After 20 winters in Maine, some of the dovetails on a cherry blanket chest (18x18x36) I built for my wife are beginning to unglue due to warping at the sides of the top of the box. In other words, the sides are beginning to “bloom” outward from the front of the chest. Also, on inspection I’ve found that one of the 6″ panels that were end-glued for the back has separated from its neighbor along the length of the glue line.
I can clamp the sides and front back into square, and plan on pegging some of the tails with Miller dowels. Since the space between the tails and pins is less than 1/32 and the glue in the joint has dried out, re-glueing may not help much. I’m sure that pegging them into place will hold the pieces tightly together, but am not sure if wood movement will split the wood in the future. I suspect not…but if anyone has a knowledgeable opinion, let me know.
As for the space between the back panels, I was thinking of gluing a strip of edging between the panels as the gap is about the thickness of edging, although that presents problems as well. The edging I have on hand is the iron-on type and I don’t know if it will hold glue spread on the iron-on glue. Also, of course, I cannot re-clamp the panels so the ‘fix’ will probably just be aesthetic and make me feel that I’ve repaired it…somewhat. Structurally it won’t help. The panel seems secure enough, but I’d be happier if it was re-glued to its neighbor.
Any suggestions or insights into wood movement and this type of construction would be appreciated. What should I have known or done before I put it together? Wood movement makes an occasional project fraught with questions that I really cannot answer. I’ve seen an article in F/W by Christian Becksvoort on the subject. Very interesting and totally confusing. Seemed like I’d need a degree in physics!
English
Replies
It's probably my own mental laziness, but I'd really appreciate a photo of the chest in order to relate to the problems you're talking about. Can you post?
DR
english,
I'd recommend trying to work some glue into the dovetail joints that have moved. Use the same typew of glue that you used before, and rub it into the joint with your fingertip. If you can, wiggle the board back and forth to further spread the glue, then repeat the rubbing in. Clamp closed. Instead of adding dowels, I'd likely wedge the dovetail by splitting the loosest pins with a chisel and driving a thin tapered wedge (dipped in glue) into them.
The problem you describe with the back is a tough one. You describe a panel that was end glued and split lengthwise. If the back is frame and panel construction, the panel should have been left to float (swell and shrink widthwise) in its groove, and not been glued. You can fill the gap with a slip of wood. But as the panel tries to swell again next summer, it will compress the slip. Then as it shrinks again in the next heating season, you'll likely see a new gap appear beside the slip you laid in. No easy fix here, unless you can somehow break one edge of the panel (and both ends) loose from the frame so it can come and go.
Regards,
Ray Pine
Sorry, don't have a picture...The back panel is made up of three 6" cherry stock 36" long. It is also dovetailed into the side pieces, i.e., dovetails front and back of chest. There's no way it could 'float' with this construction. I think you're right about the slip between the edges won't really do anything at all. If I was really worried about strength I could bridge the gap on the back with something but that would really be clunky and I don't think it's about to come apart. The dovetails are holding it in place and also the bottom which - for better or for worse - was dadoed and glued in around the chest.
Question: Would the chest have been more stable if I had polyurethaned it instead of using Watco Danish oil? Why does commercial furniture not suffer the same movement/warping, etc? What could I have done differently to ensure that the dovetails would not separate?
The joints are way too tight to sand between. I can just get a thin knife edge between the tails/pins that have opened.
English
Same as Ring-I would like to see how these dovetails are moving, assuming everything is 'normal' as far as grain orientation goes-a picture or two please.
Sorry, guys, no digital camera. It's really quite simple. The chest is 18x18x36. If the chest is facing you, the left and right hand side panels(18x18x18) are warping out from the top, i.e. away from the inside of the box and the front of the box is moving forward away from the sides. The glue has dried out and failed. That problem is moot now anyway. Yesterday I pulled everything into square and pegged it with 1x Miller dowels, which, by the way, were disappointing. Most of the dowels were oversized for the stepped drill bit and it took some searching thru the 40 cherry dowels to find 4 that would fit without fear of splitting the tails thru which they were placed.
Question still remains. If one starts out with rough lumber and are going to dovetail a piece such as mine, how do you account for wood movement in the design? As far as I can see, dovetails don't really allow for any give or take? I will assume that quartersawn would have been better than flatsawn and that kiln dried lumber would have been better than air dried?
english
If you put it together the usual way, then the grain is running "continuously" around all 4 sides, and any expansion and contraction across grain will affect all these pieces together as a unit. If a board wants to cup, then conceivably a mortised corner could pull away from the pins (not vice-versa). But in practice I haven't seen it happen because close-fitting dovetails have so much gluing surface in addition to the physical restraints of the joint that the board would probably split before the joint came undone.
If you start with reasonably dry and straight stock it should stay put forever. Sorry, but I can't get a handle on the where the problem is coming from.
DR
I agree with what Ring has said.The only movement I have ever seen on dovetails that were well made is the slightest surface differences where the pins project a little.
Did this happen suddenly? What has been kept in the chest?Philip Marcou
You have pretty much diagnosed your own problem in your posts. Your wood may have needed some time in your shop or home to acclimate before you built your chest. The orientation of the growth rings, when you glued the boards together may have contributed to the 'cupping' of the panels. The glue failure could be because you didn't have enough glue in the joints when you assembled it. Or it could have been something to do with the glue (a bad batch or time expired from sitting on the shelf too long). The splitting of the back panel could be because it has dried more than the other three panels. It may have spent some of its life in front of a heat register for example. The best way to have repaired the chest, but not necessarily the easiest way, would have been to take it completely apart and reglue it. If the split in the back panel was along the glue line you may be able to run it over the table saw blade, and then glue a piece into the saw kerf. That will depend on how the split runs through the dovetails.Using a 'hard' finish like urethane or lacquer, rather than oil, might have slowed down the seasonal moisture changes and reduced the wood movement effect. But it wouldn't eliminate it, and the inside of the chest might smell for a long time. Shellac would be the better choice for the inside.Woodworking as a learning process, can sometimes take a very long time, because it can take so long before things actually show up. It took 20 years for your chest to start showing its problems, but it has you now thinking about how you can do better for the future. That has value. Over the next 20 years you'll be able to observe what will happen with your repairs and find out how effective they will be.
Thanks for the input. Houses in Maine, at least mine, stays very dry in the winter and humid in the summer, and I think that took its toll on the piece. Taking the chest apart would be brutal as most of the dovetails stayed pretty much together and the glued-in bottom is solid. The ones on the top, although having close tolerances, just worked themselves apart in the same direction as they were put together. I didn't expect that dovetails could do so but given enough wood wiggle/movement, it can happen.
As for the next 20 years, I'm 65 and I guess the best project would be a nice casket! I'll use whatever glue I've got lying around and if the sides open up after I'm in it, I'll let you know!
Take care...english
Lee Valley sells a wood movement 'wheel' and all you need to do is rotate it to find the type of wood, then find the multiplier and use it with the dimensions that will be affected. If the inside is unfinished, that will have a big affect on the stability, too. The sides are 3 pcs, 6" wide, right? Was the center board flipped so the grain alternates from one piece to the next?
"I cut this piece four times and it's still too short."
Yes, the middle piece was 'flipped'. I think the prevailing knowledge back then was that a board with a ) grain would warp with the grain, i.e. ). But a recent article said that is not so. It will warp against that grain, i.e. (. That seems to be the case with the chest sides. I'll check out the Lee Valley 'wheel'. How would knowing the 'dimensions' work? Doesn't one have to know where not to glue to allow the wood to float? Thanks.
The wood will shrink or expand across the rings and if the board is plane sawn or rift cut, it will warp if it's too wide. Quarter sawn is much more warp resistant but it can become slightly wider, therefore, room for expansion is needed. The dimensions are needed bacause the charts use a multiplier to determine the amount of shrinkage. It's like any other material, where you take the thermal coefficient of expansion, the temperature max/min and multiply it by the dimension to find the range. Wood is one of the few materials that changes dimensionally due to humidity. Unfortunately for people like us, who try to make things with it, in a way that they'll stay together for a long time.Has the warping started recently, or has it happened every winter for years? Maybe you could raise the humidity and see if it goes back to normal, then inject some glue into the dovetails. If the sides are only finished on the outside, you could remove the contents, set a glass baking dish with damp towels inside so the wood can absorb some moisture. You don't want to change the moisture content too quickly, though. Since the 1970's, there has been something called 'Dampit', for musical instruments. This is a surgical rubber tube with holes in it, and the maker slides a strip of a sponge into it, caps the ends and attaches a clip to it so it stays where you put it. When it's dry, soak it and put it back. They sell these at music stores."I cut this piece four times and it's still too short."
Edited 1/1/2006 12:12 pm by highfigh
English,
I was shocked when I opened an old Highboy recently at the Wayside Inn in Sudbury, MA.....the dovetails joints had really morphed into something else..suspect its the moisture changes. I've often wondered how many antique piece never made it because of wood movement as opposed to poor craftsmanship.
Anyhow, I'm kinda curious about your issue relative to the finish you applied to the inside of the box? You mentioned Danish oil but did you use that only on the outside of the box? Did the cherry have any finish on the inside?
I used Watco Danish oil inside and out. I didn't flood the inside surfaces from the look of it now. I'm a bit dyslexic visually and this chest was the first woodworking project. I thought perhaps I did the dovetails backwards - if that is possible. The way the tails and pins meet, you wouldn't think the side pieces could splay outward away from the front. But that's what happened. I guess I thought it was indestructible and putting it too close to a heat source was a bad idea. I wish I could remember which glue I used then but I cannot. I do know it was a better brand made for furniture. Got it Woodworkers Supply or some such store in Cambridge Mass. I guess glue can fail if the conditions are right.
Now that I've pegged it back to square, I'll give it a good cleaning with Watco and re-fresh the surfaces. My next problem is that when I was looking at the dovetail movement, I inadvertantly moved the piece while the top was resting on the back of a couch. Since the top had no constraining hardware besides the hinges, it went #### over kettle and tore the screws holding the hinges out of the edge of the top and split off about 1/2". Since I had an overhang of 1/2", the top is now exactly the depth of the chest and will be difficult to open as there is no lip. I've got some cherry stock the same thickness and can glue up an inch to the back or, if I want to get fancy, rabbet in a 1" piece the length of the front and use that as a lip. An alternative is to just use some brass pull handles screwed into the existing lip and move on to building a plush casket!
English
I'm a big Lee Valley fan, but if you want to accomplish the same thing while using your computer, go to this site.http://www.woodbin.com/calcs/shrinkulator.htm
This forum post is now archived. Commenting has been disabled