Bridge Expansion Joint Types: Selection Guide for Engineers

Thermal movement

The base equation is straightforward:

Typical values: alpha = 6.0 x 10^-6 per degree F for normal-weight concrete, 6.5 x 10^-6 for steel. Design temperature ranges by AASHTO LRFD §3.12.2 vary from about 80 degrees F (moderate climate) to 150 degrees F (cold climate steel superstructures). State DOTs publish their own service temperature ranges; verify against the local manual.

Creep, shrinkage, and live-load rotation

• Creep and shrinkage in prestressed concrete girders typically add 0.0003 to 0.0006 in/in over the service life. Apply across the tributary length to get an additional inch or two on long spans.
• Live-load girder end rotation translates into longitudinal movement at the joint via the bearing offset, usually a few tenths of an inch on highway loadings.
• Skew and grade affect the projected movement at the joint relative to the calculated longitudinal movement. Apply skew and grade correction factors per AASHTO LRFD §14.5.

Worked example: 200 ft prestressed girder span

Assume a single-span 200 ft prestressed concrete girder bridge in a moderate climate, joint at one abutment, fixed bearing at the other.

A 3.25 inch total movement sits at the upper edge of strip seal capacity. A designer would either select a 4 inch strip seal (margin of safety) or step up to a single-cell modular system depending on traffic, expected maintenance access, and local DOT preference.

Movement range and typical applications

• Movement range: up to ~4 in total (depends on rail and gland selection)
• Typical use: prestressed concrete girder spans up to ~250 ft, steel girder spans up to ~200 ft, integral abutment retrofits
• Skew tolerance: well-suited to skew up to ~30 degrees with standard rails; higher skew uses specialty corner pieces
• Grade tolerance: handles bridge grades up to 5%+ with appropriate gland sizing

Installation and gland replacement

The economic case for strip seal is the gland: when the wearing seal fails (typically 10-15 years from de-icing chemistry, debris loading, and ozone attack), the gland alone is pulled and replaced without disturbing the steel rails. A gland replacement is a single-shift overnight closure on a typical lane. The rails themselves typically remain in service 20-30 years.

When to choose strip seal vs alternatives

• Choose strip seal when total movement is 1-4 in and the deck has a properly formed header to anchor the rails
• Step down to compression seal when movement is under 2 in and the cost difference matters more than gland replaceability
• Step down to APJ when movement is under 1.5 in and the project needs flush ride quality without a steel rail
• Step up to single-cell modular when movement is right at the 4 in edge of strip seal capacity and future traffic growth is expected

Single vs multiple support bar systems

The center beams are carried on transverse support bars sliding through bearing boxes anchored into the substructure. Two design philosophies dominate:

• Single support bar (SSB): one bar per beam, sliding through a single bearing per side. Simpler kinematics, fewer bearing wear surfaces.
• Multiple support bar (MSB): each beam rides on its own dedicated support bar with springs and equidistance devices. More components, more redundancy under fatigue loading.
• Both designs are used by major DOTs; selection often comes down to fatigue category and approving agency preference.

Movement range and load capacity

• Movement range: 4 to 32+ in total, scaled by cell count
• Live load rating: HL-93 (AASHTO LRFD design vehicle) and project-specific permit loads
• Fatigue: governed by AASHTO LRFD §6 fatigue categories, applied to the welded center beam-to-support bar connection
• Skew capability: high; modular geometry tolerates significant skew with corner detailing

Cost premium and replacement complexity

MBEJ is the most expensive joint family on a per-linear-foot basis and the most disruptive to replace. Replacement requires demolishing the deck blockout, removing the support bar bearing assembly, and recasting header concrete around the new unit. Multi-week stage construction with traffic diversion is the norm. Plan for proactive maintenance (gland replacement, bearing inspection per FHWA Bridge Preservation Guide) to delay the full replacement event as long as possible.

Cantilever vs supported finger plates

• Cantilever fingers: each finger is supported only at its anchored root, free at the meshing tip. Common up to ~4 in movement.
• Supported (sliding) fingers: tips ride on a support shelf during compression, eliminating cantilever bending. Used for 4-12+ in movement on heavily loaded decks.
• Finger length and tooth geometry are tuned to span length, traffic, and motorcycle/bicycle compatibility (some agencies require maximum 3/4 in finger gap parallel to traffic)

High-speed approach considerations

Finger plates ride well at highway speed because the wheel never crosses an open gap. The wheel transitions across overlapping steel. This makes them preferred over modular joints on long-span steel girder bridges where ride quality and noise are dominant requirements. A worn or damaged finger (impacted by debris) can produce dangerous tip loads at speed; routine inspection per the FHWA Bridge Preservation Guide is non-negotiable.

Drainage trough requirements

The drainage trough below the fingers is the actual waterproofing element and the most common failure point. Troughs clog with deicing salt sediment, pavement millings, and litter. Cleanout access (usually through a side panel) and slope-to-drain detailing matter as much as the trough material. A clogged trough sends runoff straight into the bearings, defeating the whole system.

Movement range and limitations

• Movement range: typically up to ~1.5 in total. Beyond that, the binder cannot accommodate the strain without cracking or extruding.
• Span suitability: best for bridges under ~60 ft with single fixed bearing, where total movement stays within range
• Temperature sensitivity: binder softens above ~140 F; rutting under heavy slow-moving truck loads is the dominant failure mode
• Skew tolerance: works well on skew up to ~20 degrees; higher skew concentrates strain at corners and shortens life

Retrofit applications and rapid placement

APJ shines as a retrofit. A failed strip seal or compression seal on a short bridge can be saw-cut out, the block-out cleaned, and a new APJ placed in a single overnight or weekend closure. No new headers, no anchorage drilling, no waiting for concrete cure. The minimal traffic impact often justifies the shorter lifecycle for a maintenance program managing many small bridges.

Failure modes and lifecycle

• Rutting under heavy truck loads in hot weather (binder softening)
• Cracking longitudinally over the bridging plate edge (movement strain)
• Debonding from the deck header at the block-out edge (substrate prep failure)
• Aggregate stripping under repeated wheel impact (binder fatigue)
• Typical service life: 7-12 years, vs 20+ for steel-based systems

Material requirements (AASHTO M 297)

Bridge compression seal material is governed by AASHTO M 297, which sets requirements for tensile strength, elongation, hardness, compression set, ozone resistance, and recovery after sustained compression. Test methods reference ASTM D5973 for the compression seal material itself.

Sizing relative to joint width

• Rule of thumb: the seal should remain compressed between approximately 40% and 85% of its uncompressed width across the entire thermal cycle
• Joint geometry must be milled or sawed to a consistent width; concrete header tolerances drive seal selection
• Practical movement limit: ~2 in total. Beyond that, the seal extrudes during compression or loses contact during expansion.
• Lubricant-adhesive (typically a one-part neoprene-based compound) bonds the seal to the joint walls and reduces installation friction

Lubricant-adhesive installation

Installation is the failure point. The joint must be clean, dry, and within the temperature window specified by the lubricant-adhesive manufacturer. The seal is forced into the joint with a roller or installation tool while the adhesive is wet. A seal installed in cold weather without proper joint prep will lose bond at the first thermal cycle and start leaking, and replacement requires removing all the adhesive residue from the joint walls before the next seal will seat correctly.

Lifecycle cost framework

A defensible lifecycle cost analysis combines four elements over the design life of the bridge (typically 75 years per AASHTO LRFD):

• Initial install cost (joint hardware, header concrete, traffic control)
• Maintenance cost (cleaning, inspection, gland or seal replacement)
• Replacement cost (full joint replacement at end of service life, including traffic-control premium)
• User delay cost (vehicle hours of delay during replacement events)

Modular joints often look expensive in initial cost but win in lifecycle cost on heavily trafficked urban interstate bridges, where the user delay cost of a 3-week MBEJ replacement (every 25-35 years) beats the user delay cost of repeated APJ replacements (every 7-12 years).

AASHTO LRFD design requirements

The primary federal-aid design standard is the AASHTO LRFD Bridge Design Specifications. Joints and bearings are addressed in Section 14, with thermal effects in Section 3.12 and fatigue in Section 6. Construction is governed by the AASHTO LRFD Bridge Construction Specifications, Sections 19 (bearings) and 20 (joints), which set installation, testing, and pre-set requirements.

State DOT spec variations

Every state DOT publishes its own bridge design and construction manual that tightens or modifies the AASHTO baseline. Three representative examples:

• Texas DOT: published Bridge Design Manual sets joint type by movement range and skew, with state-specific approved product lists
• Caltrans: Bridge Design Aids and Memos to Designers add seismic detailing, finger plate noise requirements, and Caltrans-specific approved devices
• Florida DOT: Structures Design Guidelines emphasize watertightness and chloride exposure resistance for coastal applications

Always specify by the local DOT's manual and approved product list, not AASHTO alone. The TxDOT LRFD Bridge Design Manual is publicly available at onlinemanuals.txdot.gov and is a useful reference for the level of detail a state-specific modification typically reaches.

Buy America domestic content compliance

FHWA Buy America rules at 23 CFR §635.410 require that all steel and iron permanently incorporated into a federally funded highway project be domestically melted, poured, and manufactured. For bridge expansion joints this affects:

• Steel edge rails on strip seals and modular joints
• Center beams and support bars on modular joints
• Steel finger plates and supporting structure on finger plate joints
• Steel anchorage hardware (studs, bolts, anchor plates) embedded in deck headers
• Steel bridging plates inside asphaltic plug joint blockouts

Elastomeric components (glands, compression seals, MBEJ bearings) and asphaltic binders are typically classified as manufactured products subject to project-specific Build America Buy America (BABA) infrastructure requirements rather than 23 CFR §635.410. Confirm classification with the project's federal-aid contracting officer; mill certificates and compliance documentation must accompany the steel deliverables.

For the broader landscape — which law applies by funding source, key deadlines, documentation checklists, and waiver criteria — see our Buy America Compliance Guide, the side-by-side BABA vs BAA vs AIS comparison, and the Domestic Procurement Standards hub.