The chain from humidity to a lifted floor
Wood exchanges moisture with the air around it until it reaches an equilibrium set by the relative humidity. That equilibrium moisture content is roughly 6 percent at 30 percent humidity and roughly 9 percent at 50, so a house that swings between those two figures over a year swings the wood in its floor by about 3 points of moisture content.
Wood then changes size with moisture content, and it does so almost entirely across the grain. The fractional change per point of moisture content is the coefficient, and it varies by a factor of several between species and between plainsawn and quartersawn material of the same species. Multiply the run across the boards by the coefficient by the moisture swing and you have the total width change of the floor. Half of it appears at each side, because a floor laid tight grows outward from its middle.
If the gap at the perimeter is smaller than that, the floor meets the wall and the movement has to go somewhere else. Upwards is the usual answer, either as a general lift or as a peak at one joint. That is the entire mechanism, and every input in it is a number you can find before installation.
Why the gap is a function of the run, not a fixed figure
Every quarter-inch-at-the-wall rule is really a rule about a room of a certain size. The same coefficient and the same humidity swing across a 12 foot run and across a 30 foot run produce very different amounts of movement, because the movement is proportional to the length of the run. A quarter inch each side is generous on a small room and inadequate on a large open plan.
Two consequences follow. First, the gap has to be checked against what the trim can actually hide, because a gap the baseboard does not cover is a gap you will be tempted to make smaller. Second, past a certain run there is no gap the trim can cover, and the run has to be broken into bays with a transition between them. That break is a real gap with a profile over it, fastened to the subfloor rather than to either floor, and it is the same detail described on the transition calculator.
Products state their own maximum run, and where a stated maximum is shorter than the arithmetic here suggests, the stated one wins. It usually accounts for things this calculation does not, including the locking system and the tolerance stack across many boards.
Acclimation, which is a moisture content target and not a waiting period
The common description of acclimation is that the boxes sit in the room for a period of days. That is the procedure, not the goal. The goal is that the material reaches the moisture content it will live at, and whether a given number of days achieves that depends on the material, the packaging, the room conditions and how far off the delivered moisture content was.
The consequence of skipping it is measurable rather than vague. If boards arrive at 9 percent and the room will hold them at 7.7, they will lose 1.3 points after installation and shrink accordingly. Across a 12 foot run that is nearly half an inch of total shrinkage, and because the floor is already fastened or locked, it appears as a gap between every pair of boards rather than at the wall. Wide boards concentrate the same total into fewer joints and each one shows more.
This is also why the room conditions during acclimation have to be the conditions the room will actually live at. Acclimating a floor in an unheated house in February and then turning the heating on afterwards achieves the opposite of the intention.
What this calculation cannot see
It is a seasonal movement model and it assumes the moisture is coming from the air. The failures that actually destroy floors usually come from somewhere else: a slab that has never dried, a crawl space with no vapour barrier, a leak under a dishwasher, a bathroom with no extraction. Wood over a damp substrate takes up moisture continuously from below rather than cycling with the seasons, and it cups because the underside is wetter than the top. No expansion gap addresses that.
So the order of work is fixed and it starts before any of this arithmetic matters. Test the subfloor or slab for moisture using the method the flooring instructions specify. Deal with whatever the test says. Then acclimate to a measured moisture content rather than to a calendar. Then lay out the gaps and the breaks. A floor that skipped the first step will fail in a way that makes the third step irrelevant.
Questions people ask
How big should the expansion gap be around a floating floor?
Large enough for the movement the run generates, and never smaller than the figure the installation instructions state. Those are two separate constraints and either can be the binding one. On a 12 foot run of solid hardwood with a 3 point seasonal moisture swing, the arithmetic gives a bit over half an inch each side, which a standard baseboard and shoe will cover. Double the run and it doubles, and at that point the trim is the limit rather than the wall. The one thing that is never right is choosing the gap to suit the trim you already bought.
Do engineered and laminate floors need an expansion gap?
Yes, and the reason is different from solid wood, which is why the coefficient field matters. Engineered wood moves less than solid because the cross-plies restrain the movement of the wear layer. Laminate has a fibreboard core that responds to humidity in its own way rather than like a piece of sawn timber. Both still move, both are laid as a single connected raft when floating, and both accumulate that movement over the whole run. Use the coefficient the product data gives rather than the solid wood default, and keep the stated minimum gap regardless of what the arithmetic returns.
What happens if I forget the expansion gap?
The floor runs out of room and the movement goes vertical. On a floating floor that shows as a general lift you can feel underfoot, or a peak along one joint line, or planks separating at their locking edges because the compression forced them apart. On a nailed floor it shows as cupping and crowning and, in bad cases, boards buckling clear of the subfloor. The diagnosis is usually straightforward: pull the shoe moulding off and look at whether the floor is hard against the wall. The repair is to cut the gap back in, which means removing trim, undercutting or trimming the flooring at the perimeter, and hoping the deformation relaxes.
How do I know the moisture content of the material?
With a moisture meter, on the actual boards, taking readings from several boxes rather than one. This matters more than any other measurement in the process, and it is the one people substitute a waiting period for. A meter also lets you compare the flooring to the subfloor, which is the comparison most manufacturers actually specify: not just that the flooring is at some absolute number, but that it is within a small range of the subfloor beneath it. Meters differ in how they read different species and thicknesses, so follow the meter instructions on species correction as well.
Where should a mid-floor expansion break go?
In a doorway or a cased opening if the run has one, because that is where a transition profile looks deliberate rather than apologetic. Failing that, at a change of direction, under a cabinet run, or at a natural line in the layout. What matters technically is that the flooring on both sides genuinely ends free with a gap between it, and that the profile covering the break is fastened to the subfloor rather than through either floor. A break that is scored into the surface but has the planks still connected underneath achieves nothing, and a profile screwed through the planks makes the floor worse than having no break at all.