4

I made some kitchen cabinet doors (1x3 maple boards on the outside, t&g joints, 1/4" plywood center panel glued into the groove). The doors are stained and finished in poly.

Then I installed them, and realized that one of them has a pronounced twist. Serves me right for not clamping it flat while I was gluing it up. The upper corner on the non-hinge side is sitting about 3/8" from the cabinet body when the door is closed. The defect is too big to disguise by playing with the hinge adjustment.

So I'm looking for a way to straighten the door. I found some articles online, but I don't know which course I should follow. Or if it would be less work to just build a new door.

enter image description here

This seems like it would work, but I'm not sure how to calculate the amount of curvature. I'd probably have to go through a lot of trial and error while cutting the cleat.

enter image description here

This is easy to adjust, but I can't tell if it's supposed to fix twist, or if it's just intended for doors that are sagging (out of square).

Lastly, some people say I could clamp it in a jig that twists it in the opposite direction and leave it sitting there for a while - but others say that doesn't work.

4
  • 2
    Honestly, if you have the spare materials, making a new door is probably quicker, easier, and more satisfying long-term.
    – MattDMo
    Oct 15, 2022 at 16:34
  • 1/4" plywood center panel glued into the groove I would expect this may give you issues in the future, too. The panels are supposed to float in the groove so the wood of the frame can move freely.
    – FreeMan
    Oct 18, 2022 at 16:12
  • @FreeMan I thought that's only when the center panel is solid wood? The way I heard it, the only reason not to glue the center panel is to allow for expansion - and plywood doesn't expand. Oct 18, 2022 at 20:10
  • Can you ascertain what's causing the bow of the completed door? It's important to consider why the door has bowed: If it's because the joints are not true/square your remedy, and chances of success, will likely be different than if the rails/stiles have themselves bowed or twisted.
    – gnicko
    Oct 23, 2022 at 14:25

1 Answer 1

5

Note: this Answer is predicated on the door being unfinished. As it has now been revealed that the door has stain and poly on it some of the tips don't apply but I'm leaving it as-is for future searchers who may be looking for help on a similar issue and whose doors are still bare wood.


3/8" is quite a lot, but perhaps still not too much to take out by counter-clamping or other remedial action. The thing is there's no way to determine whether anything is going to work other than actually trying it.

No amount of advice on this front — from experienced users or not — will actually tell you what'll happen with your door. Even if various someones who have successfully done this before had the door in front of them none could unerringly predict the outcome. They can give you their various educated guesses sure, but those are still guesses..... which is precisely why there's a 100% chance they'll vary.

This is sort of summarised neatly by something in your Question:

Lastly, some people say I could clamp it in a jig that twists it in the opposite direction and leave it sitting there for a while - but others say that doesn't work.

This is it in a nutshell: it most certainly can work, but it hasn't for everyone who has tried it. Well yeah, of course not.

The people who are saying it doesn't work are expressing themselves poorly if they state it that flatly. Presumably they are speaking from experience, and while they mean well they're simply not accounting for just how variable things can be — just the wood itself, no matter how predictable it might look, can behave very differently to other similar-looking wood. And there are a host of other variables too naturally.

Or if it would be less work to just build a new door.

Even if remedial action doesn't work I don't think it could realistically be said to be less work to build a new door, unless you have milled stock ready to go already.

If you don't, as you know, you're looking at roughly cutting one or more boards to length, jointing each piece x2, planing to thickness, ripping to width, jointing or sanding that sawn surface, milling the groove and so on. That's hardly less work than putting something in clamps for a day and seeing if it fixes the problem!

I don't know about you but I'll take a day of waiting, after a completely trivial amount of effort, over doing all the steps to make a door (even a simple door like this one) on the assumption that the fix won't work.

If the twist isn't removed or lessened enough to make the door usable really all you've lost is a day.

A few things that might be of help:

  • Heat all four joints up well before clamping or attaching the bowed cleat, and then heat some more after you've put the door aside. All the glues you're likely to have used are softened by heat, with the possible exception of foaming polyurethane. If you don't have a heat gun this is still worth doing with a hairdryer on max.

  • If you glued up using any type of PVA (white or yellow) they are softened to some degree by acetone, so applying a little here and there may also help a little. If you glued up using hide glue (any type) use water instead; do not use vinegar unless you want to try to ease the door apart into its component pieces!

  • Again if you glued up using PVA it's also suggested by some sources that dampening the joint area with water can help, even though this obviously has little to no effect on the PVA directly; can't hurt to try this with other glues as well.

  • Be patient, don't unclamp after just a few hours. I'd suggest waiting overnight at least (10-14 hours) and a full day wouldn't hurt.

9
  • Thanks, that helps. If it makes any difference, the glue I used was Titebond Ultimate. I already put stain and poly on the door, so any solvents would likely damage the finish. But I'll try the heat gun. Oct 16, 2022 at 4:01
  • And it's 3/8" not 3.8". If it was twisted by 3.8" it would no longer be a door, it'd be a sculpture. Oct 16, 2022 at 4:02
  • "And it's 3/8" not 3.8". If it was twisted by 3.8" it would no longer be a door, it'd be a sculpture." LOL, simple typo — the / key is right beside the . key on my keyboard ^_^
    – Graphus
    Oct 16, 2022 at 12:34
  • Would 12 - 24 hours really be enough time to see results? I would have thought it would take much longer than that. This inspires me to give this a try should I ever manage to build myself some warped doors.
    – FreeMan
    Oct 18, 2022 at 16:15
  • @FreeMan, well it depends, and that is the check it and see time anyway — if it's done something one can always reclamp & try for longer. Worth remembering that recs on this front often do say to counter-clamp (or whatever) for "a few hours" & that's without one or more help-along tips like dampening. I'd have thought that if one is only relying on the wood relaxing a bit due to some cycles of moisture uptake & loss, and/or the glue bonds shifting slightly under strain, then obviously it's worth waiting a lot longer, probably a week +.... which wouldn't be fun given the unsure outcome!
    – Graphus
    Oct 18, 2022 at 17:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.