Basic question here but an important one.
I built a table last year but now the top is coming apart at the seams. I often hear of this problem but I really don’t know how to prevent it. I used biscuits and glue to join the boards for the tabletop. What steps should I have taken to avoid this problem when the weather changed?
I’m going to build another tabletop and I certainly don’t want this to happen again.
Thanks,
I’m…
KooKoo for Cocopuffs!
Replies
Could be a combination of things. Some might be ..old Glue, poor jointing, vast change in moisture, oily wood/ wrong adhesive, crossgrain restraint..or gremlins
I don't see how even a vast change in moisture content could make a properly jointed and glued table top break up, assuming all the boards were at the same MC when it was made. If every other board were quarter sawn and the ones in between were plain sawn, I suppose it's possible that the difference in thickness change could be enough to start breaking the joints.
The thing that intrigues me is that this happened even with biscuits. I know biscuits don't add a great deal of strength, but they must add some. If the biscuits are pulling out of their slots, I would suspect a glue problem. Glue age, as you say, is one possibility. Is it possible that assembly time got stretched beyond the glue's open time?
Another question for KooKoo: How many boards are there in the table top, and how many of the joints are opening?
Uncle Dunc..I throw a biscut in once and a while to help with alignment (esp. if the boards are NOT all the same thickness, I did a lot of "rustic" work) to get a even top face. I do not think they are needed, and I do know they can telegraph if too close to the surface. I would bet bad glue. REALLY bad glue. Maybe like the white paste me an a girl used to run off and eat together in Kindergarten.<G>
Speaking of telegraphing: I just used biscuits to attach lipping to both sides of about 40 feet of counter top for an office. I set the biscuits closer to the bottom than the top surface, but they still telegraphed bad enough to make it hard to level the lipping, and the bumps (along with my inattention) also caused a couple of sand-throughs on the seal coat. I was using catalyzed urethane, so I wasn't too happy... Think I'll use splines, or nothing, next time.
By the way, I just checked Franklin's web site, and there specification for open time is only five minutes. They also specify clamping pressures in the range of 100 to 250 PSI. 250 PSI would require 1500 LBS of clamping pressure every 6 inches. Pretty hard to exceed with bar clamps -- at least without severely denting the wood.
Michael R
Woodwiz..the little buggers DO swell..if you don't wait long enough for the swell to subside before leveling ( it will, and then ya have the OPPOSITE problem sunken telegraph) that is what ya get..I dont know if the plastic biscuts work better. Splines are definatly a good option, but being basically lazy I rarely use anything any more other than clipped finish nails to help with slippage. Wow, FIVE mins. huh? and that is some serious clamp pressure. How'ed ya feather in the Cat/thane without a witness area?..Cat lac is next to impossible to make invisible repair.
>>Woodwiz..the little buggers DO swell..if you don't wait long enough for the swell to subside before leveling
Biscuits are compressed quite a bit, using heat. I don't think they shrink back enough to completely reverse the initial swelling. These were a week old before I got around to finishing them, so I imagine they were about as equalized as they were going to get.
>>How'ed ya feather in the Cat/thane without a witness area?..Cat lac is next to impossible to make invisible repair.
I burned through the sealer coat, not the topcoat. I just touched up the stain - painted in the base color, then the grain - and shot another coat of sealer on that panel. When you repair topcoats, it's easiest just to sand and re-coat the entire panel after the repair has been made and leveled.
Michael r
I was thinking more about a solid wood top with biscut swell being sanded away then the sunken area where the moisture from the glue swelled the wood as well as the biscut. Similar to a shrunken glue line. I missed the part about it being sealer..my bad.
KooKoo
Agree with Sphere, but probalby my biggest concern would be at what moisture content did you assemble the top an what is the moisture content currently?
Regards...
sarge..jt
Proud member of the : "I Rocked With ToolDoc Club" .... :>)
Sarge, have you ever witnessed anyone (including yourself) ever OVER CLAMPING to the point of a starved joint?..I have read this and recall in my early days (20 yrs. ago) of REALLY crankin down on sprung joints and glue (yellow then , before TB II) squeezin all over the place, no biscuts..and I still have some things now that are still holding up.
Is that whole starved joint thing a myth? Or maybe what happened to the above poster? It is not useually my first thought, but may also be cause. Any input? Duane
SPHERE
I have heard the story also, but have not experienced it. After Xmas I have to replace a kitchen table that I built 24 years ago. I CLAMPED down on it and used the old-style glue and it is still throughly in-tact. My wife just got tired of the look and wants another. (You know women, no sense to argue; just do it) ha.. ha...
It's a good question though and hopefully someone will comment if that has happened to them. I did build that old table outside in summer with just butt joints and have no clue what the moisture was. The biggest problem I have seen with glue joints coming apart is from using wet lumber and the radical shrinkage that will result when it finally dries out after assembly.
You might even start a thread on that if you are overly curious. It's definitely a good question.
Regards...
sarge..jtProud member of the : "I Rocked With ToolDoc Club" .... :>)
Sphere,
Last summer I made a router table, drop down table for the TS, etc. and decided to make some room in the shop by recycling some cabinets I had made a couple of years ago. Every single plywood dado joint came apart with little effort. Lots of glue had been applied..and still showed..but the joints were quite tight and I had clamped hard enough to change the DNA. I have stopped useing yellow TBII....I also bought better clamps
interesting, a plywood -dado-glue joint seems to lack the proper ratio of good grain to good grain for glue only. I always suppliment with pins or what ever will work (screws if a blind side cabinet). Even a dado in solid stock and glue alone presents the same problem, assuming say a shelf end grain into a plowed side/upright. That brings up the starved joint again..too tite dado, too much clamp pressure = no glue where it does the best long to long grain. I beleive that is an area where the myth may be true..we'll have to run it by Sarge.
When using PVA or waterbased urea formaldehydes adhesives, at least 200 pounds per square is needed for proper bonding of hardwoods. A standard 3/4" pipe clamp provides about 2000 psi. So, if you need to edge glue two 3/4" by 60" boards, you will need to apply 9,000 pounds. This will require five pipe clamps fully tightened. So, it is almost impossible to "starve" a joint in the average hobbiest tool shop.
Exactly my thought..I have used a vacuum bag under full vacuum to laminate 1/4" exotics to 1 1/2 alder for guitar bodies, that 1/4 " also had to bend to forma comfort cut where the rt. elbow would be..the pressure and vacuum literally sucked out the glue..it all stayed stuck.
Actually, vacuum bag glueing provides fairly little pressure. At best, 14.7 psi is the max pressure exerted at standard sea level. It's fine for veneering, but hardly sufficient for standard joint gluing.
Reading all of the responses crticizing biscuits reminded me of an article that I read many years ago on glue joints. The bottomline of which was that glue joints were strongest in shear strength and weakest in tension. That's why finger joints and the other types of glue joints that you can get router bits for are stronger than butt joints --- the joints provide larger areas where any applied load would result in a shear force in the glue bond as opposite to strictly a tension type load. As a result, it would seem to me that biscuits do play a valid role in panel type glue ups. They may not be as strong as pocket hole screws or M&T joints in frame construction, but then not all projects need the ultimately obtainable strength the other joints offer. It would be interesting to get a response from someone at Franklin or Elmer's on the tension vs shear strength issue in today's glues.
bd
>>The bottomline of which was that glue joints were strongest in shear strength and weakest in tension.
Good observation, and it goes a long way to explain why dowels perform at a level similar to M&T joints in many applications. There may be vertical shear across the dowels at the actual joint, but the stress on the glue seam is all in shear along the dowel. That's why the strength of the joint appears to vary with total glue seam area and dowel spacing regardless of grain orientation in joints like face frames and chair parts.
The actual way that glue works is fascinating: There's mechanical bonding, molecular bonding, cohesion, and a host of other factors that explain why with some glues like PVA a thin glue line (.003) is stronger than a thick glue line (.010"), while many epoxies are weaker with a thin glue line. Each glue has a different mix of characteristics, and it helps to know a little about how each one works in order to use them correctly.
>>It would be interesting to get a response from someone at Franklin or Elmer's on the tension vs shear strength issue in today's glues.
Unfortunately, glue strength varies greatly with the nature of the pieces being glued together, so any response you might get to your question would necessarily be very general. Still, it would be interesting.
Michael R.
I have the strong opinion that for edge joints, glue starvation is certainly a myth, that is as long as enough glue was applied in the first place. Glues such as PVA and Hide glue, work chiefly by being absorbed into the cell structure of the wood and forming keys. The amount of glue necessary to achieve this condition is quite small. With this in mind over clamping couldn’t possibly, starve the joint of glue, because the clamping pressure would only serve to force the glue into the wood. This does not mean that I advocate heavy clamping pressure, in fact the opposite is true. Some people try to overcome poor workmanship with excessive clamping pressure. Other glues such as epoxies, don’t rely on this keying effect to as great of an extent , but then it is a poor choice for edge jointing anyway. Now joints such as a mortise and tenon, can be glue staved , so I apply glue to both the tenon and mortise. Although I’ve never had an edge joint fail, if I did I would check the following, burnished surfaces from dull tools, joints that weren’t given enough of a spring or too much of a spring, old glue and too long an open time before clamping. The age of the glue is especially important when using yellow glue, as its shelf life is somewhat limited. The aging factor of yellow glue is one reason that I don’t use it, the other is that it is prone to creep. I’m unimpressed that yellow glue is stronger than white glue or hide glue, since both of these are stronger than the wood itself. Wood cracking at the joint, I would say, is not necessarily a sign of cross grain construction, because glue joints are typically as strong or stronger than the wood itself. I also think that biscuits and dowels are unnecessary in edge joints.
Rob Millard
To you sir, I tip my hat. I have read many of your accurate and impressively knowledgeable posts. It is a true pleasure to share this venue with those who do not feel the need to lead all the horses to THE water, rather to point out that there is much water. Chosing to drink or not is purely from the amount thirst that exists.
It might help a lot if you told us what you did. What kind of wood, what kind of glue, how well was the wood jointed, how you did the glue-up, how long you left it clamped. Too many variables to isolate just by guessing.
Second, what do the failed joints look like? Did the glue line fail? Is there glue on one side, or both sides of the joint? Is there glue visible at all? How thick is the glue line? About .002-.003 is optimum. Has the top been exposed to too much heat or moisture?
Biscuits add nothing to an edge- to-edge glue-up. Furniture tops have been made without them for hundreds of years, and still are in industry. I tried them, and found they didn't even help that much with alignment. All in all, a big waste of time, and the time they take to install could be a cause of the joint's failure. Open time for most wood glues is about 15 minutes. I do use biscuits from time to time, but never for this.
Michael R
I didn't see that anyone asked if you had floated the top to allow for expansion/contraction?? How was the table top affixed to the frame?
What kind of wood was used and what is the width of the glue-up?
PlaneWood by Mike_in_Katy (maker of fine sawdust!)
PlaneWood
When you say the top is coming apart at the seams... Is it in the glue joint? If so, is the ENTIRE seam coming apart at once or only at the ends. This is important.
Edited 12/14/2003 4:04:52 PM ET by Ken's Shop
Edited 12/14/2003 4:18:02 PM ET by Ken's Shop
This is a subject that is central to wood working. Glue joint strength. The factors can't be repaeated too often.
Biscuits add nothing to the strength of a joint noTHING! They are an aid to alignment, only, and of marginal use even there. Their popularity is wildly out of proportion to their usefulness. In the hands of a pro in a production shop who knows exactly what he is doing and where time is money, they are good. Otherwise, I think they are a case of the manufacturers tapping into tool envy and gadget fascination to sell a lot of machines. There are other, better ways to accomplish the utility they are supposed to provide.
There is no way to make a glue joint any stronger than properly preparing long grain to long grain. Two boards, properly glued and subjected to enough stress to break, should not break at the glue line.
As has been mentioned, a slightly concave joint (1/64" deviation in 6 feet of board length) greatly helps to keep the joint ends closed when the mid point is snugged up.
The best glue surface is made by a very sharp hand jointer plane prepared within hours of the glue-up. Oxydation of the wood surface in as little as 24 hours degrades the glue-wood adhesion. Next best is a jointer or planer with very sharp, correctly set knives. If not sharp, the rotary cut burnishes the wood, degrading the quality of the gluing surface. The worst thing to do is "rough up" the surface with sanding to give the glue a "tooth" to adhere to. This just generates weak, loose fibers that guarantee a joint failure.
I agree it IS very hard to starve a correctly prepared joint with hand clamping. But I assume it does happen. Then there are all the factors regarding bad glue, improper mixing of powdered glues, low temperatures, etc.
VL
Venicia,
Perfect evaluation of the usefulness of biscuits, be prepared to be yelled at.
John W.
John,
Yes. It won't be the first time. Or the last. It's rather fun being an iconoclast. The last thread about pocket hole joints was quite lively. (I'm agin' them).
The same advice about biscuits, of course applies to dowel joints. Except possibly, for an end-to-end joint, with deep-set dowels acting (well) as splines, there is not a single wood joint that benefits in strength from dowels (James Krenov's advice to the contrary, not withstanding). Yet the furniture industry promulgated this joint, a boon to alignment in assembly, as a sign of quality for years. And the public and many woodworkers bought it.
VL
Venecia,
If memory serves, Krenov often used a line of dowels to attach the overhanging tops and bottoms of his cabinets to the vertical grain sides with no visible means of attachment. Those joints wouldn't have been especially strong, but they were probably adequate for the job if the cabinet wasn't handled roughly and the top and bottom pieces were quarter sawn to minimize cupping.
Personally I use screws and bolts for the bulk of my joinery using plugs and moldings to hide the screw holes, nothing ever comes apart unintentionally and the piece can be disassembled to tweak the fit or for finishing and repairs.
John W.
P.S. I share your opinion of pocket screws, they're right up there with biscuits, splines and dowels for wasting time with no useful improvement in joint strength (in most applications).
Edited 12/14/2003 6:49:33 PM ET by JohnW
John,
Yes, the cabinets he made that way were probably strong enough as he designed them. He didn't want the visual design to be marred by a joint with the mechanical strength typically needed to hold the bottom of a wall cabinet to the sides. And Krenov being Krenov, he gets away with such practice. Not because he's right, probably because his pieces are so revered, that the cabinets would never be used to actually support any weight inside.
And, maybe he is right that more robust "correct" joinery is not needed. But I don't really buy his notion that dowels have more holding power than many believe. True, they'll never come out of the vertical side pieces, where the dowel to side piece is all long grain. But there is precious little long grain to long grain contact in the dowel to bottom piece joint. He could have done better, and very elegantly, too.
VL
>>The same advice about biscuits, of course applies to dowel joints. Except possibly, for an end-to-end joint, with deep-set dowels acting (well) as splines, there is not a single wood joint that benefits in strength from dowels (James Krenov's advice to the contrary, not withstanding). Yet the furniture industry promulgated this joint, a boon to alignment in assembly, as a sign of quality for years. And the public and many woodworkers bought it.
Lively discussion brewing here...............
Are you talking about edge joints only? If so, I agree with you.
On other joints like chair joints, what exactly is the functional difference between a dowel and a loose tenon, except for the fact that one is round and the other is rectangular? For that matter, what is the functional difference between a round tenon cut with a chucker on the end of a spindle, and a dowel?
Are you saying that an end-grain to cross grain butt joint would be as strong with glue only as it would be with dowels or biscuits? I hope not.
Joints in a chair are probaly the most highly stressed joints in woodworking. There is actually a lot of engineering and calculating that goes in to designing such furniture, and dowels are definitely the engineering equal to a M&T joint of equal cross section and surface area. Sometimes dowel joints are stronger because they allow a more robust joint. I can look up and cite the wood engineering texts if needed.
From an anecdotal perspective, I have repaired chairs with both types of joints, and really don't see much difference, except that dowels are glued at both ends, while tenons are glued just at one end. Either way, the glue almost always fails, not the wood. Broken tenons and broken dowels almost always result from movement due to glue line failure, and that failure almost always results from a combination of cross grain stresses, heat, and moisture.
I have never seen a chair assembled with plain butt joints.
Buiscuits add considerable strength to butt joints, too. I don't like them too well because the biscuits themselves don't have a lot of strength or surface area, and they don't penetrate far enough, but for assembling boxes or holding fixed shelves they are OK -- depending on the application. I've used them is store fixtures & the like when I didn't want to put fasteners through the outside surface.
Michael R
DUCK...INCOMING..yer both right
>>DUCK...INCOMING..yer both right
Is there any better way to learn OR teach than a lively, civil comparison of viewpoints and experiences? I think not -- or at least none as entertaining!
>> ... what exactly is the functional difference between a dowel and a loose tenon ...
The exact difference is that the wall area of the dowel hole is 50% end grain, or more, depending on how you define end grain, while the wall area of a properly laid out mortise is mostly long grain.
>> ... what is the functional difference between a round tenon ... and a dowel?
There isn't any. Round tenons share all the shortcomings of dowels.
Except..in chairs the round tenon USED to be 1 Bulbous and 2 Drier than the post..locks it in better in withdraw..just puttin more fuel on the fire..lol
>>The exact difference is that the wall area of the dowel hole is 50% end grain, or more, depending on how you define end grain, while the wall area of a properly laid out mortise is mostly long grain.
Good point, but the sides of the tenon are still cross - grain, and subject to deterioration due to wood movement. I confess I don't know how much stronger the pull-out strength of a cross-grain joint is than an end- grain to long grain joint, but I do know that until the glue deteriorates you will break wood before either joint comes apart. The texts I used insisted that dowels and m&t joints were functionally equivalent joints, and had the numbers to back it up. You don't want to design a chair and make 10,000 of them just to find out that a joint is too weak, and they all collapse after three months.
I like m&t joints better myself, but have trouble coming up with an objective or experiential reason why. The strongest reason I can come up with is that mortise and tenons hold up to twisting forces better than dowels, because they don't put the glue line in shear.
Michael R
>> ... the sides of the tenon are still cross - grain,
So are the sides of a dowel.
>> ... insisted that dowels and m&t joints were functionally quivalent ...
I guess that's at least plausible, as long as they're talking about joints with equal glue area. In production it's easy to see how dowel joints would be cheaper to make. As a hobbyist, thinking about drilling enough matching dowel holes to equal the glue area of a m&t gives me the willies.
Woodwiz,
Uncle Dunc answered your question to me. Most of the grain contact in a dowel joint (unless the dowel is let into a hole in the end of the board) is dowel end grain to long grain in the receiving piece - a very weak joint.
No, I do not mean that a butt joint of end grain to long grain is just as strong as one made with a dowel in there. I believe all effort should be made to eliminate such mating. The doweled joint is only marginally stronger. And most often the strength has little to do with the glue joint. It comes simply from the physical space-taking presence of the dowel. The joint simply doesn't completely fall apart until the dowel is racked completely free of its hole.
My point is that many workers assume that a doweled joint is a strong joint. It is not a strong glue joint. I think it's an expedient construction that can be improved in almost every instance by a single or double tenon that correctly places long grain against long grain and maximizes those glue surfaces, rather than minimizing them as a dowel does.
VL
>>The doweled joint is only marginally stronger. And most often the strength has little to do with the glue joint. It comes simply from the physical space-taking presence of the dowel. The joint simply doesn't completely fall apart until the dowel is racked completely free of its hole.
>>My point is that many workers assume that a doweled joint is a strong joint. It is not a strong glue joint. I think it's an expedient construction that can be improved in almost every instance by a single or double tenon that correctly places long grain against long grain and maximizes those glue surfaces, rather than minimizing them as a dowel does.
I won't disagree that a properly designed mortise and tenon joint is preferable to dowels if you want to make the best furniture you can. But people have been manufacturing furniture with dowels for about 150 years now, and if the joint had no strength at all, you'd think all those big businesses woud have got a clue by now. With automated routers and tenoners, the cost of producing m&t joints isn't a whole lot higher than dowelled joints, yet the factories keep on using dowels. Why? Yes, it is expedient, but expedience doesn't do any good if the method doesn't work.
The excerpt below should suffice to show that dowel joints are at least a close equivalent to mortise and tenon joints. It's from a study done by a european wood technolgy institute. The joint tested is a typical back leg to side rail joint. You have to read it carefully, because the tenons and dowels are not equivalent in width or shoulder spacing. Compare line 3 and line 6 for equivalent joints, for example, and read the conclusions. This is what I was taught 30 years ago, and wood hasn't changed, although adhesives have improved somewhat.
Even the most negatively biased view of this data would put dowels as a near equivalent of M&T joints in initial strength. I don't think they resist fatigue as well, but as I said earlier, in almost every joint I have repaired, the glue has failed due to cross-grain wood movement, which can't be avoided with either joint. Your long grain to long grain contact in a m&t joint is cross grain, and if you look at it, there's actually more vertical cross grain movement in a m&t joint than with dowels, horizontal movement being the same in both.
The text in the study did not specify tolerances on the joints, and the lower performance with PVA glue on dowel joints is unaccounted for, but I suspect it may be an artifact of tolerances, because the authors go on to say that they believe performance would not have dropped off with a "gap filling" PVA glue.
With all respect, I don't think it is useful to condemn a whole class of joinery out of bias founded on only theory and speculation. Real world data trumps theory every time. I agree with your preferences, but not to the point of judging other methods unfairly, and promulgating that unfair judgement as fact on a forum read by less informed people. Sort of noblesse oblige, I guess.
Michael R.
Assembly no.
Type
Tenon width or dowel space mm/inch
Schoulder on apron top/bottom mm/inch
Depth of tenon or dowel mm/inch
Average breaking load Nm. Urea glue
Average breaking load Nm. PVAC
1
Tenon
41/ 1½”
6,4/6,4 / ¼”/ ¼”
19 / ¾”
172
173
2
Tenon
41 / 1½”
9,5/3,2 / 3/8”/1/8”
19 / ¾”
172
170
3
Tenon
35 / 1 3/8”
16,0/3,2 / 5/8”/1/8”
19 / ¾”
139
132
4
Tenon
41 / 1½”
9,5/3,2 / 3/8”/1/8”
16 / 5/8”
153
145
5
Tenon
41 / 1½”
9,5/3,2 / 3/8”/ 1/8”
22 / 7/8”
203
198
6
2 dowels
32 / 5/4”
11,0/11,0 / 7/16”/7/16”
19 / ¾”
164
145
7
3 dowels
16 + 16 / 5/8” + 5/8”
11,0/11,0 / 7/16”/7/16”
19 / ¾”
195
160
8
2 dowels
22 / 7/8”
16,0/16,0 / 5/8”/5/8”
19 / ¾”
150
120
9
2 dowels
22 / 7/8”
19,0/13,0 / ¾”/ ½”
19 / ¾”
139
112
10
2 dowels
32 / 5/4”
11,0/11,0 / 7/16”/7/16”
16 / 5/8”
151
129
11
2 dowels
32 / 5/4”
11,0/11,0 / 7/16”/7/16”
22 / 7/8”
202
161
The results show that an extension of the tenon and dowel from 16 to 19 mm (5/8 – 7/8”) has resulted in an increase of the strength by approx. 15%, which is due to the larger glue-seam-area. When loaded, the glue seam is exposed to displacement tensions on both sides of the tenon and the strength of the joint mainly depends on the size of the actual glue-seam-area.
When a dowel joint is loaded, the greatest part of the load will be on the top dowel, why the strength of the joint depends on both the location of the dowel as well as the glue-seam-area. In other words, it is the location of the dowel, which has caused a reduction of strength in the joint with 2 dowels with 22mm (13/16") spaces in proportion to the joint with 32mm (5/4") spaces.
The survey also shows that the dowel joints glued with regular PVAC glue are approx. 20% weaker than the equivalent joints glued with filling urea glue. This difference would hardly have occurred in case one of the filling PVAC types (tenon glue), presented on the market after the tests had been carried out, had been used in stead.
The survey has shown that it is not possible to give a generally applying answer to which assembling method is the strongest for chairs. A series of circumstances such as demand of strength, designing, dimensions, wood species and production machinery are of importance for the choice of construction. Never the less experiences indicate that negative results usually are not due to a wrong choice of construction, but on the contrary that the employed assembling method is not a thought through idea, or it is not manufactured to satisfaction, e.g. with a too small glue-seam-area due to too short dowels or lack of glue application
Woodwiz,
I do not mean to condemn a whole class of joinery. But I'm sticking to my story. There is simply no way to explain away the fact that there is precious little sound gluing surface in a joint that mates a dowel to the side or edge of a board. The laws of physics and geometry are pretty firm on this point.
All those "big businesses" that need to have their wood chairs last, use chairs with M&T joinery. All those big businesses that sell chairs with dowel joinery do so because people buy them. And the joints fail. And craftsmen repair them.
How many chairs with M&T joinery do you repair compared to those with dowel joints?
VL
You are entitled to your opinion, but as a former engineering major, I have to disagree. I can't take the time to do a detailed response on the theory, but I doubt it would convince you. I have seen several studies over the years that give essentially the same results, and at the risk of committing argument from authority, how do you think your opinion outeweighs that of people who study these things for a living? And how do you explain the results in the article?
>>All those "big businesses" that need to have their wood chairs last, use chairs with M&T joinery.
Who would that be? You've really got me curious here. Can you actually name five large manufacturers that actually use M&T joinery in their chairs? I ran furniture factories for over twenty tears, attended way too many furniture shows, and never talked to a manufacturer who used that joinery, so if there are any, I would honestly like to know about them. The factories I ran produced $10 to 20 million at retail and were considered small. But if you can give me even three makers who produce that kind of volume, I will humbly concede your point.
>>How many chairs with M&T joinery do you repair compared to those with dowel joints?
Since probably 99% of the chairs in America have been produced with dowels for the last 150 years or so, I see very few with M&T joinery, but the percentages are consistent, even though my sample is too small to have any validity. I currently have in my shop six Chippendale chairs with pegged mortises. On most of them, the tops of at least one of the front legs has split at the peg. I will have to disassemble the joints, repair the splits and missing wood, and re-assemble. Considerably more work than regluing or replacing a dowel. I also have an 1820's low spindle back chair with broken and worn tenons that will have to be built back up or replaced. It's in my "someday" pile. I also have a pair of Georgian chairs available at an acquaintance's shop that my not even be worth restoring because the joints are in such bad condition.
I currently have one dowelled chair in my shop with broken back spindles. This does not prove anything except the invalidity of anecdotal evidence.
>>There is simply no way to explain away the fact that there is precious little sound gluing surface in a joint that mates a dowel to the side or edge of a board. The laws of physics and geometry are pretty firm on this point.
Again, no disrespect intended, but have you considered the idea that you may not understand the laws of physics and geometry as well as you think you do? And how do you account for the fact that every properly designed study and every wood technology text comes to the same conclusions?
>>And the joints fail. And craftsmen repair them.
All joints fail eventually, if stressed. And craftsmen repair them.
Michael R
Thanks for your opinion on dowelled joints vs. mortice & tenon. As a new woodworker, I have recently started building chairs and have used dowells with great results. I will continue to do things this way because as far as Iam concerned, the mortice and tenon joint is just too complicated for the average home woodworker, and dowels are a practical and strong alternative.
thanks again
Venicia L
I agree 100%. Biscuits are completely over rated. The first thing my high school shop teacher taugh me in 1972 was that a glue joint is stronger than the wood itself. And back then we were using crappy white Elmer's glue, the same thing kids use for kindergarten projects! It is VERY important to have properly seasoned wood and to joint it with a very slight gap in the middle.
Every time I happen across The New Yankee Workshop I see 'ol Norm slappin' in some more biscuits. He's selling that little Delta tool. Actually, I don't have anything against Norm. Anyone who's a Delta guy is OK by me. I just want the name of that company he's got money invested in, you know, the one that makes those biscuits.
I only use biscuits or "Splines" when glueing up end grain (like a checker board), plywood, and to reinforce miters.
Edited 12/14/2003 5:28:02 PM ET by Ken's Shop
Ken,
Yes, Norm is just fine. I'd like to meet him. There is room in this world for everything. I'm sure he laughs all the way to the bank. If he infuences 100 new people into this craft and one of them discovers real craftsmanship and art and moves on from the tool envy, gadget-oriented beginning, that's a good thing.
VL
Sharp plane, hide glue, rubbed joint, no clamps...seen this last for a century and more <G>
>> Biscuits add nothing to the strength of a joint noTHING!
If the joint surface area added by the biscuits is greater than the surface area lost to the slots, I don't see how the biscuit could avoid adding some strength. At least in tension and shear, maybe not in bending.
I also seem to remember a joint test in FWW that found that in the particular configuration they were testing, four biscuits were stronger than two.
Edited to add that I'm not defending the use of biscuits in edge joints. I agree that they're not necessary. This is just a quibble over your statement that they add nothing it all.
Edited again to add that I'm assuming the strength of the biscuits is comparable to the strength of the wood being glued. If that's not the case, then it's easier to understand how they could add nothing to the strength of the glue joint.
Edited 12/14/2003 6:44:40 PM ET by Uncle Dunc
Uh oh..another can o worms..ya just had respond..we need to do some serious hi jacking in the future..ha ha ha LMAO..
Uncle Dunc,
The article you refer to was by John D. Wagner titled "Choosing the Strongest Joinery for Doors." I don't know in which issue of FWW it appeared, but I have it in "Practical Design Solutions and Strategies," Essentials of Woodworking, Taunton.
It was presented as a scientific evaluation of methods of construction including M&T, dowel, butt, biscuit, lag bolts, loose teneon. I remember first reading it in FWW and thinking that the editors should have been ashamed for publishing it. It is full of uncontrolled variables, variations in method from one joint to another, complete absence of documentation of the fit of the joints, especially the mortise and tenon joint and inconsistencies that would have embarrased a first year statistics student.
The only accurate statement was the author's comment, "It would be a stretch to call these tests a rigorous scientific examination." Yet that's just how it WAS presented in both Taunton publications.
To complicate any conclusions, the mating surfaces had a tongue and groove in addition to the joint method under study. The M&T joint gapped at 1200 lbs pressure and failed at the glue line at 2700, deflecting more than the loose spline joint when it failed. (huh?)
But two boards mated with just the toungue and groove did not gap at all, shattering the wood, not the joint at 1300 lbs.
"The result? Biscuited joints were the strongest, but M&T construction may still be the best choice. To know why, you'll have to know more about how the tests were run."
I'll say!
The author admitted he thought the M&T joint was glue starved.
One example, and only one, of each joint type was tested. Very poor work. The information is worse than useless. It's misleading.
VL
No, I don't think that's the article I'm remembering. I also remember reading the article you reference, and like you, thought it was pretty cheesy. IIRC, the article I'm thinking of had to do with gluing something on a bed frame.
A very forceful response to the second paragraph of my message, but no reply to the first paragraph. Does this mean you accept the logic of the first paragraph? :)
Uncle Dunc,
No I don't really agree with the logic of your first paragraph. Biscuits do not create more sound gluing surface than the area they eliminate in the joint space. They fracture under stress and while they swell, they do not mate with the mortise like a properly fit spline/floating tenon.
I'm not arguing the point. It IS logical as far as it goes, it just doesn't get translated into real behavior in the joint. And since a well constructed glue line is stronger than the surrounding wood anyway . . .
VL
While we wait for KooKoo to respond, how about we speculate somewhat?
What odds that: the boards were not perfectly straight or every slightly concave? i.e. the edges were slightly convex or even irregular (either along or across the joint) the top was fixed to the base with no allowance for movement the "gap filling" glue wasn't
I stand by my post ...Gremlins I tell ya.
or bojums !
ian,
Yup.
VL
Thanks for all the good info. It looks like the problem could be one of many things or that it could be a little bit of a number of things.
I'm leaning towards thinking the glue just failed (Titebond II) and that I misled myself thinking the biscuits would give added strength. I know I didn't starve the glue joint. And moisture problably was the instigator. I built the top in an unheated garage, then moved it inside the house.
How can I measure moisture and adjust it, to avoid repeating the problem in the future? (Practically speaking that is. For example, I wouldn't want to bring the wood inside for a couple of weeks and then work inside the house - unless I guess it was absolutely the only way).
By the way, some of you wanted additional info. The tabletop is made of nine, pine boards, of which I have developed a tremendous problem along one glue joint. Two more boards have developed less serious separations. All the separations have one end involved but the separations are not all the way the length of the board. The main separation is almost the length of the boards, however -- about three foot separated and about one foot not separated.
The glue was Titebond II.
What is the best glue for this? What is a brand name and where can I get it?
Thanks for the tremendous responses.
KooKoo for Cocopuffs!
AKA Richard
A couple of other points.
I had a breadboard edge around all the boards.
I attached the top to the table with one screw at each corner. I didn't elogate the screw holes, however, figuring that might end up giving the table too much play.
BTW, I would have gotten back sooner but I had to do a little research about the glue type and other details.
Thanks again for all the info.
KooKoo
AKA Richard
You've answered your problem. Your previous post mentioned assembling it outside and then bringing it inside where the relative humidity is probably less. Attaching the top to the apron with screws at the corners and not allowing for movement, the top tore itself apart when it shrank. How did you attach the breadboard end? By gluing it along its entire length (bad) or only along the center (good)?
I have to admit I glued the breadboard edge the whole length of the boards. When I got through with the tabletop the joints looked pretty tight but I probably did pull them together a bit with clamps. That is, they probably looked better when I got finished, than they really were.
Well, thanks for all the great clues. I am going to go through all these posts and make a list of all the tips. When I make my tabletop replacement, I'll apply them all and I'm sure I'll have much better luck.
Finally, does anybody have a recommendation for glues, specifically by brand names?
Also, anyone know of a good value brand moisture meter?
Thanks,
KooKoo
AKA Richard
KooKoo,
No matter what the edge treatment was, or the glue used, or the moisture content of the boards, or anything else for that matter, if you glued the breadboards onto the ends, the top would have split.
I would guess that 90 percent of the posts to your question wouldn't have been made if it had been known, at the start, that the top had breadboard ends, somebody should have asked.
There are two good articles in the October 2003 issue of Fine Woodworking, one concerning wood movement and the second on how to prepare lumber. You should get the issue, the information in those two articles alone will save you a lot of trouble in the future.
You don't need a moisture meter unless you are going to be doing a lot of work with thick planks where you also don't have any information on how they have been dried and stored.
If you are working with stock that has been handled carefully, and been stored out of the weather, whether kiln dried or air dried, the wood just needs to be brought into the shop and stickered for a few weeks before you use it to stabilize its moisture content with the indoor environment. As long as the wood's moisture level is stable when you start working with it, the furniture will hold up well if it is properly designed to handle humidity changes.
John W.
Edited 12/17/2003 1:20:51 PM ET by JohnW
>>Finally, does anybody have a recommendation for glues, specifically by brand names?
Titebond is certainly one of the best all-around assembly glues. That type of glue is almost universally used in industry for generally assembly. In my shops we glued up literally thousands of panels from 4/4 to 8/4 pine and hardwoods over a period of twenty years, with never a reported glue failure. Did get some end checking once in a while, due to poor moisture control. We made a lot of restaurant table tops out of 8/4 red oak, and its really hard to find suppliers that can or will dry it uniformly.
>>Also, anyone know of a good value brand moisture meter?
I think almost any inexpensive meter would do. I like the little Mini LIgno ones (kind of expensive at around $150), but they're all pretty good. As John W. pointed out, they're not strictly necessary, but they are quick and convenient and can save you from problems later on. In twenty years of manufacturing furniture, moisture problems were one of our biggest and most persistent headaches. In your case, you don't have to do much, just wait until the moisture content (internal, not surface) quits going down, and make sure all the boards are within about 1/2 % of each other.
You can indeed make furniture successfully with high moisture content wood, as long as you are aware of it and make allowances, but it's a lot easier when the wood is in equilibrium with its intended environment. I'm sure you have noticed all the furniture that comes over from Southeast Asia and India, and promptly cracks and warps in American houses. Especially the fake antiques. Wet wood, in equilibrium with the environment where it was made, but way too wet for American central heating and air conditioning. If you build with wet wood, the wood must be very straight grained, and cut straight as possible to the grain. Any grain runout will tend to cause bowing, twisting, and crooking as it dries. Flat sawn wood from the outer parts of the tree will tend to cup. Having the wood properly dry tends to minimise these problems.
The equilibrium moisture figures I gave above are a very general industrial standard. They vary somewhat by species and local climate.
The tips about roughing your lumber to a little bit oversize to let is stabilize are very good, too. It is general practice in a lot of small shops.
HTH
Michael R.Correction on the Mini Ligno. Amazon has the Mini Ligno E on sale for $105. Good value, and perfect for your needs (and mine). Measures 6 to 36% moisture content. I think you can set them for the specie of wood you are checking, rather than using a cinversion table. At least the one I had (stolen) would do so.
Edited 12/17/2003 7:32:01 PM ET by Woodwiz
Edited 12/17/2003 7:37:00 PM ET by Woodwiz
I think Ben has it mostly nailed. That kind of end checking usually results from wood shrinkage with changes in moiature content.
When you are gluing up panels, the wood shouuld be at the equilibrium moisture content for where it is going to stay. For pine, that's around 11%, most hardwoods, 6%. Equilibrium moisture content outdoors is much higher, and there is a lot of wood movement between those levels. A pine 1 x 10 will move on the order of 1/8" from 19 to 11% moisture content. The only solution for you is to keep the wood inside for several weeks before working it. I am sure there are articles in the archives or on the web for determining the moisture content using an oven or a microwave, and there are decent moisture meters available for not a great deal of money. Be advised that all the softwood lumber you will find at most lumber yards and big box stores will be way too wet to use for furniture, and will require a lot of acclimatization before you can use it with confidence. Softwoods bought at a hardwood lumber supplier have a lot better chance of being drier.
The second thing is that when you glue up panels, they have to fit well before gluing them up. That means no daylight to be seen anywhere along the joint, or, at most, a very little gap or "spring" in the middle of the joint - 1/64" or so, depending on the length of the joint, the reasoning being that ends always seem to shrink more than the middle of boards in a panel. You can't force the joint together with clamps and expect it to hold.
Third, you may have been below the "chalking" temperature of Titebond, which the manufacturer says is around 55 deg. F. The glue doesn't work as well below that temp.
The point about attaching the breadboard end is sound, too. Gotta let the wood move, because you can't stop it.
Michael R
KK,
One question I haven't seen raised in the discussion is how long was it between the time you machined the rough lumber and assembled the table top? If there were any stresses in the wood, it might have taken awhile for the boards to reach a steady state. Straight boards cut from crooked trees may be held straight during the drying process by the restraints put on the pile of lumber in the kiln or while air-drying, e.g., banding or weight added to the top of stack. After machining a lot of "straight" boards may curl, twist or cup because of the internal stresses that are still there. It is best to let freshly dressed lumber set a day or two before doing any assembly --- of course that goes directly against the advice about doing glue ups as soon after machining as possible to avoid oxidation and other surface changes that cause poor glue adhesion.
bd
bd,
It's best to "rough mill" lumber to slightly oversize and let it settle down and move if it's going to for a few days or later final fitting and assembly. Then final jointing and planing removes the last 1/16" or so of the oversize without much more danger of wood movement. If it's not glued up at that point, later a last miniscule shaving of the joint surfaces before glue-up is all that's needed.
VL
<<how long was it between the time you machined the rough lumber and assembled the table top? If there were any stresses in the wood, it might have taken awhile for the boards to reach a steady state. Straight boards cut from crooked trees may be held straight during the drying process by the restraints put on the pile of lumber in the kiln or while air-drying, e.g., banding or weight added to the top of stack. >>
I can't remember what the answer is to this question. It has been some time since I made the tabletop and I've just forgotten. It is a consideration that I'll certainly consider in the future however.
Thanks,
KK
This forum post is now archived. Commenting has been disabled