The door has warped: is it a furniture defect or the room?

In short. Wood exchanges moisture with the air and moves: on its own that is not a defect. A panel bows when the exchange happens on one face only, and the asymmetry can sit in the construction or in the room. The direction of the bow tells you which face gave up moisture, and therefore which of the two to investigate.

Why wood moves, and why moving is not a defect

Wood is hygroscopic: it exchanges water vapour with the air until it comes into equilibrium with it. If the air is drier the wood gives up moisture and shrinks; if it is damper it takes moisture up and swells, and almost all of that movement happens across the grain, in width. Wooden furniture that does not move does not exist, and a specification requiring it not to move is asking for something no supplier can give.

What can be required, and required in full, is that the movement be symmetrical. A panel that takes up or gives off moisture equally on both faces changes dimension and stays flat; one that does so on a single face bows. The defect, where there is one, is not in the movement: it is in the asymmetry.

What makes a panel bow?

The face that loses moisture shrinks and becomes shorter than the other, and the panel turns concave on that side. That gives a reading rule which costs thirty seconds on site: the hollow always faces the drier side. If a wardrobe door bows towards the room, it is the room-facing side that has given up moisture — or the wall-facing side that has taken it up; if it bows towards the wall, the reverse holds. It is the reading that separates two families of causes put right in opposite ways.

Which are the real construction defects?

They are the cases where the asymmetry sits inside the piece. There, the room is only the occasion.

• unbalanced veneering. Species, thickness and grain direction have to match on both faces, including the one nobody sees;
• finish on one face only. Primers and lacquers are a vapour barrier, so a door lacquered at the front and raw at the back exchanges moisture on one side alone. It is the commonest cause of all, and it is a defect: in contract furniture that face has to be closed even if it faces the wall;
• unsealed edges. End grain takes moisture up far faster than the faces, and the forgotten edge is nearly always the bottom one;
• solid wood rigidly restrained. A top screwed rigidly across the grain either bows or splits: the fixing has to be designed so the piece can move;
• undersized hardware. A tall door on two hinges instead of three does not cause the bow, but it does nothing to resist it.

What does the building do, and why does it never reach the report?

The other family of causes is not in the furniture, and on a hotel contract it is the commoner of the two.

• two kinds of air. A wardrobe set against an external wall has its back in still air, cooler and damper, and its front in a heated room: the same panel in two climates;
• the building drying out. Screeds, plaster and paint give off water for weeks. Furniture that arrives in that period takes it up, and when the heating goes on gives it off quickly: the bow appears some weeks after delivery, which is to say exactly after the report has been signed;
• closed rooms. Between delivery and opening, rooms are often shut up, unheated and unventilated: the worst condition of all, and written down nowhere;
• local sources. A radiator under a top, underfloor heating beneath a side panel, a door in front of sunlit glazing: they heat one face and not the other.

None of these is a defect of the furniture, but all of them are foreseeable, and that is what makes them something an order can deal with.

How do you measure it, before arguing about it?

• the amount: a straightedge laid on the piece and the deviation measured at its worst point. By eye a bow can be seen but not measured, and what is not measured can neither be claimed nor defended;
• the direction: hollow towards the room or towards the wall. This is the diagnostic reading, and it is lost if it is not noted at once;
• the spread: how many pieces, in how many rooms, on which facade. If they all bow the same way, the cause is in the piece; if only the rooms on one side of the building bow, the cause is the building.

It is the same reasoning as the guide on fabric pilling: the question «defect or not» is not settled by looking closely at the piece, but by looking at how the phenomenon is distributed.

Can a warped door be straightened?

Sometimes yes, on one condition: remove the asymmetry instead of fighting the bow. Closing the face that was left raw and letting the panel come back into equilibrium works, slowly, provided the piece has not been taken past its elastic limit; clamping it flat without removing the cause puts it back into bow as soon as the clamp comes off. Where the asymmetry sits in the construction of the panel instead, there is no recovery and the piece is replaced.

And here is the thing worth saying first, which almost nobody says: if the cause is the room, replacing the piece without correcting the room produces a new piece that bows like the first. The building first, then the piece.

What should the order and the report say?

• the tolerance: the deviation allowed on a door and on a top, measured as above. It is the figure that turns «it has warped» into something verifiable, and it is agreed with the maker beforehand, not afterwards;
• the balance: panels finished on both faces and edges sealed, including where it does not show. It is one line, and it closes off the whole family of real defects in advance;
• the state of the building at delivery: two lines in the handover report. Whether screeds and plaster were dry; whether the building was heated and ventilated. It is the only way to hold, weeks later, a document that says what the air in those rooms was doing;
• who ventilates the rooms between delivery and opening. If the furniture goes in while the site is still live, the question becomes one of custody as well;
• when the check happens: acceptance looks at the piece as it arrives, and the bow arrives later. What counts as a defect at delivery and what counts as a latent fault has to be settled in writing beforehand.

Where supply ends and works begin changes the liability regime (guide on warranty), and with it when the final payment falls due.

How we work

We do the woodwork inside our own manufacture, alongside marble, metal and upholstery, and that changes exactly the part of the job this page is about: balancing a panel and closing its hidden face are decided in production, and no inspection on site can put them right afterwards. On a hotel contract the most valuable thing is time, and a door to be remade once the rooms are furnished costs more in time than the panel costs in material. What the mock-up room settles, and how a specification clause is written, are in the two dedicated guides.

Frequently asked questions

How much can a door move before it is a defect?

There is no figure that holds in the abstract, and if the order does not set one the discussion stays one opinion against another. Where no tolerance was agreed, the reading that matters is the spread: one piece out of fifty and fifty out of fifty are not the same problem.

Can wooden furniture be delivered into a building that is still damp?

It is done, and it is normal. What changes everything is writing it down, because the state of the building on the day of delivery is precisely the reading that is missing when the complaint arrives.

Can a specification require a given air humidity?

It can set out the expected service condition. What matters is the part that is almost always missing: who verifies it, and at what point. A condition nobody measures protects neither the buyer nor the supplier.

Talk to us The contract division

This page describes assessment and selection criteria. It cites no technical standards and quotes no prices, no response times, no tolerance values and no company case histories. No numerical figure for the moisture content of wood or for air humidity is given, because the applicable limits depend on the project and on the use of the rooms and have to be agreed between the parties. The guidance on panel balance, edge sealing and the restraint of solid wood is that of good woodworking practice and has to be verified case by case.