AASHTO Guide Specifications for Seismic Isolation Design: What Bridge Owners Need to Know

Search for the guide specifications for seismic isolation design and most of what comes back is a store listing. That is a thin answer for an owner who has just been told a bridge will sit on isolators. This article is the longer answer. It covers what the AASHTO Guide Specifications for Seismic Isolation Design are, which edition is current, how they sit beside the AASHTO LRFD Bridge Design Specifications and beside ASCE 7 Chapter 17 for buildings, how state departments of transportation apply them, and what to take from them when procuring isolators. We write from the owner's side. The engineer of record designs the bridge, and the owner's job is to know which questions to ask.

What the guide specifications are

The publisher is the American Association of State Highway and Transportation Officials, the association of state transportation departments that also issues the LRFD Bridge Design Specifications. A "guide" specification binds nobody on its own; it becomes binding when a state DOT references it in its design manual or an owner writes it into a contract, and the large seismic states do exactly that.

The current edition is the Fourth Edition, published in 2014, with Interim Revisions issued in 2023 under item code GSID 4 I1. The interims adopted a risk targeted approach to the design spectrum and the 2018 United States Geological Survey National Seismic Hazard Model, for uniform risk across the country. Caltrans's January 2025 bridge design memo cites the document as the "4th Edition with 2023 Interim Revisions," a fair indication of what practice treats as current.

The fourth edition expanded the 2010 third edition mainly through Appendix B, a set of 14 design examples built on two benchmark bridges with six variations each, developed under NCHRP Project 20 07, Task 262, led by Ian Buckle at the University of Nevada, Reno. The first edition dates from 1991. The 1999 edition introduced property modification factors and lower substructure response modification factors, and the 2010 edition rewrote the guide as a supplement to the LRFD specifications.

How it relates to AASHTO LRFD and to ASCE 7

The guide is written as a supplement. An isolated bridge is still designed to the LRFD Bridge Design Specifications for everything the isolation guide does not address. Where the two conflict, the isolation guide governs. Washington State's bridge manual states the combination directly: isolation bearings are designed to the LRFD specifications, the AASHTO Guide Specifications for LRFD Seismic Bridge Design and the isolation guide together.

The hazard level is the first difference from buildings. AASHTO adopted a 1,000 year return period for conventional bridge design in 2007, a 7 percent probability of exceedance in 75 years, and the 2023 interims bring the isolation spectrum onto a risk targeted basis. Buildings under ASCE 7 Chapter 17 are designed at the risk targeted maximum considered earthquake, as we cover in our guide to ASCE 7 Chapter 17 seismic isolation design requirements. The physics of seismic isolation is the same on both sides. The code environment differs, and a specification that cites the AASHTO guide on a building, or ASCE 7 on a bridge, signals a document assembled by copy and paste.

Analysis procedures

The guide offers four analysis routes. The simplified method treats the isolated bridge as a single degree of freedom system with an effective stiffness at the design displacement, an effective period from that stiffness and an equivalent viscous damping ratio from the area of the hysteresis loop. The single mode and multimode spectral methods carry the same equivalent linear idea into a full model of the bridge. The time history method, with nonlinear or equivalent linear isolator properties, is reserved for complex structures and highly damped systems.

Property modification factors and bounding analysis

Isolator properties are not fixed numbers. They shift with temperature, aging, loading velocity, cumulative travel and wear, contamination of sliding surfaces and, for elastomeric bearings, scragging. The guide handles this with system property modification factors, the lambda factors. The nominal characteristic strength and post yield stiffness, defined for fresh and scragged specimens at normal temperature, are multiplied by a minimum and a maximum factor to produce a lower bound and an upper bound set of properties, and the bridge is analyzed with both.

The minimum factor is taken as 1.0, so the lower bound equals the nominal properties. The maximum factor is the product of component factors for temperature, aging including corrosion, velocity, travel and wear, contamination for sliding isolators and scragging for elastomeric isolators. A system property adjustment factor then recognizes that all maxima are unlikely to coincide: 1.0 for critical bridges, 0.75 for essential bridges and 0.66 for other bridges, applied to the portion of the maximum factor above unity. Default values are tabulated for both bearing families, and project specific values may be established by test.

Every isolation design therefore carries two answers. The lower bound usually governs the displacement demand; the upper bound usually governs the force transmitted to the substructure. Neither bearing family escapes this; each has its own component factors. Our article on seismic isolation design under ASCE 7, 2022 edition shows the same bounding logic in the building code.

Restoring force and displacement

The guide sets a minimum restoring force. The post yield stiffness must be at least 0.025 times the weight carried by the isolation system divided by the design displacement, which means the lateral force at the design displacement must exceed the force at half that displacement by at least W/80. The period computed from the post yield stiffness is capped at 6.0 seconds. The reasoning, as the FHWA and MCEER manual puts it, is that bridge failures in earthquakes have mostly come from excessive displacement, and a strong restoring force limits permanent displacement.

The requirement applies to every isolator type. A curved sliding bearing meets it through the geometry of its surface; an elastomeric bearing meets it through the stiffness of the rubber. Beside the restoring force sit the displacement provisions: clearances at abutments and joints for the total design displacement, vertical load stability at that displacement, and uplift. Caltrans goes further and prohibits tension in isolation bearings under seismic load.

Testing requirements for elastomeric and sliding bearings

The testing article of the guide, Section 13 in the 2010 and later editions, has three tiers, and an owner should expect all three in a bearing specification. Characterization tests build the manufacturer's database of properties under varying velocity, pressure, cumulative travel and temperature, and they are the basis for the lambda factors. Prototype tests are project specific. They are run before production, typically on two full size isolators of each type and size, and they confirm that the bearings behave as the design assumed. Production tests are quality control on every bearing delivered.

The prototype sequence described in the FHWA and MCEER manual, based on the 1999 edition, shows the scope: thermal cycles, 20 cycles at the wind and braking service load, three cycles each at 0.25 to 1.25 times the total design displacement, 10 to 25 cycles at the design displacement, a repeat of the service load test, three verification cycles, and a stability test under increased and reduced vertical load beyond the design displacement. Cold region bearings are tested again after low temperature exposure, and a wear test of at least one mile of cumulative travel checks the effect of decades of thermal and traffic movement.

Two points deserve an owner's attention. The tests are run at the frequency of the isolated bridge, because heating during cyclic motion changes the properties of lead cores and sliding interfaces alike. And the acceptance criteria are tied to the design values and their bounds; a prototype that misses the nominal criteria may still be accepted if a bounding analysis with the measured properties shows the bridge works. That decision belongs before the test. Settled afterwards, it turns into a negotiation. We go through the contract side in seismic isolator specifications and testing.

Inspection and replacement

Neither the guide nor the DOT manuals assign an isolator a life in years. What they require is the ability to inspect and, if needed, replace. Caltrans lists periodic inspection and possible replacement during the design life among the consequences of isolation and asks designers to detail for it. Washington requires the designer to obtain inspection and maintenance requirements from the bearing suppliers so the bearings can be shown to function through the design life and after seismic events. Access and jacking points follow from that, and so does a commissioning record, as we cover in seismic isolator lifespan and maintenance.

How state DOTs use it

Caltrans treats isolation bearings as nonstandard elements. Every isolated bridge needs project specific design criteria that adopt, supersede or amend Bridge Design Memo 20.33, the Caltrans Seismic Design Criteria and the AASHTO isolation guide. The memo of January 2025 grew out of Memo to Designers 20 22 of 2019. Two bearing types are prequalified, the lead rubber bearing and the friction pendulum sliding bearing, and the 2019 memo recommends that the design accommodate both as viable alternatives unless a site factor rules one out. The ultimate displacement capacity of the bearing must be at least 1.5 times the design displacement, and columns must stay essentially elastic until the bearings reach 1.25 times it. The bridge must also keep a secondary earthquake resisting system, usually column plastic hinging, for an event larger than the design hazard.

Washington State's Bridge Design Manual, Chapter 9, reaches a similar place by a different route. Isolation requires approval of the State Bridge Design Engineer on the basis of a cost benefit comparison between a conventional ductile substructure design and an isolated design. Once approved, preliminary plans go to at least three bearing suppliers to confirm they can meet the requirements. The bearing capacity must be at least 125 percent of the total design displacement, and the column shear capacity must exceed the isolator force at that point. The manufacturer submits stamped calculations and shop drawings before fabrication, with agency quality assurance at the plant. Elsewhere the pattern repeats. The guide supplies the method, and the DOT supplies the hazard along with its own approval route and submittal rules.

What an owner or agency should take from it when procuring isolators

Specify a parameter set and leave the product name out. The guide works from characteristic strength, post yield stiffness, design displacement, vertical loads and service movements. Caltrans's plan sheet parameter table is a good model, written so that either prequalified family can bid against it. A specification that names one product and asks the others to match it is neither neutral nor competitive.

Write the lambda factors into the contract and define the acceptance criteria against the bounds before the first test. Give each test tier its own line, with laboratory, specimen count, test frequency and the owner's right to witness. Ask for the restoring force check at both bounds in the design submittal. Our checklist on how to choose a seismic isolator manufacturer covers the supplier side, and our system selection and BOQ preparation consulting turns these questions into a bid package.

How the building codes differ

The differences follow from the operating environment. A bridge lives outdoors and moves every day with temperature and traffic. It has no occupants to protect from fire, so the bridge guide spends its effort on temperature, travel, wear and restoring force. A building is occupied and permitted through a building official, so ASCE 7 brings mandatory review and fire protection of the isolators, and it sets the moat. The review side is explained in ASCE 7 Chapter 17 peer review requirements. In Europe the same split holds: EN 1998 Part 1 for buildings, EN 1998 Part 2 for bridges.

What this means for an owner

The guide specifications are the method behind isolated highway bridges in the United States, and the method is neutral. It does not prefer lead rubber, high damping rubber, flat sliders or curved sliders. It asks for bounded properties, a minimum restoring force, a displacement the structure can accommodate, and tests that prove the delivered bearings match the design. The useful moment to get help is before the parameter table is fixed, when the lambda factors and the test matrix still cost nothing to change. Book a consultation and bring the governing DOT manual and the draft specification.