Indian arm of Vexcolt UK · manufacturing in India since 2012info@vexcoltindia.com · +91 95992 91005
HomeBlog › How far apart should expansion joints be?
Design stage

How far apart should expansion joints be?

The 45 metre rule, where it comes from, the four things that override it, and the joints a building needs that have nothing to do with temperature.

This is the question we are asked earliest and answer least often, because by the time a joint supplier is involved the spacing is usually already fixed. It should not be. Where the joints go is a framing decision taken at scheme stage, and it costs nothing to get right then and a great deal to get wrong.

The short answer, and why it is only a starting point

Indian practice puts an expansion joint in a reinforced concrete building at roughly every 45 m of continuous length. IS 456:2000 deals with expansion joints in its clause 27 and refers the designer on to IS 3414 for the design and installation of the joint itself. Around 45 m is the figure that has settled into practice from that.

Treat it as a default, not a rule. It is a single number standing in for a calculation with at least four variables in it, and there are common building types where 45 m is plainly too far and others where a joint at 45 m is an expensive answer to a problem the building does not have.

Codes moveClause numbers and figures here are given as a working reference. Check them against the current edition of IS 456 and IS 3414 before they go into a drawing or a specification — the responsibility for the spacing sits with the structural engineer of record, not with the joint supplier.

What the 45 m is actually protecting against

A concrete frame that is prevented from changing length does not simply stay still. The restraint turns into stress, and concrete is roughly ten times weaker in tension than in compression, so the stress arrives as a crack — usually at the weakest section, which is rarely where anyone would have chosen to put a joint.

Three separate effects push in that direction, and they do not cancel out:

  • Thermal movement is reversible and cyclic. The frame gets longer in May and shorter in January, and does a smaller version of the same thing between two in the afternoon and four in the morning.
  • Drying shrinkage is one-directional and permanent. Concrete keeps losing moisture and shortening for months after the contractor has left. IS 456 takes a total shrinkage strain of the order of 0.0003 for design purposes — on a 45 m bay that is on the order of 13 mm of shortening that never comes back.
  • Creep keeps deforming the frame under sustained load for years afterwards, and relieves some — but not all — of the restraint stress that shrinkage builds up.

The spacing rule is a way of keeping the accumulated movement over one continuous run small enough that the frame, its reinforcement and its finishes can absorb what is left without visible distress.

Four things that override the default

1. An abrupt change in plan

This is the one IS 456 singles out, and the one most often missed. An L, a T, a U or a cruciform plan concentrates stress at the re-entrant corner regardless of how long the individual limbs are. A 30 m by 30 m L-shaped block is inside the 45 m rule on both arms and still wants a joint at the corner, because the two arms move in different directions and the corner is where they argue about it. The same applies to a sudden change of width, a wing set back from a main block, and a plan that steps.

2. A change in stiffness or foundation

A tower next to a two-storey podium is two structures, not one. They have different masses, different lateral stiffness, different foundation pressures and different settlement rates. Whether or not the length rule is triggered, the line between them needs to be free to move, and that separation is a settlement and seismic decision before it is a thermal one. The same reasoning applies where a block on raft foundations meets one on piles, or where a new wing is built against an existing building that has finished settling.

3. Exposure

Ambient air temperature is not what the concrete experiences. A shaded internal frame in a conditioned building sees a much narrower effective range than the weather does. An exposed roof slab, an open podium deck or a multi-storey car park sees a much wider one — surface temperatures on a dark exposed slab in a north Indian summer afternoon run far above air temperature. Two buildings of identical length in the same city can need different spacing purely because one is wrapped and the other is not.

4. Where you are in the country

The annual temperature range across our markets varies by more than a factor of two. Delhi NCR runs from about 4 °C to 45 °C and beyond; Chennai sits in a band closer to 20 °C to 40 °C. An identical 45 m bay accumulates more than twice the thermal movement in Delhi that it does in Chennai. That is the whole argument for treating 45 m as a default rather than a rule — and it is worked through with the arithmetic in our companion piece on calculating the gap width.

A single number cannot carry plan shape, exposure, structural system and climate at once. The 45 m figure is the answer when none of those are unusual — which, on real projects, is not most of the time.

The joints that have nothing to do with temperature

Three other kinds of separation get lumped in with "expansion joints" on drawings, and the confusion is not harmless, because they are sized by completely different calculations.

  • Settlement joints separate parts of a structure that will settle by different amounts. Sized by the geotechnical assessment, not by length.
  • Seismic separation keeps adjacent blocks from colliding when they respond out of phase in an earthquake. It is derived from the design storey drift under IS 1893, amplified for inelastic response, and in a tall building in Zone IV or V it is frequently the widest gap in the project by a large margin. We have written that up separately in what the seismic zone actually changes.
  • Control and construction joints are not building separations at all — they are crack management and pour management within a continuous element. Calling one an expansion joint on a drawing is how a 6 mm saw cut ends up on a schedule next to a 100 mm structural gap. That distinction is set out in expansion, control and construction joints.

Where more than one of these lands on the same line — and on a podium-and-tower scheme they usually do — the widest requirement governs. A joint sized for thermal movement and then asked to also absorb seismic drift is the most common way a correctly manufactured system ends up damaged by an event it was never specified for.

What we would ask for at scheme stage

If the joints are still movable on the drawing, four inputs settle almost everything:

  1. The longest continuous run in each direction, and where the plan changes shape.
  2. Which parts of the structure are exposed — roof, podium deck, open car park — and which are inside the envelope.
  3. The seismic separation the structural engineer has calculated between blocks, if any.
  4. What crosses the joint at each location: floor to floor, floor to wall, wall to wall, wall to ceiling, and what the finish is.

With those four we can return a joint schedule that names a gap, a system and a movement class per location — which is what a consultant needs in order to write a specification that a contractor cannot substitute their way out of.

Send us the drawingsIf you have GFC or concept drawings, we will return a location-by-location joint schedule at no cost, with the gap, the system and the movement class named for each. It is the fastest way to find out whether the spacing on the drawing is going to work.
Expansion joint spacingIS 456IS 3414Design stageStructural movementSpecification

Working on a joint detail right now?

Send us the drawings. We'll recommend the right system per location and return TDS and CAD details your consultant can specify directly.

Talk to our technical team