Engineered Flooring Dimensional Stability: The Numbers

Stabilité dimensionnelle du parquet contrecollé : quels sont les facteurs qui la déterminent réellement ?

Parquet en chêne contrecollé à larges lames, avec une illustration du noyau à trois couches à grain croisé, garantissant une meilleure stabilité dimensionnelle face aux variations d'humidité.

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:

Diagram showing tangential, radial and longitudinal wood dimensional change, with tangential shrinkage greater than radial and longitudinal shrinkage.
AxisTypical changeWhere it shows on a floor
Tangential (across the growth rings)8–11 % total, green to oven-dryBoard width — the dominant movement
Radial (perpendicular to the rings)Roughly half the tangential figureBoard width, reduced by quartersawing
Longitudinal (along the grain)0.1–0.3 %Negligible; length is effectively fixed

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:

EspècesTangential coefficient (per 1 % MC)Radial coefficient (per 1 % MC)Quartersawn reduction
Red oak0.003690.001582.3×
Chêne blanc0.003650.001802.0×
Hard maple0.003530.001652.1×
Black walnut0.002740.001901.4×
Black cherry0.002480.001262.0×

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.

Cross-section comparison of solid oak flooring and three-layer engineered flooring, showing a 90-degree cross-grain core that restrains wood movement.

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.

La constructionCore arrangementMovement per 1 % MCRemarques
Solid woodSingle piece≈ 0.30 %Movement across full board width
Two-layerTwo layers at an angleLower than solidThinner build-up
Trois couchesThree layers, centre at 90°≈ 0.10–0.15 %Standard engineered specification
Multicouche5–9+ plies, alternating anglesLowest, approaching plywood behaviourPlywood core moves isotropically at ≈ 0.5 % on both axes

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:

BoardMovement per 1 % MCMovement over 5 % MCOn a 200 mm plank
Solid oak≈ 0.30 %≈ 1.5 %≈ 3.0 mm
Engineered (3-layer)≈ 0.10–0.15 %≈ 0.5–0.75 %≈ 1.0–1.5 mm

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:

Chart showing the relationship between relative humidity and wood equilibrium moisture content at 20°C, including the 35–55% NWFA recommended range.
Approximate equilibrium moisture content of wood at 20°C across different relative humidity levels. The shaded area indicates the commonly specified 35–55% indoor RH range. Actual EMC may vary with species, temperature and sorption history.
Relative humidityEMC at 20 °CTypical condition
20–30 %≈ 4.5–6.3 %Heated interior, winter
40 %≈ 7.5 %Cooler winter interior
50 %≈ 9.5 %Typical interior, year-round average
60 %≈ 11.3 %Humid summer interior
70 %≈ 13.5 %Unconditioned space
80 %≈ 16.0 %Covered exterior

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:

SourceRelative humidity rangeStatus
NWFA (National Wood Flooring Association)35–55 %Industry installation guideline
Manufacturer warranties (typical)35–55 %, year-roundCondition of coverage

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:

Relative humiditySolid woodEngineered
40–55 %StableStable
60–70 %Moderate risk of cupping and gappingPerforms without significant change
Above 70 %High risk of swelling and bucklingRemains dimensionally controlled

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:

ComponentMovement patternRate
Plywood coreIsotropic — same on X and Y axes≈ 0.5 % on both axes
Solid oak wear layer, along grainVery low≈ 0.1 % longitudinal

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.

SymptomConditionDirection of change
End gaps at board endsLow RH, winterLength shrinkage
Gaps at side jointsLow RH, winterWidth shrinkage
CuppingMoisture differential, top to bottomUneven MC through the board
Crowning / bucklingHigh RH, sustainedSwelling under restraint

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.

ParameterTypical figureSource basis
Perimeter expansion gap≈ 12 mmNWFA installation guidance (½″ at fixed objects)
Gap sizing ruleProportional to board width and expected MC swingSection 3 calculation
Fixed objects (columns, door frames)Same allowance as perimeterMovement cannot be restrained

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

What is dimensional stability in flooring?

It is the change in board width and length for a given change in moisture content, expressed as a percentage per 1 % change in moisture content. Solid oak is about 0.30 % per 1 %; cross-laminated engineered boards are about 0.10–0.15 %.

Is engineered flooring more dimensionally stable than solid wood?

Yes. The cross-laminated core restricts movement in the width direction. The USDA Wood Handbook puts the reduction in dimensional instability at approximately 50–60 %.

How much does engineered flooring move seasonally?

Over a typical 5 percentage point moisture content swing, engineered flooring moves about 0.5–0.75 % of board width, roughly 1–1.5 mm on a 200 mm plank. Solid oak of the same width moves about 3 mm.

What humidity should engineered flooring be kept at?

The NWFA recommends 35–55 % relative humidity, and most manufacturer warranties require that range year-round. At 20 °C, 50 % RH corresponds to about 9.5 % wood moisture content.

Why does my floor gap in winter?

Low relative humidity causes shrinkage. The plywood core and the solid wear layer shrink at different rates along the length, producing end gaps proportional to board length. This is normal seasonal behaviour and the gaps close as humidity rises.

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.

Partager l'article :
Obtenir des solutions exclusives en matière de produits de revêtement de sol en bois dur

“Abonnez-vous pour connaître les dernières tendances en matière de fabrication de planchers de bois franc contrecollé ! Entrez votre email maintenant ! !!”

En tant qu'importateur de revêtements de sol en bois, les solutions fiables et innovantes sont la clé de notre succès. Notre collaboration avec duramagicfloor a permis d'enrichir notre gamme de produits avec des revêtements de sol en bois très stables, améliorant à la fois l'esthétique et la fonctionnalité.
michael
Emre Yılmaz
PDG
Demande de devis
fr_BEFrench

Parlez de votre projet

Notre équipe est prête à discuter de vos besoins à tout moment. Que vous ayez des dessins de conception spécifiques ou que vous souhaitiez en savoir plus sur nos produits habituels, nous serons heureux de vous fournir de plus amples informations et des échantillons. Soumettez les informations relatives à votre projet et entamez notre collaboration.