Troubleshooting Warping in Veneered MDF Doors

A cabinetmaker gets advice about a set of bath cabinet doors that has cupped after finishing. February 15, 2015

I'm just about finished up on a nice little walnut bathroom project when I see a problem starting with two doors! These two doors in particular cannot be replaced and could make things go badly if they don't stay flat, but let me back up and start from the beginning.

There's a total of 16 - 3/4" MDF doors on the project. Most of them are in the range of 14"w x 44"h +/-. They're factory laid up sequence matched walnut veneer both sides. I banded, sanded, stained and top-coated with two coats of MLC Duravar same exact treatment both sides. I installed the doors about two weeks ago and except for two they're all fine. It's two particular doors though that have started to develop a slight bow outward away from the cabinet at top and bottom! These two particular doors are on medicine cabinets that flank the vanity. They're very shallow cabinets that by their nature are "sealed" a bit more so than the rest of the cabinetry (standard Euro style wardrobes and vanity). My current thinking is that now that we're in heating season the dry air is affecting the outside faces faster than the inside. I figure the other cabinetry may have more/easier air transfer in the gaps between doors, and that the med cabs are a bit too sealed?

I opened the doors today and will leave them open for the rest of the week to see if they "balance" out. I am open to ideas suggestions to see if I'm going in the right direction. These cannot be replaced easily as the grain is matched in the panels above and below. There are switches and outlets in both lower panels and a soffit connecting them at the top! So these doors must stay flat! Again, the other 14 doors which are all the same batch of material, were all made the same way, and are slightly bigger, have all stayed flat?

Forum Responses
(Architectural Woodworking Forum)
From contributor H:
I think you are on the right track by opening and leaving them open to see if they acclimate. I bet they will based on your description of construction and finishing. Your finish might not be fully cured in only two weeks. I use a plant on finished end on euro cabinets which would make this sort of situation much less apparent as you can't see the door edge.

From Contributor E:
If you're in a position to remove the doors and lay them flat with the convex side exposed to the heat, you'll be able to straighten them much faster.

From contributor M:

I can't say we've had problems with MDF doors after finish. I do know we've seen some crazy movement before finish, caused by temperature and humidity swings. One case where we had doors done, ready to begin finishing, finding them warped like crazy, at the end of an especially hot day. Went home feeling totally bummed out wondering what we're going to do about it. Came back the next morning, they were straight again. We've also had luck on straightening MDF doors by bowing them in the opposite direction. How much bow and the length of time is somewhat trial and error. I'm optimistic you can get the two doors to work.

From contributor C:
I have seen this before. Your fix sounds like it will work. Laying them flat with the inside facing up towards the heat source is a good idea.

From the original questioner
So I'm a bit more optimistic this will work. I could remove the doors if needed but I probably won't be back in town until the end of the week anyway, so we'll see how things look then.

Contributor H, these actually do have finished ends but since they're almost flush with the face of the door they won't hide much. It's a tricky design to explain and I couldn't get a pic since the electrician hadn't got the lights wired up yet. Everything is contemporary Euro style flush all around, which looks great if everything stays flat.

From contributor N:
Iím with the others laying them face down, but I would take it a bit farther. I keep a little space heater out in my shop, which I use to blow warmed air into a tent, using a tarp with an opening cut into it to match the heater outline. It is quick to set up, stores easily, and can get up to 150ļ inside, which will speed up the process.

From Contributor V:
Are these the only two doors that are up above counter level? If so you might be having a problem with steam from a shower.

From contributor N:
To contributor V: I think they are going the wrong way for steam. I was thinking there is probably a register blowing more air on this pair than the others.

From the original questioner
There is no steam presently, or plumbing in general. This is a second, (master), bath so the clients are using the other bath while this one is being done. As far as the heights the wardrobe cabinets go floor to ceiling so those would be the highest reaching doors. No direct heat at this time either, just the heat from the unit, which is about 71 deg F.

From Contributor V:
It might be interesting to check the humidity level inside that vanity cabinet as vs. inside other cabinets and the room in general.

From Contributor V:
I wouldn't be able to do that now as I've left the doors open for the time being. I'll be back there at the end of the week to check progress and hopefully install the last cabinet Iím waiting on an electrician. I think if I were to try to check it I'd have to leave the gauge in there for a day or more then come back and check? I'm not sure, I'll have to mull that over a bit.

From Contributor O:
Radio Shack has a temperature and humidity gauge with two channels so you get a reading where the unit is, and another off the end of an 8' wire. Usually meant for inside outside temps, it is handy for inside a cabinet and outside a cabinet. It also records highs and lows. But I'm betting you will go back and the problem will have solved itself, once everything has settled down and acclimated.

From contributor M:
If the medicine cabinets are in the wall, then moisture from inside the wall might be keeping the back side MC higher. New wood framing? No vapor barrier behind the cabinet? Just a thought. Even if you have to fix the doors, you can still keep the veneer intact, so don't sweat it.

From Gene Wengert, forum technical advisor:

It is surprising how much a thin piece of veneer can move a large panel when the MC changes. MDF has very little movement when the MC changes, so cupping is more likely due to the veneer movement. Now, MDF would cup lengthwise and widthwise, while the veneer (unless lots of cross grain) would only cup widthwise. With some veneer from a tapered log, we get grain angles that result in twisting. In this case, it is important to have the veneer on both sides "opposite faces out" so that the twist in one is cancelled by the other sheet.

Because this happened quickly, we also know that quick changes in MC are much more likely to cause warp than slow changes, even though the total MC change is the same. So, the good news is that if the doors straighten, then this phenomenon will probably not happen again. The bad news is that the MC when you brought the doors to the home was not in equilibrium with the home, your shop was more humid perhaps or the wood was stored before you made the doors in a more humid location. So, it is this first MC change in the home that is large and is causing the problem. The RH sensors mentioned take about 20 minutes to give you a valid reading (unless they are cold before you bring them in). So, check the RH in your storage area, in your shop and in the home several times day and night.

From contributor N:
I'm not sure there would be any way to check this out, but just one more potential cause. In my area, my supplierís warehouses have large overhead doors which are always open, so exposed to outside atmosphere. While the bundles may have cover sheets, they may not always put that back on top after they are busted. So when I get to be there to pull my sheets, I always decline taking the top sheet if it doesn't pass my inspection for flatness. I'm wondering if the shop that laid up the veneers, ran a top sheet which had a high MC on one face might be causing this now that the substrate has reached equilibrium?