Engineered Flooring Dimensional Stability: What Actually Determines It

Engineered flooring dimensional stability refers to the change in board width and length for a given change in moisture content, expressed as a percentage per 1 % change in moisture content. On that measure, solid oak moves about 0.30 % per 1 % moisture content change across the grain, while an engineered board with a cross-laminated plywood core moves about 0.10–0.15 % — roughly a third to a half as much. The rest of this page sets out why that difference exists, how far it goes, and where it stops being enough.

Why does wood move at all?

Wood changes dimension only below the fibre saturation point, which averages 28–30 % moisture content. Above that point, additional moisture sits in the cell cavities and has no dimensional effect. Below it, water leaves and enters the cell walls themselves, and the walls change size.

Movement is not the same in every direction. Wood shrinks and swells anisotropically:

The tangential-to-radial ratio averages about 1.8 to 2.1 across common hardwood species, which is why quartersawn flooring moves about half as much as plainsawn of the same species. Two numbers follow from this that matter for a specification:

Source: USDA Forest Products Laboratory, Wood Handbook, dimensional change coefficients.

To see how those directions map onto the ply counts used in real boards, see our guide to two-layer, three-layer and multilayer construction.

What does the cross-laminated core actually do?

Cross-lamination reduces dimensional movement because adjacent plies are oriented at 90°, so the expansion of one layer is restrained by the stiffness of the layers above and below it. A solid board is free to move its full width; a laminated board is not, because each ply is held by its neighbours.

The USDA Wood Handbook puts a figure on it: laminated wood structures reduce dimensional instability by approximately 50–60 % compared with solid wood of the same species.

The last row is worth reading carefully, because it is where the mechanism shows up as an isotropic result: a plywood core changes dimension at roughly the same rate on its X and Y axes, around 0.5 %. A solid wood face, by contrast, moves 0.1–0.3 % along the grain and 8–11 % across it. Those two behaviours are not identical, and section 6 covers what the mismatch produces.

In specification terms this is why two-layer and three-layer boards are not interchangeable, even with the same surface species: the core arrangement, not the face, sets the movement figure. Boards built on two-layer and three-layer cores carry the lower figures in the table above.

How much movement does that mean in millimetres?

Percentages are hard to specify against; board width is not. The same moisture content swing produces movement proportional to the width of the board, so the millimetre figure scales with plank width while the percentage does not.

Over a 5 % moisture content swing — a normal seasonal range in a heated interior:

Two consequences follow, and the second is the one that catches people out.

The first is that the difference between the two constructions is about 1.5–2 mm per plank over a normal season. Accumulated across the width of a room, that is the difference between a floor that stays tight and one that needs wider expansion gaps.

The second is that board width matters as much as construction. The percentage movement is fixed by the material, but the millimetre movement is proportional to width. A 240 mm plank moves twice as far in absolute terms as a 120 mm plank of the same construction, for the same moisture swing. This is why wide-plank engineered flooring is specified with wider perimeter gaps than narrow-plank floors, even though the board is “more stable” per unit of moisture.

What decides how much moisture the floor actually sees?

Moisture content in service is set by the equilibrium moisture content (EMC) of the surrounding air, not by the flooring. At a given temperature, wood comes to rest at a moisture content fixed by relative humidity:

Source: USDA Forest Products Laboratory, Wood Handbook, moisture relations table.

The working range in a normal heated building is therefore roughly 6 % MC in winter (20–30 % RH) to 11 % MC in summer (60 % RH) — a swing of about 5 percentage points. That 5-point figure is the one to use in the calculation in section 3, and it is the number the material actually sees over a year.

Two organisations put an approved operating band around it:

The warranty point deserves to be stated plainly: an engineered floor that gaps in midwinter is usually sitting below the humidity range its warranty assumes. The floor did not fail — the indoor condition moved outside the specified band. The pre-installation half of the same question — getting the boards to the moisture content the building will hold them at — is covered in our guide to checking flooring moisture content against site conditions.

Does engineered flooring handle humidity better than solid?

Yes, and the gap widens as relative humidity rises. The two constructions behave differently once the air gets damp:

The mechanism is the same one that drives the numbers in section 2: the laminated core resists movement that a solid board cannot. The practical effect is that engineered flooring carries a much smaller safety margin requirement in variable climates. It does not make the floor immune — it moves the risk threshold from about 55 % RH to well above 70 % RH. Above 70 % RH for extended periods, moisture can still reach the core and cause swelling.

What actually goes wrong in low humidity?

Below the fibre saturation point, moisture loss causes shrinkage. In dry conditions the failure mode is therefore gaps, not swelling — which is the opposite of what most installation guides are written to prevent. In wet conditions the direction reverses and the risk becomes swelling and cupping.

The specific problem in very low humidity is end gapping, and it follows directly from the mismatch in section 2:

The wear layer and the core therefore shrink at different rates in the length direction. The result is small gaps at board ends, proportional to board length, which appear in winter’s low-RH season and close again as humidity rises. This is normal behaviour, not a manufacturing defect — and it is why prefinished flooring is supplied with beveled or eased edges, which make the seasonal shape change far less visible than the flush, square edges of a site-sanded floor.

There is a documented way to reduce the low-RH problem at source: dry the wear layer to a low moisture content before it is bonded to the core. Boards whose lamella is dried to 4–5 % MC before lamination have less moisture to lose on site, so they shrink less when the building’s air dries out in winter.

How much expansion gap is needed?

Seasonal movement has to be accommodated, not resisted. Expansion gaps are left at perimeter walls and fixed objects, sized to the expected movement rather than to a fixed number.

The gap figure is what the calculation in section 3 is for. A 120 mm plank and a 240 mm plank in the same room over the same moisture swing need different allowances, and the difference comes from the width-proportional term.

Dimensional stability is a property of the construction, not of the species. The same oak surface behaves differently depending on whether it sits on a two-layer, three-layer or multilayer core — and the figures in this page apply to the construction, before any question of surface species arises.

FAQ

Project Specification Takeaway

The calculation in section 3 is only as good as the two inputs you put into it: the moisture content swing the building will actually see, and the board width you intend to lay. The second one is where wide-plank projects most often run into trouble, because the movement the specification has to absorb scales with width rather than staying flat. If you want the millimetre figure worked through for a plank width you are considering, tell us the width and the destination climate and we will set it out.