Hi,
I’m coming to the end of a long kitchen project. All of my doors and drawer-fronts are flat frame and panel. I just installed the doors, and to my horror discovered that a number of them are pretty badly twisted. Is there any good way to get the twist out? Unfortunately they are already painted. The frames are 3/4 x 1 1/2 poplar, and the panels are 1/4 birch ply. The corners are joined with slot mortises. It is possible that some of twist comes from the mostises or tenons not being exactly centered on the rail or stile. Eyeballing through the paint they all look OK, but the margin of error in my tenoning jig was probably as much as 1/16.
Mike
Replies
Are the ply panels glued in place? What paint did you use?
Ron
Thanks for all of the advice. I figured the doors are toast, but thought it worth asking. I suspect the cource of the problem is the joinery. Due to the limitations of my shop and equiptment, I think in the future I will miter and spline the corners -- there's a little more magin for error that way. In response to Ronaway's question, I primed with oil-based Kilz, and finished with two coats of California Paints Semi-gloss Latex.
The simplest and best thing to do is somehow overlooked in almost anything I've ever read or heard in making frame and panel doors. Dry fit the frame without the panel, and lay it on a true flat surface (your workbench) and see if it is flat. If you are not sure how to know if your bench is flat, then you need to go thru that exercise first, before gluing up any doors. When gluing the frame and panel doors together, use the flat surface and block up with two sticks of the exact same size to elevate the assembly equally off the flat surface. As the glue sets, insure that the door assembly remains in this flat position, stacking the doors with equal size spacers to accommodate the clamps.
Frames will not be flat due to out of square stock, out of parallel mortise and/or tenons, loose joints, clamping in a non flat position, or wood movement (changing MC after assembly).
Dave S
http://www.acornwoodworks.com
Dave,Good advice.Beautiful work on your website.Rich
The only doors that I've had warp were made from poplar and had a plywood panel glued into them and to maximize the problem they were finished with a waterborne paint type of lacquer. Even at that only 2 out of the 6 warped. If I had it to do again I would seal the doors with shellac before applying the waterborne product.Ron
A common question, and so far as I know there is nothing that you can do that will flatten out the doors. The usual suggestion, that you clamp the panels down stressed in the opposite direction, can sometimes flatten the door out, but the effect is almost always very temporary.
There are two likely causes for your problem, although there are other possibilities. One is that the joints weren't cut quite square, the slots and the tenons weren't quite parallel to the length of the frame pieces. The second possibility is that the frame stock has warped, it should all be carefully selected quarter sawn wood. In either case there is nothing that can be done to correct the problem once the door is assembled.
The one possibility, but it is a slim one, is that the frame has warped because of moisture absorbed from the paint and the doors will flatten out in a few days as the excess moisture evaporates into the air
John White, Shop Manager, Fine Woodworking Magazine.
I don't think that using quartersawn material for doors is always possible, practical, or necessary, although yes, it is the ideal. In fact, it would be best used for everything built of wood, I think, because of the stability of it.
I have found when asked by others about this issue, that the problem can sometimes be caused by material that was not flattened to begin with. I am not saying that this is the case here, but I know that many woodworkers buy material that is already surfaced and seemingly ready to use.
This material may seem flat and straight initially, but when cut into small boards, it will often twist, curve, or warp. A flat door can not be made of crooked boards.
I find that it is often best to cut lumber into rough sizes, just 1/4" wider and thicker than the finished product.
Then: flat joint, plane, flat joint again prior to reaching final thickness, plane to final thickness, edge joint, cut to width 1/32" wider than the final size, cut to length, and joint second edge.
Hal
http://www.rivercitywoodworks.com
Hi FM,
I tend to get wordy, but here's some things that may help you get flat door panels. I learned these lessons recently when I started to make my first panel doors as well.
In addition to what Hal is saying, I was told that if your finished stock isn't dead flat, your rails and stiles aren't exactly the same respective lengths, and your mortises aren't exactly parallel to the faces of the frame members, your doors will not be square and flat.
At first, I didn't want to believe what the local woodworkers were telling me about the absolute importance of having access to a planer and a jointer. Too expensive! But I found that while I could make a nominally decent looking door without them, I couldn't make a consistently flat door without them.
So, I found a good used jointer and another used planer and rebuilt them so that they would prepare stock like the books say that they should. Then I spent some time with pine 2/4's learning how to set-up and use them.
In order to make doors that wanted to go together square, I learned to cut my rails and stiles using stop blocks so that I could get rail or stile pairs that were exactly the same length.
In order to make doors that wanted to go together flat, I made about 4 versions of a mortising jig to use with my router before I could consistently make mortises for floating tenons that were always parallel to the stock's face, and always registered to what would be the finished door's face. For me, the mortising jig is the key! There's lots of plans out there for mortising jigs. Pick one that appeals to you, build it, learn to use it, and then refine it so that it's fast to set up and use.
After all that, in order to make doors that stay flat and square while the glue drys, I made a panel-door assembly and clamping jig that I found posted somewhere. It's simple in concept and easy to make, and after I used it for the first time, I was glad I took the time to make it.I don't consider this a production tool because it only does one frame at a time, but it does it darn well!
The assembly/clamping jig is simply a flat (clamp it to the workbench) 30" square, 3/4" thick panel of masonite with a jointed, 1/1" hardwood lip glued and screwed to two edges at 90 degrees to each other, and finished with paste wax to keep the glue from sticking. If you build this, leave a 1/4 inch gap at the corner where the two lips come together, or you'll have problems getting the sides of the frame to register on both at the same time. Glue up the frames in the glue up/assembly/clamping jig. The 90 degree corner formed by the hardwood lip makes it easy to keep the frame square, the flat panel makes it a snap to keep the frame flat while the glue sets up, and the clamps are easy to apply because the whole thing is held flat, immovable, and accessible on my workbench.
I hope this helps you up the learning curve.
Mike D
Why isn't it possible? Every log, no matter how you saw it, will have at least a few quarter sawn boards produced, so why not pick them out of the pile and use them in the most critical applications? If you aren't getting quarter sawn wood in your purchases, somebody else is picking through the pile before it gets to you.
Why isn't it practical? Once you have the quarter sawn stock it cuts and joins just like flatsawn.
As to being necessary, I think that it is. The problems of dealing with a warped door are way more than the small amount of extra time spent sorting out the quarter sawn from the wood pile before you begin.
Also, quarter sawn looks better in a door's frame and forms a nice contrast to the panel which can be flat sawn and chosen for its pattern.
There are a lot of reasons why a frame may not turn out flat, but starting out with poorly chosen frame stock can cause problems that no amount of careful stock preparation and joinery can overcome. Why go through all the trouble of building a perfectly flat frame only to have it possibly warp later?
John W.
John,
Well now we know who has been cherry picking all of the quartersawn material out of the units!
I hope I don't come on as being argumentative here, but I have to dissagree with you on a couple of things.
There are millions of frame and panel doors have been made of material that is not quartersawn. It is not necessary. Reaction wood can never be used though. It will always warp.
As far as whether quartersawn material looks better or not, and whether or not the panel should be flatsawn is a design choice. You may like it, someone else may not.
For example, I once used quartersawn zebrawood for panels and they looked great with the flatsawn padauk frames and ebony raised moldings that matched the rest of the cabinet. Wouldn't have been as good the other way around.
I have also built cabinets of all rift white oak that would have looked weird with quartersawn door frames and flat cut panels. I had no warped doors though.
The reason I say that it isn't possible is that if every manufacturer & cabinet maker did what you are saying, there would be no quartersawn material for you or me to use. It would be presold to the larger manufacturers, and all of the doors on the cabinets at Home Depot, Lowes, etc. would have quartersawn doors.
Cutting quartersawn material specifically out of a log is considered a wasteful way to cut lumber, which is why all boards are not quartersawn.
Hal
Wood warps due to changes in M/C, period. It does not have some magical, inherent sense of "wanting" to warp, with it being our duty to overcome this with will, technique or quartersawn. Reaction wood is the only exception, and should always be avoided except at the woodstove.
Quartersawn earned the"stability" label due to its half-as-much-shrinkage-across-the-width nature, not a natural abilty to resist the rumored tendency for all wood to warp. And to take it a bit further, all wood has quartersawn grain in it, sometimes it matters more than others due to orientation in service.
Start flat and stay flat is rarely taught. Jointed, flat, square stock is essential. But so are flat work surfaces or reference points.
Dave S
You are wrong about this, quarter sawn wood will stay flat across it face as it gains and loses moisture, flat sawn wood won't, it will cup.
The cupping is the result of the difference in the tangential shrinkage versus the radial shrinkage, it is caused by the basic structure of the wood. Some species cup more, some less, depends on the T/R movement ratio. A flat sawn board, if it cups, will probably distort a frame so it is best avoided whenever possible.
In addition, as you note, quarter sawn wood shrinks less across its width compared to flat sawn, so less allowance needs to be made for shrinkage and expansion of the frame in the opening it fits into, allowing for tighter fits that still won't bind with humidity changes.
John W.
I'm a little confused. Does the quartersawn imperative address "possibilities" or are we addressing normal results?
I've always assumed that if a frame and panel door is made "right", i.e., constructed of properly prepared stock, joints per the book, rails of equal length, stiles of equal length, etc......and then glued up and allowed to dry flat on flat surface, then perhaps (this being wood after all) the occasional door may "twist", but that's not likely and it's not often.
Is this right or is this wrong? If it's right, is it fair to say that going one step further and using all quartersawn wood eliminates the few instances "not likely and not often"?
Mike D
Okay everyone, settle down.
These stiles and rails that we are talking about are only 1 1/2" wide.
It is obvious that the guy either had twisted boards or crooked joints at the time of assembly.
Case closed. Whew!
Hal
OK, It's morning, and I'm calm now.
What Telemike says makes sense to me as well. I've been lucky so far, but making sure that you don't introduce the doubling effect into machining errors is something that I'll think about the next time I make anything.
Mike D
You can trim the tenons until the thing lays flat if the error was caused by machining. You'll probably need to use thickened epoxy to glue it up once you're done with this exercise.
If you slide the tenon into the mortise very slowly and watch the reaction vis-a-vis a flat surface you can figure out where the twist is and the direction of the twist. Do it one corner at a time and when you're done it should go together flat, although the fit will be gappy.
Edited 11/8/2006 10:43 am ET by VeriestTyro
Mike,
Quarter sawn wood stays flat as it gains and loses moisture. Flat sawn wood will cup at least a little bit and some times a lot depending on the board's grain and the species of wood. If the grain of the wood runs a little off from the long axis of the stock the cupping will turn into a twist. These are both "normal results", not an occasional odd happening in a especially troublesome board.
If you want to avoid even the chance of the wood cupping and distorting the door frame, the safest route to go is to use quarter sawn. Being that the frame will also usually look better and have smaller shrinkage in width there is little reason not to use quarter sawn whenever possible, and there should be some in every wood pile.
For a short explanation of the way that wood moves when it gains and loses moisture read the booklet that comes with the Wood Movement Reference Guide from Lee Valley. For a long technical discussion of the process, Bruce Hoadley's "Understanding Wood" is a good reference.
John W.
Edited 11/8/2006 6:11 pm ET by JohnWW
Thanks John. Great explanation.
Mike D
John,
While I agree that quartersawn stock makes good door frame material, it is by no means completely problem free. The growth ring orientation that keeps qtrsawn material flat on its face, transfers more movement to its edges. That is to say, I've found narrow quartersawn stock, as you'd use in door frames, to be as prone to longitudinal bowing (edgewise) as flatsawn stock is (looking down its face). It is as important to let quartersawn stock "relax" after roughing it out as it is for flatsawn, allowing plenty of extra width for straightening after any sideways movement takes place. Especially an issue if the doors are tall, and/ or the quartersawn material came from near the very center of the tree, as one edge of qtrsawn stuff nearly always is. Also, proximity to a knot will telegraph across the whole width of qtrsawn stock, with attendant warpage problems, perhaps worse than in flatsawn where any distortion will be localised more or less perpendicularly to the face, rather than parallel to it.
In my experience, it is more important that stock chosen for door frames be selected for mild grain, than that it be either flatsawn, riftsawn, or quartersawn. By mild, I mean straight grain, no knots or reaction wood, with little runout end to end, without transition from heart to sapwood, and not from near the center or pith of the tree.
I bet we'll be in agreement that it's a good idea in any case, to get out extra stock, and let it rest as long as possible, before final sizing. This lets any instability show up, to be culled out before assembly of the door.
Regards,
Ray Pine
Ray,
All of your points are valid, important and very clearly explained. Isn't it is amazing how many things need to be considered when "simply" choosing a board for an application?
I also always try to cut extra stock and give it time to settle down before final joint cutting, but once the joints are cut I try to assemble the joints as soon as possible.
John W
John,
One shop I worked in, the foreman got the idea to stockpile furniture components. We would typically build up 2 or 3 chests, when an order came in for one. He figured there were savings to be had by cutting 6, building 1 or 2, and stacking the remainder on shelves, to be assembled "on demand" at a future time. They would take up less room, not get shopworn in the showroom, and save on set-up time in the shop. Unfortunately, the time savings in the machining, was offset by the hand fitting of all the joints that swelled, or shrank, on the shelf. Assembly was problematic, to say the least.
Actually, just a day or two will affect the fit of a mortise-tenon joint, if the weather changes.
Ray
John,
That's a pretty good explanation. You have done pretty well at not letting me get to you. Good for you. You are very knowledgeable and I always enjoy reading your posts.
But isn't it possible that the grain can run off the long axis of the stock in quarter sawn lumber as well?
I know I sound like I am giving you a big hassle on this, but I have seldom gotten quarter sawn boards that are perfect from end to end.
The big thing for me about quarter sawn lumber is that it shrinks and swells less than flat sawn lumber. Because of this I like to use it for wide panels such as table tops. But even here, the problem can be one of swelling and shrinking in thickness which can also cause problems if the heart edge of one board is glued adjacent to the outer (sap, minus the sap) edge of another.
I make all of my apprentices read Bruce Hoadley's book "Understanding Wood," and test them on it. It is still, after all these years, great information for anyone that deals with wood whether they are buying it, selling it, or working with it.
Here's one for you. If we are going to talk about the ideal wood for these stiles and rails, then shouldn't we also recommend that the only wood that can be used is Genuine (Honduran) Mahogany since it is probably the most stable of all wood, quarter sawn, of course.
Hal
http://www.rivercitywoodworks.com
Hello John ,
First off , imo Q sawn is beautiful and I would strongly agree that it is typically stable and flat across the grain . I have built miles of mostly Q sawn Oak case work out of this wonderful grained wood and it is one of my all time favorites to use .
It is not always practical nor wise to use Q sawn in all applications for your , mainly because the grain pattern will not match the other non Q sawn materials in the job .
I have been in business for over 25 years and have literally built thousands of frame & panel doors , the majority of these door frames were not made of Q sawn material . Honestly I could count on 1 ( one ) hand the doors that were warped enough to re build , I know most folks would agree that is a safe margin to work with and less than a gamble . So Q sawn is clearly NOT necessary .How much cup can there be in a door stile 2-3" wide ?
As far as Q sawn looking better , it really depends on the application . On one job the client complained about several Q sawn parts that were used and thought they were defective and they did not look the same as the rest . So beauty is in the eye of the beholder .
I am surprised that with all the contributors of this thread there has been no mention of sighting down the side of the doors while in clamps and tweaking them to be flat. I have found that they will stay flat when they set up in a flat plane .
If using Q sawn is your preferred choice then that's great , but to suggest it is almost mandatory for a flat door is a misleading and inaccurate statement that can cause much anxiety especially for a learning wood worker . Perhaps you would better serve the people on the learning curve by toning your statement down a tad .
dusty
Dusty, reading your message has prompted me to add my sixpence worth:
I'm having some trouble understanding why there is skirmishing over small frame and panel doors, painted , for kitchen use of.
The twist/warp: most likely cause in this case is edges being planed out of square to face, then the problem can be worsened by incorrect clamping. On the other hand it can sometimes be rectified by compensatory clamping especially if the frame stock is softish wood-not a good practice but it happens.
How wide are the rail and stiles?? Not wide at all I would assume-something like 2 inches or so? So how much cup can a pessimist expect from a measly 2 inch width-or even a twist? Or even a combination? Not a lot to be sure-certainly nothing that can't take a bit of planing out.
As for the necessity of q/s stock for rails and stiles- it is absolutely not necessary when we talk of a) narrow widths and b) painted kitchen stuff. If one is making long cabinet doors in specialised circumstances such as air tight book cabinets etc then one just might have a case (sorry bad pun there), but I believe it more important to select sound, acclimatised stock with no trouble potential such as knots and crazy grain, and no mixing of q/s and f/s.
Q/s really looks good when one is talking of certain Oaks , Beech etc when quarter sawing exposes the medullory rays-so from an appearance point of view there is a case, but not for painted kitchen cabinets.
Philip Marcou
Phillip ,
Perhaps I missed Mr. Whites reference to painted work , but his inference that QS MUST be used is simply overstated . If you need a micrometer to measure the cup , warp or twist and you can't measure it on a tape measure then in reality what difference does it make ?
Your comments and understanding come truly from hands on experience and real world work scenarios as mine do , thank you I respect and appreciate your input .I think your observation of face to edge being less than square was largely overlooked by most of us as well .
regards dusty
You must lay the stiles and rails on a flat surface in their correct orientation to make sure all the parts are lying in the same plane. If they are twisted up off of a known flat surface then don't even waste your time cutting the joinery on the boards as the only way to make the door lie flat would be to cut twist into the mortise and tenons themselves in a fashion what would offset the twist in the workpieces. Of course, this would be a fools errand.
You need to locate relatively straight grained stock during the rough milling process and set it aside for the door parts. You don't want 'interesting grain' for door stiles and rails. You want bland, straight stuff if not quartersawn as somebody else mentioned, although spec'ing QS lumber for every cabinet door you make shouldn't be necessary. But it's never a bad thing to have on hand.
It also helps to mill extra parts in case a stile or rail proves unruly.
If you are doing serious woodworking then you will finish mill your own stock from roughsawn lumber as you need it. I can't imagine the frustration of trying to find flat, S4S stock at your dealer and hoping it will stay flat when you get it home. You have to have the ability to four-square it yourself.
Edited 11/6/2006 10:29 am ET by TaunTonMacoute
As JohnWW said, your doors are probably kindling. However, when you make new ones, don't repeat the error -- allow for a smidgen of error in your joinery by making your joints just a tad loose and dry fit them on a flat surface before you glue. Fix any problems and then glue, again on a flat surface.
Hey, look at it as valuable experience! ;-)
Mike Hennessy
Pittsburgh, PA
Forgive me if this message is a little esoteric. I've also run into this problem. Ultimately, the warped mortise and tenon doors were the fault of my mortiser technique. I have a cheap far eastern mortiser that I've used for mortises for about 20 years. When I made a batch of kitchen cabinet doors for this house, each end of the stiles had symetrical mortises so I put in a stop block and flipped the boards end-to-end to cut opposite ends.
After some examination (and cuss words), I discovered that the head of the mortiser had a small amount of deflection under pressure (when cutting a mortise) that led to the mortise being at a small angle. When I flipped boards end to end, it doubled the angular error. When you do this flip the board one set up approach, the doubled error always warps the door because two diagonal corners have matching angular errors that are opposite the other two diagonal corners.
Now, I cut mortises into each end of a stile without flipping the board even if I have to break the set up and set it up again to cut the second mortise on each style. The angular error is small and angles the same way on each end. The error is usually not noticeable when I cut the mortises this way.
This problem can apply anywhere where the mortise is not exactly 90 degrees to the edge of the board - that is, almost anytime you cut one whether with a mortiser, a drill press, or a router. Next time, cut all the mortises on one end of the board with the first set up, then the others with a second, opposite set up.
This explanation is a little hard to understand, but it made all the difference to me.
I guess the cardinal sin is not catching the problem at the dry-fit stage, no? Gluing and hoping is not a route to success in the shop.
I guess I'll add my 2 cents here,as I've experienced the same problems. First, your wood must be the same moisture content when building as when hung.Wood will always move when you cut it. This usually means drying longer than you would like(usually 3weeks to a month), all depending on your shop humidity, as compared finished product location. The next important factor is joints. If your joints are not perfect, then when you clamp for gluing, they will twist slightly to match your cuts. If you clamp the doors pefrectly flat, first,and then clamp the joints not quite as tight as you would like, they will dry flat with slight gaps at the joints. If you are painting the doors, these gaps can be filled and hidden. If you are not painting, it will be a toss up, tight joints or flat doors. The key to all of this is dry wood, and proper joints. I work with locally grown and sawn woods, and don't have the option quartersawn or whatever, and still make it work.
This forum post is now archived. Commenting has been disabled