Checked against current official sources: 3 September 2026.
A bridge looks still.
Cars pass over it.
Trains may pass over a viaduct.
People walk across.
The concrete and steel seem permanent.
But permanence is not the same as immobility.
Temperature changes.
Materials shrink.
Concrete creeps slowly under sustained load.
Traffic bends the structure.
Foundations can settle.
If the structure were forced to behave as though none of those movements existed, stress would have to go somewhere.
An expansion joint works because a strong structure sometimes survives by being deliberately flexible at the right place.
Quick Read
Singapore works partly because bridges and viaducts are designed to accommodate movement rather than pretending movement can be eliminated.
LTA’s current Civil Design Criteria for Civil and Structural Works, dated December 2025, requires structural movement joints to accommodate shrinkage, creep, thermal effects, variable actions and settlement. The joints must be sealed against water and granular material or deliberately drained, maintainable, replaceable where parts wear, safe for pedestrians and cyclists, and designed so they do not unduly impair vehicle ride quality or create excessive noise and vibration.
LTA also requires designers to calculate the design movement range of proposed joints and the loads generated as the joints accommodate those movements. The number of movement joints is to be minimised, and very large movement capacities require specific approval. That tells us something important: a joint is not a decorative gap. It is a calculated interface between parts of a structure that are expected to move relative to one another.
The deeper mechanism is:
temperature and loading change structural dimensions → bridge deck wants to expand, contract, translate or rotate → movement joint provides a controlled deformation zone → adjacent structural members avoid taking the full incompatible strain → sealing or drainage manages water and debris → replaceable joint components take wear → inspection and maintenance restore the interface before local wear becomes wider structural or ride-quality failure.
This article does not claim that expansion joints explain bridge safety by themselves, or that every bridge needs the same joint arrangement. It isolates one mechanism: a rigid-looking system can become more durable when it is given selected places where movement is expected, measured and allowed.
It also owns a different job from eduKateSG’s existing article on MRT rail expansion devices and special joints. That article concerns how the rail itself handles continuity and thermal movement. This article concerns structural movement joints in bridge and viaduct structures: the places where decks, spans or structural units accommodate movement relative to one another.
Wait, What? Concrete Moves?
Yes.
Not like a door.
Not like a wheel.
But enough to matter.
Heat generally causes materials to expand.
Cooling causes contraction.
Concrete also changes over time through shrinkage and creep.
Loads from vehicles, trains, wind and other actions deform structural members.
Foundations can experience settlement.
Each movement may be small relative to the bridge length.
But a millimetre-scale movement repeated across a long structure can create large internal forces if the structure has nowhere appropriate to accommodate it.
structures do not have to move dramatically for movement to become an engineering problem.
Thermal Expansion Is Predictable Enough to Design For
Singapore does not have winter ice and summer snow.
That does not mean bridge temperature is constant.
Solar heating changes deck temperatures.
Day and night differ.
Rain cools surfaces.
Different materials absorb and release heat differently.
The engineer therefore estimates a design movement range rather than assuming tropical weather means no thermal movement.
Shrinkage Means New Concrete Does Not Keep Its Original Size Forever
Concrete changes as it cures and loses moisture.
Some of that change is shrinkage.
It develops over time rather than arriving in one instant.
A movement joint must therefore accommodate not only today’s thermal cycle but also longer-term dimensional change in the structure.
This is a useful distinction:
some movements repeat; some accumulate.
Creep Means Load Can Keep Changing the Structure After Construction Ends
Concrete under sustained stress can continue deforming slowly.
This time-dependent deformation is creep.
The bridge therefore has a memory of load.
Part of the structure’s future geometry depends on what forces it has carried and for how long.
LTA’s current criteria explicitly includes creep among the movements structural joints must accommodate.
Variable Actions Add Faster Movement on Top of Slow Movement
A bus crosses.
A heavy vehicle passes.
A train enters a viaduct span.
Wind acts.
The structure deforms elastically under changing actions.
That means the joint has to accommodate both slow long-term movement and repeated operational movement.
A bridge joint lives at the intersection of several time scales.
Settlement Moves the Support, Not Just the Deck
A joint can be designed perfectly for thermal change and still face another movement source.
Foundations interact with ground.
If support conditions change through settlement, relative geometry between structural units can change.
LTA includes settlement among the effects movement joints must accommodate.
This reminds us that structural interfaces must sometimes absorb movement caused outside the component itself.
The Joint Is a Deliberate Weakness That Makes the Whole Structure Stronger
This sounds contradictory.
Why put a gap in a bridge?
Because the alternative may be forcing incompatible movements into the deck, bearings, supports, surfacing or other structural components.
The joint localises movement at a component designed to tolerate it.
a controlled discontinuity can protect a larger continuity.
Why LTA Wants Fewer Movement Joints, Not More
If movement joints are useful, why not install them everywhere?
Because each joint is also a maintenance interface.
It experiences traffic.
It can admit water or grit if sealing fails.
It can generate noise.
It contains wearing components.
LTA’s December 2025 criteria therefore says the number of movement joints should be minimised.
The optimum is not:
maximum flexibility.
It is:
enough deliberate flexibility to accommodate necessary movement with as few high-maintenance interfaces as practical.
Movement Capacity Is Calculated, Not Guessed
LTA requires designers to submit the design movement ranges for proposed joints.
That range represents the expected relative movement the joint must survive.
Too little capacity and the joint can bottom out, tear, bind or transfer damaging forces.
Too much unnecessary capacity can increase size, complexity, cost and maintenance burden.
Good design fits the joint to the movement envelope.
The Joint Also Generates Forces While Allowing Movement
A movement joint is not frictionless nothingness.
Its components resist, seal, guide and transfer loads.
LTA therefore requires calculations for resulting loads generated by the joint as it accommodates movement.
The interface changes the structure while being changed by the structure.
an accommodation mechanism can itself become a load source.
Water Is One of the Joint’s Quietest Enemies
Rain lands on the bridge deck.
Traffic carries grit.
If water and debris enter a joint uncontrollably, they can reach bearings, concrete surfaces, steel components or drainage paths below.
LTA’s criteria therefore requires movement joints either to be completely sealed against ingress of water and granular material or deliberately designed to drain those materials away without accumulation.
The joint has two jobs at once:
- allow movement;
- control what crosses the opening.
Sealing and Movement Pull in Opposite Directions
A perfect rigid seal would resist motion.
A perfect free gap would allow water and debris through.
Joint engineering exists in the compromise between those goals.
The component must remain flexible enough to move and controlled enough to protect the structure below.
Interfaces often exist because two valid requirements conflict.
The Road User Is a Receiver Too
Drivers feel the joint.
Cyclists cross it.
Pedestrians may cross it.
LTA therefore requires movement joints not to unduly impair vehicle ride quality and not to present a danger to pedestrians or cyclists.
Where pedestrians have access, LTA’s current criteria calls for a non-slip cover plate, while open gaps on riding surfaces are constrained.
The joint is structural infrastructure and a user interface simultaneously.
Noise Is Evidence of an Interface Being Crossed
Drivers know the sound.
Tyres cross a joint.
There is a change in sound and vibration.
LTA requires joints to avoid undue noise or vibration from vehicular traffic.
This matters particularly where structures sit near homes or where repeated impacts can worsen component wear.
A good joint therefore has an acoustic receiver as well as a structural receiver.
Replaceable Wear Parts Turn Failure into Maintenance Instead of Reconstruction
Movement joints are exposed.
Traffic passes directly over them.
Water and grit attack them.
They move repeatedly.
LTA explicitly requires parts liable to wear to be easily replaceable and joints to enable easy maintenance.
This is modular durability.
when a component is expected to wear, design the wear into a replaceable layer rather than pretending the entire structure will remain untouched forever.
Maintainability Is Part of Structural Design
A brilliant joint that requires closing an entire transport corridor for a month every time one seal wears is not operationally brilliant.
LTA’s criteria requires maintainability and replaceability because infrastructure lives inside an operating city.
The maintenance receiver is therefore another stakeholder:
- Can crews inspect the joint?
- Can worn parts be replaced?
- How much traffic disruption is needed?
- Can water and debris be cleaned out?
- Can damage be detected before it threatens adjacent structure?
The Bottleneck Is the Interface, Not the Span
A bridge span can be structurally strong.
The joint can still become the recurring maintenance bottleneck.
Why?
Because the joint concentrates several difficult jobs into a narrow strip:
- movement;
- vehicle impact;
- sealing;
- drainage;
- surface continuity;
- noise control;
- and replaceability.
Interfaces often become bottlenecks precisely because multiple systems meet there.
Why Very Large Movement Capacity Gets Special Attention
LTA’s December 2025 criteria says movement joints with total longitudinal movement capacity exceeding 100 mm should not be used unless otherwise approved by the Authority.
This does not mean 100 mm is a universal danger threshold for every bridge everywhere.
It means unusually large joint movement deserves additional scrutiny under LTA’s design framework.
Large movement tends to increase interface complexity.
More movement can mean larger gaps, more elaborate mechanisms, greater maintenance demands and stronger consequences if something binds or fails.
The Expansion Joint and The Ramp Use Opposite Flexibility
The Ramp makes a fixed level difference easier for people to traverse.
The Expansion Joint makes changing structural geometry easier for the bridge to survive.
One gives the human a gentler path.
One gives the structure a controlled movement path.
Both reveal that geometry can reduce stress when abrupt incompatibility would otherwise create failure.
The Expansion Joint and The Road Hump Use the Same Road Surface for Opposite Purposes
The Road Hump deliberately changes surface geometry to make excessive speed uncomfortable.
The Expansion Joint tries to accommodate structural movement without unduly disturbing ride quality.
One wants the user to feel a designed constraint.
The other tries to make a necessary structural discontinuity feel as continuous as practical.
The Expansion Joint and The Noise Barrier Share a Maintenance Reality
Both are exposed interfaces.
Both can perform well on the day they are installed and degrade later.
Both have geometry that matters to performance.
Both can produce unwanted noise if condition or design is poor.
Asset performance is not frozen at commissioning.
Competing Explanation: Why Not Build the Whole Bridge as One Continuous Piece?
Engineers often try to reduce the number of movement joints precisely because joints are maintenance-intensive.
Continuous structural systems can sometimes distribute forces and eliminate some joints.
But continuity does not eliminate movement.
It changes where movement and restraint are accommodated—through bearings, structural flexibility, supports, foundations or other details.
The correct question is therefore not:
joints or no movement?
It is:
where should unavoidable structural movement be accommodated with the best balance of strength, durability and maintainability?
Model Limit: Expansion Joint Is a Convenient Name for a Wider Family of Movement Joints
Every movement at a structural joint is not purely thermal expansion.
LTA’s current criteria uses the broader term movement joint and explicitly includes shrinkage, creep, thermal effects, variable actions and settlement.
“Expansion joint” is familiar public language.
The engineering reality is broader.
The article therefore uses the familiar title while keeping the mechanism precise.
Primary-School Lens: Give the Bridge Two Temperature States
Draw a long bridge on paper.
Mark one version “cool morning” and another “hot afternoon.”
Ask the child:
if the bridge becomes slightly longer when hot, where can that extra length go?
Then draw a controlled joint.
The child learns that strong things can still need room to change.
Secondary-School Lens: Separate Repeating and Accumulating Movement
Give students five movement causes:
- day-night thermal cycle;
- seasonal temperature variation;
- concrete shrinkage;
- creep;
- vehicle loading.
Ask which repeat, which accumulate and which depend on traffic.
Then ask how one joint must survive all of them together.
The lesson becomes superposition of structural effects across time scales.
JC Lens: Compatibility, Strain and Boundary Conditions
At JC level, the movement-joint problem can be understood through compatibility.
A restrained member that wants to expand develops stress because displacement is constrained.
A joint changes the boundary condition.
Instead of forcing all thermal strain into internal stress, the system allows a controlled relative displacement within the design movement range.
The engineering question becomes:
which movements should be resisted, which should be accommodated, and how should the interface transfer necessary loads while remaining sealed, durable, safe and maintainable?
Thought Experiment: Seal the Bridge Solid at Its Hottest State
Imagine a joint locked completely while the deck is thermally expanded.
Night arrives.
The deck wants to contract.
It cannot move as designed.
Forces relocate into adjacent components.
The thought experiment shows that restraining motion does not delete motion demand.
Thought Experiment: Perfect Movement, No Seal
The joint moves beautifully.
Rainwater and grit pour through.
Bearings and concrete below remain wet and dirty.
The movement problem is solved.
The durability problem is created.
Interfaces have to solve several constraints simultaneously.
Thought Experiment: Perfect Joint, Impossible to Replace
The joint lasts fifteen years.
A wearing component finally fails.
Replacement requires demolishing a large part of the bridge deck.
The original component was durable.
The lifecycle design was poor.
This is why LTA explicitly requires easy replacement of parts liable to wear.
Why Singapore Works Does Not Mean Every Bridge Joint Is Quiet or Problem-Free
Joints can wear.
Seals can fail.
Debris can accumulate.
Impact noise can develop.
Unexpected settlement can exceed assumptions.
Poor detailing can create cyclist or pedestrian hazards.
The best structural solution may sometimes reduce or eliminate movement joints through another bridge concept.
The serious claim is narrower:
LTA’s current structural design framework treats bridge and viaduct movement as a calculated reality, requiring movement joints where used to accommodate thermal and time-dependent deformation, live actions and settlement while remaining sealed or drained, safe to cross, maintainable and replaceable.
The joint does not make the bridge rigid.
It makes rigidity less brittle.
The Fifteen-Question Expansion Joint Test
- Movement source: Which thermal, shrinkage, creep, load or settlement effects drive movement?
- Range: What total longitudinal, transverse and rotational movement must be accommodated?
- Restraint: Which movements should the surrounding structure resist instead?
- Loads: What forces does the joint itself generate while moving?
- Seal: Can water and granular material be kept out?
- Drainage: If not sealed, can water and grit be deliberately removed?
- Ride quality: Does the interface remain acceptable for vehicles?
- Cyclists: Does the gap avoid trapping or destabilising wheels?
- Pedestrians: Is a safe non-slip surface provided where required?
- Noise: Does traffic over the joint create undue impact sound or vibration?
- Wear: Which components are expected to deteriorate first?
- Replacement: Can worn parts be replaced without major reconstruction?
- Inspection: Can condition be observed and maintained safely?
- Joint count: Could the bridge concept use fewer joints without transferring risk elsewhere?
- World return: Do measured movement, leakage, noise and condition data still match the original design assumptions?
Frequently Asked Questions
Why do bridges need expansion joints?
Because bridge structures move through thermal effects, shrinkage, creep, variable loading, rotation and settlement. A movement joint provides a controlled location for relative displacement between structural units.
Does LTA call them expansion joints?
LTA’s current Civil Design Criteria uses the broader term “movement joints” and requires them to be designed in accordance with bridge expansion-joint requirements. The broader term is useful because the movement is not caused only by thermal expansion.
Why must the joint be sealed or drained?
Because water and granular material entering the joint can affect components below and create durability or maintenance problems. LTA requires either effective sealing or deliberate drainage that prevents accumulation.
Why does LTA want to minimise the number of movement joints?
Movement joints are necessary interfaces but they also require inspection, sealing, replacement and traffic interaction. Fewer joints can reduce lifecycle maintenance where structural design can safely accommodate movement another way.
Is this the same as a railway rail-expansion device?
No. Rail expansion devices manage movement and continuity in the steel rails themselves. This article concerns structural movement joints in bridge and viaduct structures.
What is the main student lesson?
Strength does not always mean refusing to move. Durable systems often become stronger by deciding where unavoidable movement will happen and designing that interface to tolerate, inspect and repair it.
Sources and Further Reading
- Land Transport Authority — Civil Design Criteria for Civil and Structural Works, December 2025.
- Land Transport Authority — Materials and Workmanship for Civil and Structural Works.
Final Thought: The Gap Is Part of the Strength
A child sees the thin line across the bridge and may think something was left unfinished.
The engineer sees something else.
Room for heat.
Room for shrinkage.
Room for load.
Room for decades of structural life.
That is why Singapore works, in another quiet way:
the city understands that some structures endure not because every part refuses to move, but because movement has been given one disciplined place to go.