level, displacement-based, performance-driven framework.
2. Performance Objectives & Hazard Levels
Unlike buildings, which are designed for a single performance objective under the MCER, AASHTO/NCHRP 12-49 bridge design explicitly requires checking two earthquake events:
| Earthquake Event | Probability of Exceedance | Return Period | Performance Target |
|---|---|---|---|
| Frequent Earthquake | 50% in 75 years | ~108 years | Minimal damage; bridge open to all traffic after inspection |
| MCE (Rare Earthquake) | 3% in 75 years | ~2,475 years | Significant damage acceptable; bridge may need replacement but does NOT collapse |
The Life Safety performance level (minimum required for all bridges) defines what damage is acceptable at each earthquake level:
| Earthquake Level | Service Level | Damage Level |
|---|---|---|
| Frequent | Immediate — open to all traffic | Minimal — essentially elastic response |
| MCE (Rare) | Limited — open to emergency vehicles only | Significant — plastic hinging in columns permitted |
3. Earthquake Resisting Systems (ERS) & Elements (ERE)
One of the most important innovations in NCHRP 12-49 is the explicit classification of Earthquake Resisting Systems (ERS) and Earthquake Resisting Elements (ERE) into three categories. The designer must select these early in the design process:
✅ PERMISSIBLE
Preferred systems with predictable, ductile behaviour. Examples: plastic hinging in columns above ground, abutment resistance limited to passive resistance, spread footings designed to rock.
⚠️ PERMISSIBLE WITH OWNER'S APPROVAL
Special consideration required. Examples: full passive backfill resistance relied upon (requires specified backfill material), in-ground column hinging (not easily inspectable after earthquake), isolation bearings.
❌ NOT RECOMMENDED FOR NEW BRIDGES
Brittle or non-inspectable behaviour. Examples: non-ductile piles used to resist seismic forces, wall piers used as the primary ERS without adequate ductility detailing.
4. Site Classification for Bridges
Bridge site classification follows the same soil categories (A through F) as ASCE 7, based on the average shear wave velocity v̄s, SPT N-value, or undrained shear strength in the top 30 m (100 ft). However, bridge geotechnical investigations are often more detailed than those for buildings, because foundation spring stiffnesses must be explicitly modelled.
| Site Class | Description | v̄s (m/s) | Fa (typical) | Fv (typical) |
|---|---|---|---|---|
| A | Hard Rock | >1,500 | 0.8 | 0.8 |
| B | Rock (reference site) | 760–1,500 | 1.0 | 1.0 |
| C | Very Dense Soil | 360–760 | 1.2 | 1.7 |
| D | Stiff Soil (default) | 180–360 | 1.6 | 2.4 |
| E | Soft Clay — ⚠️ High amplification | <180 | 0.9–2.5* | 2.4–3.5* |
| F | Special (liquefiable, peats) | — | Site-specific study required | |
*Site Class E amplification factors are highly period- and amplitude-dependent. Values from AASHTO tables; always check against local geotechnical investigation.
5. Constructing the Design Response Spectrum (AASHTO/NCHRP)
The AASHTO/NCHRP 12-49 design response spectrum is constructed using the same shape as ASCE 7, but must be developed for both earthquake levels (MCE and Frequent). For sites with liquefiable soils, a third, reduced spectrum for the liquefied condition must also be developed.
Olympia WA Example (MCE, Non-Liquefied — from MCEER/ATC-49-2):
| Parameter | MCE (Non-Liq.) | Frequent EQ |
|---|---|---|
| SS (mapped) | 1.175g | 0.261g |
| S1 (mapped) | 0.411g | 0.081g |
| Fa | 0.9 | 2.46 |
| Fv | 2.4 | 3.5 |
| SDS | 1.058g | 0.642g |
| SD1 | 0.986g | 0.284g |
| TS | 0.933 s | 0.442 s |
| T0 | 0.187 s | 0.088 s |
6. Seismic Design & Analysis Procedures (SDAP)
The SDAP replaces the older “analysis method” choice in bridge design. It is determined from the Seismic Hazard Level (I–IV) and the bridge's regularity and performance objective:
| SDAP | Method | When Applied | Pushover Required? |
|---|---|---|---|
| A1 | No seismic analysis | Very low seismicity only | No |
| A2 | Limited check | Low seismicity (Hazard Level I) | No |
| B | Equivalent static | Regular bridges, Hazard Level II | No |
| C | Elastic response spectrum (uniform load or single-mode) | Regular bridges, Hazard II–III | No |
| D | Elastic response spectrum (multimode) | Irregular bridges or Hazard III | Optional |
| E | Elastic multimode + Displacement Capacity Verification (pushover) | High seismicity (Hazard IV), liquefiable sites, large RB used | YES — mandatory |
The Olympia, WA bridge in Design Example 8 required SDAP E because: (1) Seismic Hazard Level IV (FaSS = 1.06 > 0.60), (2) the site has liquefiable layers causing potential inelastic foundation deformations, and (3) full passive abutment resistance was relied upon.
7. Seismic Detailing Requirements (SDR)
Separate from the analysis procedure, the SDR governs how the structure is detailed for ductility. SDR 1 through 6 correspond approximately to the old Seismic Performance Categories (SPC) A through D, but with finer granularity:
| SDR | Key Detailing Requirements |
|---|---|
| 1 | Minimum connection forces only; no special ductility detailing |
| 2 | Basic seating length requirements; nominal connection force checks |
| 3 | Minimum column ductility detailing; abutment seat width checks |
| 4 | Full ductile column detailing (confinement, lap splice restrictions, shear design); capacity design of connections; column aspect ratio limits |
| 5 | SDR 4 requirements + enhanced foundation detailing; pile confinement through potential hinge zones |
| 6 | All of SDR 5 + special inspection requirements; near-fault considerations |
8. Response Modification Factors RB
The bridge response modification factor RB (distinct from ASCE 7's R for buildings) reflects the ductility and overstrength of the bridge seismic system. A key difference from building design: the RB factor is period-adjusted because inelastic demand in short-period structures exceeds the equal-displacement assumption.
| Substructure Type | RB (MCE, Life Safety) | RB (Frequent) |
|---|---|---|
| Wall piers — strong direction | 2 | 1.3 |
| Wall piers — weak direction | 5 | 1.3 |
| Single-column bents | 4 | 1.3 |
| Multiple-column bents (SDAP E) | 6 | 1.3 |
| Connections (to cap/foundation) | 0.8 (elastic design) | 0.8 |
9. Structural Modelling of Bridges
For multimode response spectrum analysis (SDAP D and E), a 3-D spine model is typically used. Based on MCEER/ATC-49-2 Design Example 8 for a 5-span CIP concrete box girder bridge, the following modelling principles apply:
9.1 Superstructure Modelling
- Spine (stick) model: Single line of 3-D frame elements along the bridge centroidal axis. Adequate for straight, regular bridges.
- Nodes at quarter points of each span (minimum) to correctly capture mass distribution — most programs lump mass at nodes.
- Uncracked section properties for superstructure (concrete box girder) since the superstructure is designed to remain elastic.
- Density adjustment: Include additional dead loads (wearing surface, barriers, utilities) as increased density, not additional loads.
- Abutment end diaphragm: Connected to a longitudinal passive soil spring at the back face of the end diaphragm.
9.2 Substructure Modelling
- Cracked section properties for columns (effective moment of inertia Ieff ≈ 0.35–0.50 Ig for typical axial loads).
- Cap beam: Modelled with artificially high torsional stiffness to correctly distribute forces to individual columns when a single-point superstructure-to-bent connection is used.
- Rigid links from column top to superstructure centroid.
- Foundation springs at the base of the pile cap (or seal) representing the pile foundation stiffness in 6 DOF.
10. Foundation Spring Modelling
For SDAP E, foundation stiffness must be included in the model — a rigid base assumption is not acceptable. Foundation springs represent the combined lateral, rotational, and axial stiffness of the pile group.
10.1 Pile Lateral Springs (p-y method)
Lateral springs (p-y curves) are generated from soil properties and pile dimensions using the approach of Reese, Matlock, or COM624P/LPILE software. For design purposes, the pile lateral stiffness at the top is extracted and input as a single spring constant.
10.2 Passive Abutment Soil Spring
For abutments in direct contact with backfill (stub-type abutments with overhanging end diaphragm), passive resistance from the backfill provides a significant and often dominant longitudinal resistance. The passive spring value is typically estimated as:
11. Displacement Capacity Verification & Pushover Analysis
For SDAP E, a static pushover analysis (displacement capacity verification) must be performed for each pier. This is the defining feature of SDAP E and represents the shift to displacement-based design in bridge seismic engineering.
11.1 Purpose
The pushover analysis verifies that the displacement capacity of each pier (ΔC) exceeds the displacement demand (ΔD) obtained from the multimode elastic response spectrum analysis:
11.2 Pushover Procedure
- Apply gravity loads (dead load, permanent loads) to the model.
- Define plastic hinges at potential locations (column tops and bases, pile tops) using moment-curvature analysis of the section.
- Push the structure laterally in the direction of interest, increasing displacement incrementally.
- Track base shear vs. deck displacement — the pushover curve. Identify peak capacity ΔC.
- Compare ΔC vs. ΔD. If ΔC < ΔD, redesign is required (stiffer/stronger columns, or more ductile detailing).
12. Seismic Column Design & Confinement (SDR 4)
For SDR 4 (the Olympia bridge example), full ductile column detailing is required. The key provisions for circular RC bridge columns are:
12.1 Flexural Design
Columns are designed for the modified design forces (elastic force divided by RB, adjusted for the period-based modifier RT). The column interaction diagram (P-M) must envelope all load combinations including the orthogonal seismic load combination (100%+40% or SRSS rule).
12.2 Shear Design in Potential Plastic Hinge Zones
12.3 Confinement (Volumetric Spiral Ratio) — SDR 4
13. Abutment & Connection Design
13.1 Abutment Seat Width
A fundamental unseating check must be performed at all abutments and intermediate piers with expansion joints. The minimum seat width N (mm) must accommodate the elastic demand displacement plus a margin:
13.2 Shear Keys (Transverse Connection)
Shear keys transfer transverse seismic forces from the superstructure end diaphragm to the abutment stem wall. They are designed as sacrificial fuses in many designs (designed to yield, protecting the abutment stem wall from damage) or as capacity-protected elements in others. The design choice must be clearly identified in the ERS selection step.
14. Liquefaction Considerations
Liquefaction is one of the most damaging earthquake hazards for bridges. It occurs when saturated loose sands temporarily lose shear strength due to pore pressure build-up during ground shaking. Effects on bridges include:
- Loss of lateral pile support: p-y springs in liquefiable layers are dramatically reduced (often to near zero)
- Lateral spreading: Liquefied ground flows laterally toward free faces (embankments, riverbanks), imposing large horizontal forces on pile foundations
- Settlement: Post-liquefaction reconsolidation causes surface settlement of 0.1–1.0+ m
- Reduced bearing capacity: Spread footings may lose bearing capacity entirely
For SDAP E bridges on liquefiable sites, a separate structural analysis must be performed with liquefied foundation springs (p-y curves modified for liquefied soil) to determine the additional demands on pile foundations and column bases.
15. Worked Example: 5-Span Bridge, Olympia, WA (MCEER/ATC-49-2)
Bridge Summary
| Type | 5-span continuous CIP concrete box girder |
| Total Length | 500 ft (5 × 100 ft spans) |
| Location | Olympia, WA (Lat. 47.0°N, Long. 122.9°W) |
| Site Class | E (soft clay over liquefiable alluvial sands) |
| Substructure | Two-column integral bents with 24-in. CIP piles with steel casings |
| Abutments | Stub-type with overhanging end diaphragm (passive backfill resistance) |
| SDAP / SDR | SDAP E / SDR 4 |
| RB (MCE) | 6 (multiple-column bent, SDAP E, Life Safety) |
| Analysis Tool | SAP2000 Nonlinear v7.40; Mathcad for hand calculations |
12-Step Design Process (per MCEER/ATC-49-2)
- Preliminary Design: Static load (LL + DL) design; define ERS (plastic hinging in columns + passive abutment backfill). Classified as “Permissible with Owner's Approval.”
- Basic Requirements: SSS = 1.175g, S1 = 0.411g (USGS). Site Class E. Fa=0.9, Fv=2.4. SDS=1.058g, SD1=0.986g. Seismic Hazard Level IV.
- SDAP/SDR: SDAP E required (Hazard IV + liquefaction + full passive abutment). SDR 4.
- Elastic Seismic Forces: SAP2000 multimode response spectrum analysis (CQC combination). Separate MCE (non-liquefied), MCE (liquefied), and Frequent earthquake runs.
- Design Forces: Applied RB=6 and period modifier RT. Modified design forces combined using 100%+40% orthogonal rule.
- Primary EERS Design: Column interaction diagram P-M checked. Pushover model set up.
- Displacement Checks: Pushover run in transverse and longitudinal directions. ΔC ≥ ΔD verified. Seat width checks at abutments.
- Structural Components: Column shear design using overstrength plastic hinge forces. Confinement (spiral) steel designed for ductility per SDR 4. Cap beam and joint design.
- Foundation Design: 24-in. pile design for combined axial + lateral seismic loads. Pile top interaction diagram developed.
- Abutment Design: Passive soil spring force checked. Shear key design for transverse loads.
- Liquefaction: Separate analysis with liquefied p-y springs. Design for lateral spreading forces on pile caps.
- Design Complete: All checks satisfied. Construction documents prepared.
16. Tips, Facts & Common Mistakes
✓ Top 6 Bridge Seismic Design Tips
- Identify the ERS and plastic mechanism before starting analysis — it controls everything.
- Always run both liquefied and non-liquefied cases for Site Class E/F sites — one often governs foundation design, the other governs column design.
- In SAP2000/ETABS, lump mass correctly — superstructure mass not at nodes will cause wrong mode shapes and periods.
- Iterate the passive abutment spring: check if the modelled force exceeds Fmax and re-run if needed.
- Use moment-curvature analysis (not simplified formulas) for column plastic hinge properties in pushover — it dramatically affects ΔC.
- Check the 100%+40% orthogonal combination in both directions — it often produces a different critical column than the SRSS combination.
❌ Common Mistakes
- Using gross section properties (Ig) for columns in the seismic model — overestimates stiffness, underestimates demand displacement.
- Neglecting foundation springs (rigid base assumption) in SDAP E — code non-compliant and unconservative for displacement demand.
- Using the ASCE 7 two-thirds reduction (SDS = 2/3 SMS) for bridge design — AASHTO/NCHRP uses full MCE values; this is a 50% underestimate of seismic demand.
- Designing only for the MCE — the frequent earthquake often controls column and bearing design.
- Forgetting the vertical seismic effects on long-span bridges and cantilever elements (especially for near-fault sites within 10 km of an active fault).
- Not providing a clearly identified load path from superstructure to foundation — the provisions explicitly require this to be documented.
17. References & Further Reading
- MCEER/ATC-49-2 (2003) — Design Examples: Recommended LRFD Guidelines for the Seismic Design of Highway Bridges. ATC/MCEER Joint Venture. (Primary source document for this article.)
- MCEER/ATC-49 (2003) — Recommended LRFD Guidelines for the Seismic Design of Highway Bridges (Part I: Specifications; Part II: Commentary). NCHRP Project 12-49.
- AASHTO LRFD Bridge Design Specifications, 9th Edition (2020). Section 3.10 — Earthquake Effects: EQ.
- AASHTO Guide Specifications for LRFD Seismic Bridge Design, 2nd Edition (2011). AASHTO.
- Caltrans Seismic Design Criteria (SDC), Version 2.0 (2019). California Department of Transportation.
- Priestley, M.J.N., Seible, F., and Calvi, G.M. (1996). Seismic Design and Retrofit of Bridges. John Wiley & Sons.
- Youd, T.L. et al. (2001). Liquefaction resistance of soils: Summary report from the 1996 NCEER workshop. ASCE Journal of Geotechnical and Geoenvironmental Engineering, 127(10), 817–833.
- FHWA Seismic Retrofitting Manual for Highway Bridges, FHWA-RD-94-052 (1995). Federal Highway Administration.
- Reese, L.C. and Van Impe, W.F. (2011). Single Piles and Pile Groups Under Lateral Loading, 2nd Ed. CRC Press.
- USGS Seismic Design Tool: https://earthquake.usgs.gov/designmaps/
This article is based on the MCEER/ATC-49-2 Design Examples report (NCHRP Project 12-49) and current AASHTO LRFD provisions. All design must be verified by a licensed structural engineer against the applicable code and project-specific conditions.
