I am planning a cabinet with two frame & panel doors that would ideally be approximately 6’0″ tall by only about 1’0″ wide. The Frame would utilize 4/4 cherry stock (3/4″ thick) and 1/4″ cherry ply as the panel.
I would like the frame to be this long and narrow WITHOUT intermediate cross bars. Initially I wondered weather the frame this length might flex or otherwise contort and more recently I wondered if I could simply glue the plywood panel to the frame?
I’ve always understood that panels should “float” in frames like drawer bottoms but plywood is pretty stable…..
Is my thinking right or wrong?
Any experienced help would be gratefully appreciated!
Replies
You can safely glue the plywood panel in the door. You would get more strength with a 3/8" panel...
You are correct that the plywood will be stable, however, the solid wood stiles and rails can move. If you are using a bead and cope router bit set up for the joinery, the plywood may not fit the groove very tightly. Gluing in the panel isn't particularly easy, given the way the joint may fit. Since you would have to glue to the face of the frame, in a groove, this will be a possible weak link. Movement may fracture that weak area.
I think the more important part of the construction would be selecting good flat straight stock for the frame. You need good joinery with 90° edges along with proper glue application. Making sure that everything is flat, straight and square during glue up makes all the difference. You will have plenty to do just getting the frame glued and clamped up. I don't think gluing the panel will add anything other than introducing other issues. I know I'd just make a big mess and if the panel didn't fit just right...it could get nasty.
Beat it to fit / Paint it to match
ello Lad ,
72" tall with no cross mid rails is pushing the envelope imo , Could you make the end rails large ? Can you make two seperate doors with no cabinet face between on top of eachother ?
Industry standards will typically not glue panels in wood frames . I have made a few thousand F&P doors and have never glued a panel in , and have never had to replace a door because of excessive warpage , honestly !
Most of the case work I do gets 1/2" Apple ply drawer boxes and , Yup I glue the bottoms in , I've never had a come back on a drawer box , except one that had water stains on the finish from the sink .
Choosing the right stock ( flat and straight ) is really key to your success on these door frames . Wood will move , design with that in mind and your work will stand the test of time .
good luck dusty
Lad, there is nothing much if anything to be gained by glueing the panel in. Go for straight grain stock for the frame, quarter sawn if the look will take it, but don't mix quarter with flat sawn. Obviously well seasoned stuff is called for, preferably mortice and tenon joined with generous haunch especially on the bottom of the lower rail and top of the upper rail.
Leave minimal clearance all round for the panel in the groove as there is only the seasonal movement of the frame members to take into account.
If the look can take it the greatest aid to rigidity will be if the design can take either top and bottom rails as wide as possible or at least the bottom rail to be wide as possible relative to the width of the stiles.
Assuming you can get a reasonably good fit for the plywood I think you would gain substantial stability from gluing in the plywood panel. But the advantage of virtually eliminating racking has got to be great. When glued, the panel is essentially similar to diagonal braces running from the top corners to bottom corners.
Sure, the rails and stiles can move, but with stiles only a couple of inches wide the amount of movement will be pretty minor, perhaps 0.026" since it is divided among both sides of both stiles. (And this assumes central heating in winter in a house that't isn't very tight and no air conditioning in summer, as well as having the wood unfinished.) Stresses ought to be less than the typical mortise and tenon joint, which after all is a cross grain construction, where gluing the plywood panel in grooves gives a glue joint more similar to long grain against long grain.
With just a little dry fit "tweaking" to ensure no surprises when you pull the frame together I can't see how the glue up would be a problem.
Hi Steve ,
Whether many know it or not , even plywood has a moisture content ,while certainly more stable than wood it does exist .
The theory of the floating and stable frame and panel is centuries old , this is not a hunch but more of an accepted practice . Do you make F & P doors and such ? If so do you always glue in your panels , or are you saying because of the very tall application of the OP he should be able to get away with it ?
I consider you a true Master finishing Guru , but , your advice seems contrary to that from professionals on this issue .
It would be interesting for other pros besides the few that already have , to chime in and tell us if they glue panels in .
dusty
The "accepted practice" derives solely from solid wood panels, and doesn't take account the properties of plywood that makes its expansion and contraction more like lengthwise solid wood, not widthwise, whether radial or tangential. Why allow for the change in dimensions of a panel that won't change in dimension?
Industrial or commercial practices may not add the glue even with plywood panels but I think that's a cost savings matter, not a "it wouldn't work" matter. For doors in the normal range of sizes it's not needed, but for this tall door there would be a real risk of racking that could be eliminated by gluing the panel.
Steve,
I agree that there is much to be gained by gluing the plywood panels into this door, nothing to be lost.
Frame and Panel construction came into being before plywood existed. It was (is) the answer to the expansion and contraction of a panel made from solid material. The panel is free to move in the frame's grooves. Since there is ample strength provided by the frame (in most frame design) the panel is not needed (and can't be used) as a structural member.
This door is a frame and panel in name only and uses the style for aesthetic reasons only. There's no need to worry that gluing all the parts together is a problem. Just the opposite, the opportunity to do so solves other problems.
In this door, the frame strength is marginal. Since the panel is as stable (more so) than the frame itself, the whole thing can be glued up as a unit to gain stability.
Rich
Rich, what you are saying sounds logical, but in practice things may be different. I have seen a few examples of house doors and ordinary cabinet doors (kitchen) where there was no clearance- the frame had nowhere to go and there was both distortion and cracking. Glueing the panel implies a close fit between panel and frame, otherwise it is not effective. Furthermore, if the glue holds on the edges of the groove, and the rail shrinks, there can be splitting, especially if there is moulding which reduces wood thickness in the area.
For the doors Lad is contemplating frame strength is not that important. What is important is for the timber to be sound, and the dimensions of the rails and stiles to be appropriate.The frame should not be relying on a ply panel for rigidity-that is the job of proper mortice and tenon work.
Maybe Lad should glue one door and not the other- if he lives in a climate of extremes we await the (inevitable) result (;).Philip Marcou
If gluing this panel will cause the rails to split, then every glued mortise and tenon joint wider or deeper than about 1 1/2" is doomed to split too, with the cross grain movement of both parts creating stresses. I can't think anyone would consider M&T joints of such modest sizes to be even the slightest bit unsound. But there is no greater stress created by gluing the plywood panel to the frame. It shouldn't be unsound either.
You can't glue a solid wood panel--that would show significant change of width over 10", but the plywood's dimensional change over such a length is almost nil and that makes all the difference.
Without gluing the panel, the structure is a frame 6' long and 1' wide, made of something like 2" wide stiles into 3" or 4" rails on top or bottom. Racking forces would be considerable, as there is little to resist downward sag of the outside of the door frame. Glueing the panel substantially resists the sag. I can't help but think that risk of failure of the joints through racking the door frame would be much greater than the risk of failure through splitting the frame.
Amateur here. I vote with Steve--use glue.
In the mid 1970s I built two (2) tall, narrow cabinets with doors. The cabinets are probably 7 feet tall by about 4 feet wide. No one told me any better, so the door corners are mitered, the plywood is glued into the channels in the frame--all the way around. I didn't have clamps, so I wrapped the outside perimeter with a few turns of string, and then slid small blocks of wood under the string, toward the corners to clamp the miters. I put finish on one cabinet. It started its life in Virginia Beach, air conditioned house, then moved here to Washington state, spent four or five years inside, then went to the garage. The unfinished cabinet has had a similar life, although it has been stored and not used since the early '90s. Neither one suffers from sag nor splitting. I used fir for the frames, which is similar in expansion ratio to cherry, and 1/4 inch ply for the panels.
Another thought is that manufactured floor I-joists have a manufactured material continuously glued in a groove in natural wood for lengths often exceeding 20 feet.
Of course, highly efficient, rigid joints can be obtained by bonding plywood to itself or to heavier wood members, such as those needed in prefabricated wood I-joists, box beams, and stressed-skin panels. --Wood as Engineering Material
Comparing a M & T joint to edge gluing of a panel and frame may not be a suitable or similar example in general .A stile is all long grain , going in the same direction , not so with a M&T .
You are under estimating the amount say a 2 1/2" stile can and will move , a 1/16th would not be uncommon or excessive for each stile's movement in it's life .Regardless of where you got your calculations from , mine are from making thousands of F&Ps .
In this case using wider rails top and bottom and creating sound joinery ( whatever type used ) is really the most important factor in keeping this frame flat . The structural integrity in F&P comes from the joints not the panels .We all agree that careful selection of the frame material is also key .
It has been pointed out also that some stile and rail details on the inside of the frame is shaped to a much thinner profile at this juncture . To glue the panel to a frame that moves can create tension and cracking and the racking effect you try and avoid may be created .
Using the proper amount of hinges will also help this frame .
I would use at least 2 1/2" wide stiles and as wide a rails the design will tolerate . Really this is a poor design , a center rail would solve this dilemma before it becomes a problem , I think most of us will agree on that .
dusty
"Knowledge without experience is simply information"
Now I am confused. The M&T with cross grain has more, not less, stresses to cope with than the rail to panel joining.
The total width of a 2 1/2" rail that is exactly tangential grain might change 1/16th of an inch in the very worst case. However, since the rails and stiles are completely unconstrained on the outside dimensions, much less than 1/16th of an inch needs to be accommodated in the form of compression of the panel. If it weren't for the rails on the ends the width could change entirely with no stresses being created within the door. Putting a plywood panel between the stiles is just the same as putting a center rail in place. If that's not a problem then gluing in the plywood panel isn't either. Remember, the experience with solid wood panels isn't relevant here.
The worst case will occur only if furniture that has been significantly dried in a house that stays at under 20% relative humidity for long enough for the wood to reach equilibrium then swings to over 70% relative humidity, again for long enough to reach equilibrium. Since most furniture is finished, such a wide range seldom occurs on occupied homes. This is a 10 percentage point range in moisture content. To get the full range there probably has to be a substantial period of storage in unconditioned space. Nonetheless, furniture should be designed for conditions at least close to the most severe.
As far as maxim's go, the same experience repeated a thousand times creates no learning.
Now Steve old Boy ,
I said stile to panel is going in one direction , I did not mention rails and glue .
Maybe if the OP glues the panel to the stile without the hinges this could be safe .
The frame can become constrained when you glue a panel in , well that's just in my experience , unlike an attorney or physician I'm not practicing I'm doing it for real .
As far as the statement you made of " the same experience repeated a thousand times creates no learning " ,,, if you say so , but I say it creates a lot of experience and proficiency . History is the test of time , I have never had a come back on a F&P of any type , so if you think I'm doing it wrong , I'm not buying your line .
You never actually answered the question posed to you , do you glue in all your panels ? and have you actually made many that way ?
Happy Easter to you and All
dusty
Steve,just discussing here, you understand...
I was referring to splitting on the moulded inner edge of the rail and stiles- glue may hold here more than on the bottom of the groove- I was not contemplating frame failure or joint failure.
I don't believe there is any sagging factor to worry with- the thin ply weighs little and more importantly there is little moment about the hinge stiles because of the short length(leverage) of those rails which are not carrying the weight of a heavy panel and are themselves wider than the stiles.
On further thought, and hoping that Lad reads all this, a compromise may be to do a version of what Uncle Pablo has suggested: instead of a groove for the ply he could just do a wide rebate, say 11/2" or 2" by a hair over the thickness of the ply say5/16ths" on the back of the frame, then he can glue the ply on effectively. A suitable moulding on the front inner edges would complete the picture.Philip Marcou
Hi Philip ,
After thinking about the rebate or fastening the panel in behind similar to the way you mentioned .
About 25 years and more ago it was a common method to make a frame , shape the inside detail and then glue a panel (then plywood ) behind the frame . These doors were called back panel doors . Every one of them will cup and warp away from the back , some more then others .
So if the frame was attached to something it would stay flat , otherwise maybe not .
cheers dusty
Lad,
Are the doors inset, overlay, rabbetted, etc.? What you are doing there, in a way, is like making a plywood door with a wider than normal solid edgebanding. I would approach it from that point of view. It needs only appear as a frame and panel in a true sense. As someone has already suggested, you may want to consider a thicker ply and go from there. With the dimensions you gave I would rely on the plywood for stability.
-Paul
London,
If I may -
Your main concern here is not whether or not to glue the ply. You can do it either way and it won't make much difference.
My problem is with the choice of the ply in the first place. The chance that it will cause these long doors to bow is much more real than any of the scenarios raised in the learned discussion. I would absolutely use veneered MDF for this door.
David Ring
http://www.touchwood.co.il/?id=1&lang=e
That's one of the few sensible things mentioned in this discussion, though I wonder if 1/4" MDF really would be an advantage over 1/4" ply. That MDF would remain flatter than ply is certainly true, but MDF is weaker than ply for preventing external forces (like the racking in the frame) from flexing the panel. Think of how ply is better from preventing shelf sag; I suspect it will also be better at restricting the frame from flexing.
I'm just surprised at how many posters think wood movement in the frame is a reason for not gluing in the panel, when it's the panel movement in a solid-wood panel that is the reason solid wood doors are not glued. Gluing in a plywood panel will create additional torsional strength to Londonlad's door, which is exactly what he needs to lessen the risk of racking. Think of the glued panel as an extremely thin torsion box construction...Recommending the use of "Hide Signatures" option under "My Preferences" since 2005
But the door is vertical, so there's no sag element to consider here. My shop stopped using veneered ply for these doors about 3 years ago because the veneered MDF is much better in this application. We commonly use 8mm MDF which comes to about 9.5mm thick after veneering. But I've used thinner stuff as well, and it's fine. I'm assuming here that the frame is made from good stock that doesn't need to be "straightened"; all I'm asking from the panel is to sit there and not get into trouble.
We don't glue the panel in the frame because ultimately it's a waste of time. It doesn't hurt but it's just not necessary in a well-made frame. David Ring
http://www.touchwood.co.il/?id=1&lang=e
Well-made frame is the key.
There is no sag, but there is considerable risk for racking when the door is 6' tall. Good cup hinges might keep the door flat by pressing it closed, but if the frame has non-straight grain, you could develop a twist in the door. Gluing in the panel will lessen the chance the door will twist. A plywood panel will lessen the chance of twisting more than an MDF panel will due to the structure of plywood. My point was this same difference in structure is most often noted in the case of shelf sagging, where MDF shelves sag considerably more due to the lack of long grain in MDF.
The real question in this discussion is to glue or not to glue. I submit that glueing is better because (1) the plywood obviates the risk of movement problems present in solid panels, so it can be glued; and (2) gluing will create some additional torsional restriction to door racking. I can't quantify how much it will help, but it will help some. Perhaps not enough to warrant the extra time, but if time isn't critical then it certainly would be a bit of extra insurance against racking. Recommending the use of "Hide Signatures" option under "My Preferences" since 2005
Hiya Pondfish ,
I would totally agree with the fact MDF or PB will have more deflection or sagging in a horizontal shelf type application . The best grade veneer panels we use in door frames does come with an mdf core , 5.2mm seems the most common here in the Western states .
The mdf panels are much flatter and free of internal voids and offer a more perfect surface than veneer core or plywood panels do especially for doors .
Since there are no crossbands or core voids there is no grain memory telegraphing through the veneers .
I guess if a person wants to glue the panels in great I have never known a professional maker that practiced this method as a regular way on ply or mdf panels for that matter .
It's also true we can all become set in our ways and neglect to try different approaches , part of the beauty of our craft is there are many ways of doing the same thing , none are wrong , some are better than others , but it's all good .
dusty
I agree! The only that isn't good about making sawdust is when termites do it...Recommending the use of "Hide Signatures" option under "My Preferences" since 2005
Folks, I admit to be intrigued by the way some people build in imaginery problems and or insist on applying "rules" rigidly come what may, so I would like to have one more go at putting my case-which is that glueing a 1/4" ply panel into a door frame of these dimensions is irrelevant.
Look at it this way:-
1)If the frame is made from stiles that are 2" wide , the top rail 4" wide, the bottom rail 8" wide this would give a strong frame and the ply panel would only be between 8 and 9 inches wide by just over 5 feet long-hardly heavy or capable of distorting the frame, especially if good quality 5 ply.
2)Since the ply is light and flimsy this is one reason why there is no question of sag. Another reason why there is no sag is because of the dimensions of the rails-both their width and their length- hell they are only 12 inches long, and carry only the negligible weight of that ply and two slim stiles.Obviously there will be substantial haunches.Neither is there a question of distortion or racking- the ply is flimsy and the joinery can be designed for optimum strength by preferably by use of 7/8"ths stock over 3/4" and the use of wide rails: all the stiles have to do is stand there.....and one can have 3 or 4 hinges.....
3)There is also another admittedly remote reason for not gluing: the chances of splitting the moulding if there is one on the rails , should they shrink back and by sod's law the glue has stuck like s"ht to a British army blanket there....
A tall slim door of these proportions is quite feasible, provided the frame stock was decent timber with appropriate joinery.
I have harped on about this because it is an example which has ramifications in other applications, but I now rest my case apart from stating that Uncle Pablo hinted at an effective way to glue that panel for those who just must do it (;)Philip MarcouEdited 4/8/2007 8:44 pm by philip
Edited 4/8/2007 9:06 pm by philip
You must be surprised by the answers you've had to what seemed to be a basic woodworking question. Not only do you have theorists that may have never built a long skinny door but you have experienced pros saying the opposite from each other. It just goes to show there are a lot of ways to apply woodworking principles.
How about a simple compromise? Glue the panels in the four corners.
Beat it to fit / Paint it to match
Interior Center Rail
I'm new to this stuff but perhaps you could add a thin but wide center rail to the interior of the door using lap joints. Plywood panel is still one piece and no center rail on the outside. If you are really concerned about strength, add a diagonal to the interior instead.
This forum post is now archived. Commenting has been disabled