Dowel Joinery Opinions

[Tony Shea]

I am currently thinking about building a cabinet. I have read most of Krenov's books and am basing the design a little on some of his work. I originally planned on using dowel joinery for the carcass as Id like the top and bottom of the cabinet to over hang the sides slightly and also to create a partially inset door. Basically using the joinery methods that Krenov used extensively. But I am presently reading a book by George Nakashima in which he states that dowel joinery should never be considered in fine furniture.

I am just curious what the general consensus is here on the merit of dowel carcass joinery. Would it be worth the extra effort to go through with M/T joinery in place of the dowels or would dowels suffice in a fine cabinet that is meant to last 100 or more years?

[Zach Dillinger]

I don't think there is a woodworker alive who wouldn't think of Krenov's work as "fine furniture". I've used dowels before and I've never had a problem with them. They go together really quickly and, when used right, you cannot tell the difference. My only caveat is to make sure that the dowels aren't too close to the face of a board, as they are sometimes noticeable as a bump under the surface.

[Don Dorn]

I think that dowels can certainly be used in fine furniture. However, in my opinion, there is both and upside and downside with the upside being that the joint will be piston fit. The joint will also be very strong if multiple dowels are used and they help square things up.

The downside is that glue is the only long term holding factor - eventully it after it fails, the joint will too as it isn't self locking like a M&T with a drawbore or dovetails. Granted, that will be after we are done looking it over in this lifetime, but if the piece is handed down over generations, someone will have to fix it someday. The good news is that it can easily be done, but I just don't think it will ever hold over time like some other traditional joints.

[Jim Koepke]

I am presently reading a book by George Nakashima in which he states that dowel joinery should never be considered in fine furniture.

Does he say why?

I have worked in situations where statements like this about techniques were made with no explanation. When I asked why, I was told the subject was not open for discussion.

I have used dowels in joinery. My use was usually for edge joining or with a dado for right angle joining. One person who had me repair and build furniture gave me his opinion on dowels. He did not like flush cut dowels since that tended to look like a mistake someone was trying to hide. His opinion is the main reason I tend to leave dowels standing proud.

Back to the original question, I might consider using through mortise and tenon with proud tenons and a proud dowel through the tenons. Maybe something in the style of Greene & Greene.

One of the problems with overhanging edges on cabinets is when it comes time to move most people including many "furniture movers" will tend to use the top as handles to lift the piece instead of lifting at the base. If the top is not secure, the it can lift off. That is a nasty surprise when going down stairs.

jtk

[Jon van der Linden]

If the cabinet is small and light with thin sides like most of Krenov's were, then dowels will work just fine. If you have some reason for being worried about the strength, the only practical alternative is wedged through tenons.

I doubt that there will be more problems over time with a properly executed dowel joint than with the alternative. Generally looking at the (very) long term, glue failure will have the same results in each case. More likely things will go fine until there is some kind of catastrophic failure... like when I saw two technicians try to open a Riesener desk that had swollen from high humidity... they split its front in half! Let's just say the museum director who was showing me around wasn't pleased.

Bottom line, I wouldn't worry, at least not about that detail. There is much more chance of someone unintentionally destroying it than having the piece self destruct.

[Mike Henderson]

You can think of dowels as a form of loose tenon. However, when you compute the long-grain-to-long-grain gluing surfaces, you have to put in a lot of dowels to equal what you'd get in a loose tenon.

Furniture manufacturers (especially cheap furniture) used dowels in places where they're not at their best, such as attaching the chair seat to the back rails. The joint will almost always fail and people blame the dowels. In that case, they're correct because there's just not enough surface area - a mortise and tenon would be much better.

Mike

[Prashun Patel]

I loves me some Nakashima, but in my humble opinion, "Fine Furniture" is a matter of humble opinion.

I bet woodworkers 150 years ago would have cringed at the thought of using yellow glue in fine furniture.

Sam Maloof used screws and dowels in some of his chair construction. If that ain't fine, I don't know what is.

Dowels are plenty strong if used properly - at least that's my opinion.

[Mike Henderson]

I loves me some Nakashima, but in my humble opinion, "Fine Furniture" is a matter of humble opinion.

I bet woodworkers 150 years ago would have cringed at the thought of using yellow glue in fine furniture.

Sam Maloof used screws and dowels in some of his chair construction. If that ain't fine, I don't know what is.

Dowels are plenty strong if used properly - at least that's my opinion.
If yellow glue had been available 150 years ago, our woodworking ancestors would have been all over it. Our ancestors were pragmatic people, just like we are today.

Mike

[Pam Niedermayer]

I'd say that using small round objects for attaching furniture parts is up to the maker. Of course, I'd try to avoid that furniture when buying (Yeah, right, when have I bought furniture in the last 40 years? Never.); but I'd also say that there's a big difference in quality between purchased dowels and hand made, whittled, tree nails. Dowels today are pretty crummy, short grained devices; whereas tree nails are long grain, pretty good connectors.

Personally, I prefer M&T in most cases.

Pam

[Bob Lang]

I'll be the one to step forward and say "but the Emperor isn't wearing any clothes." Krenov may be everyone's hero, but Nakashima had a better understanding of putting pieces of wood together and having them stay together for a long, long time. Dowels aren't the way to do it because of what happens as they shrink and swell in response to seasonal wood movement. Many years ago R. Bruce Hoadley published an article in Fine Woodworking subtitled "Why Round Tenons Fall Out of Round Holes" that explains it thoroughly. The short explanation is that dowels become out of round after a few seasons, and what little long grain to long grain glue area there was to begin with is reduced to one or two small points of contact. At that point dowel joinery is doomed to failure.

Dowel joints may get by for a while, and the failure may be delayed by using a bunch of them (as in some of Krenov's cabinets) but for the long haul Nakashima is correct. Go to an auction or antique mall and look at a bunch of furniture made 80-100 years ago. If you see a joint coming apart, chances are there is a dowel in there instead of a proper joint. If you're trying to build something to "last 100 or more years", put down the drill.

Bob Lang

[Jim Koepke]

The short explanation is that dowels become out of round after a few seasons, and what little long grain to long grain glue area there was to begin with is reduced to one or two small points of contact. At that point dowel joinery is doomed to failure.

Bob,

Thanks for giving us a good "why" to answer the why not?

jtk

[Sean Hughto]

Dowels aren't the way to do it because of what happens as they shrink and swell in response to seasonal wood movement. Many years ago R. Bruce Hoadley published an article in Fine Woodworking subtitled "Why Round Tenons Fall Out of Round Holes" that explains it thoroughly. The short explanation is that dowels become out of round after a few seasons, and what little long grain to long grain glue area there was to begin with is reduced to one or two small points of contact. At that point dowel joinery is doomed to failure.

Bob, I respectfully would like to suggest you are overstating things here. Are any early (now over 50 years old) Krenov pieces known to be failing - or degrading at all? I'm going to guess no.

My personal experience of somethings I've made with "cylindrical floating tenons" is that they hold up extremely well. I've even used them in a step stool that has gotten heavy use for years with absolutely no degradation at all.

The failing antique dowel jobs I have seen tend all to be dresser drawer joints that use smallish dowels. Drawers are often over loaded and treated roughly. I would agree that dowels are not good choices for that application, but even there, I'm guessing a good woodworker who made appropriately sized dowel and otherwise took care in gluing and executing the joint might well make dowels succed even in a drawer box.

Here's a very early shop furniture effort of mine - about 20 years old now - where I used dados reinforced with dowels. Not the slightest hint of degradation much less failure.
http://farm3.static.flickr.com/2185/2233245819_4eeaaf9918_z.jpg?zz=1

[Dave Anderson NH]

Since my furniture is primarily Queen Anne and early Chippendaale (Rococco) I'm a bit of a traditionalist and only half jokigly say, "Friends don't let friends use dowels." Bob Lang is correct and to carry things further think about situations where you are running the dowel long grain in one piece of wood (pretty good bond) and cross grain into the other piece of wood (not so good). Couple this situation with limited glue surface and the only way to make up the surface area deficit is with lots of dowels. More drilled holes and dowels to my mind equals more opportunities for something to get out of alignment and make assembly more difficult. One of the basic tenets of the adhesive world is that tensile, cohesive, and shear strength is only partial dependent on the type of adhesive. Increased surface area increase the bond strength of all adhesives. My personal choice as someone else said is a foxtailed or wedged set of tennons.

[Bob Lang]

Hi Sean,

Not to quibble, but the joints you show aren't what I would consider a dowel joint, those are dados with dowels used as pegs for reinforcement. Not the same thing as it would be if the only connection between the bottom, shelf and uprights were dowels. Most of the failures I see are face frames and doors falling apart, and rails and aprons ready to leave table legs. Feel free to disagree, but I'd suggest reading what Hoadley has to say on the matter.

Bob Lang

[Mike Henderson]

I'm no fan of dowel joinery but let me make a few comments, hopefully balanced between for and against.

I want to compare the long-grain-to-long-grain gluing area of dowels to a tenon. A dowel is a cylinder and the formula for the area of the sides of a cylinder is 2*pi*r*h, where r is the radius and h is the height of the cylinder. Since the dowel is usually going into end grain, only part of it is long grain to long grain. As Bob pointed out, theoretically there are only two lines that are exactly long grain to long grain but right next to that line there's still long grain to long grain, with a little bit of end grain. This continues around the dowel until you reach the 90* point where it's long grain to end grain. I maintain that the effective long grain to long grain is half the surface area of the cylinder.

So let's compare 3/8" dowels to a tenon. I'm going to assume a joint such as the chair seat into the back rail and further assume that the tenon can be made 2" wide. You can do the calculations if you want to make other assumptions. I'm further going to assume that the dowels have to go into the same space as the tenon.

Next, we have to make some assumptions about how deep you're going to sink the dowels. Whatever I use for the dowels, I use the same depth for the tenon.

So here are the numbers, calculated on an Excel spreadsheet:

1" depth
Each dowel has effective gluing area of .589 sq in
The tenon has effective gluing area of 4 sq in (2" by 1" on each side)
It would take 7 dowels to equal the tenon

1.5" depth
Each dowel has effective gluing area of .883 sq in
The tenon has effective gluing area of 6 sq in (2" by 1.5" on each side)
It would take 7 dowels to equal the tenon

In fact, for this specific example, a 2" tenon, it will always take seven 3/8" dowels to equal the gluing area of the tenon.

For a 1/2" dowel, it will take 6 dowels to equal the same tenon.

Can you put 6 or 7 dowels into a 2" space? Do people put that many dowels into the joint I described? Most chairs I've seen have two dowels in them at that joint.

Mike

[Sean Hughto]

As I said, I don't disagree with the general proposition that dowels are not good choices in many high stress situations - drawers, doors, and aprons where lots of racking, impact, movement, etc. is expected. What I said was that I think you are painting with too broad a brush - overstating things. "Never a dowel a fine furniture" as as the sentiment seems to be shaping up for some on this thread is overkill and arbitrary.

Also, I was questioning your statement about dowels failing because of shrinkage in a matter of a "few seasons" (I think that is way too short a time span). If they come out of round and that causes their failure in such a short time, shouldn't that happen in every use of dowels with the exception for applications like drawbore pinning where the dowel is held in place by the surrounding pieces?

I just think we ought to use our common sense and experience to guide joinery decisions for given circumstances and applications rather than overgeneralize with rules about "never."

[Tony Shea]

All great points here and Bob, I really like your explaination. I have seen no where that someone has had any kind of failure with any of Krenov's cabinets with dowel joinery. But then again I'm not sure any one would report such a failure in a public fashion that we could hear of.

After all the input I'm still unsure of what my joinery method will be. I do think that using M/T joinery is going to be the way but still have not got rid of the idea of using dowels.

[Mike Henderson]

I think Sean (above) made a good point. There's nothing wrong with dowels in certain applications. There's nothing wrong with biscuits in certain applications. But for joints with lots of stress, such as on chairs, M&T is a better choice. You just have to use your judgment.

Mike

[Frank Drew]

How are Krenov's wall-hung cabinets attached to the wall? Is there any support under the bottom?

[Jon van der Linden]

How are Krenov's wall-hung cabinets attached to the wall? Is there any support under the bottom?

They're hung from small brass plates which are inset into the back of the case. They have a keyhole type hole in them and are essentially hung from two screws in the wall.


Most of the failures I see are face frames and doors falling apart, and rails and aprons ready to leave table legs. Feel free to disagree, but I'd suggest reading what Hoadley has to say on the matter.

This is where we go back and look at the question... and see that the construction asked about is none of the above. Then we look at Hoadley's article from 30 years ago... and see that if each dowel had the strength he considered failure, it would take 560 lbs to pull the joint apart! If everything was properly prepared and good adhesives used, it would take much more than this. To be blunt, you're simply wrong.

Although I personally seldom use dowel joints, they are in my experience very suitable for the posters application. They are significantly less suitable for many of the applications which are erroneously being used for comparison.

[george wilson]

Dowels do not hold as well as mortise and tenon joints. Yet,there are low stress applications where they will do. Certainly not for use in holding chair legs on,or in other places where stress and some twisting might cause them to fail. I have seen cheaper type old Queen Ann style chairs made in England just fall apart because their legs were doweled on.

[Johnny Kleso]

I always thought I would like to try this Bead Lock Jig
http://www.rockler.com/product.cfm?page=18092&filter=dowel%20jig

197365

[glenn bradley]

I skip dowels for the reasons stated but, Krenov and Maloof used them and I consider their furniture "fine". I stick to loose tenons for joints that require that sort of thing due to the increased long face grain. For the joinery you describe I use sliding dovetails. There is more than one way to skin the cat. Use what works for you.

[Frank Drew]

[Krenov's cabinets are] hung from small brass plates which are inset into the back of the case. They have a keyhole type hole in them and are essentially hung from two screws in the wall.


Thanks, Jon.

But with regards to his hanging cabinet design featuring sides doweled to a bottom that's often proud of the case sides, I'd be very nervous, over the long term, putting anything heavier than a sake cup or maybe a couple of dried flowers on that bottom. Bottom doweled to the sides, not my preference, but at least the joinery method is working with the dowels' sheer resistance against any reasonable load.

Did he ever do through dowels on alternating angles (sometimes called dovetailed doweling)?

[jamie shard]

So here are the numbers, calculated on an Excel spreadsheet:
...
In fact, for this specific example, a 2" tenon, it will always take seven 3/8" dowels to equal the gluing area of the tenon.

For a 1/2" dowel, it will take 6 dowels to equal the same tenon.


I really like this analysis, thanks!

... since no one has mentioned it yet, I think the book "Understanding Wood" talks about dowel joinery and how dowels become oval over time. One "solution" is to have a split in the dowel. That way the glue line remains intact around the outer surface of the hole/dowel and the gap opens up in the center of the dowel.

[george wilson]

Dowels are oval when NEW,not just over time. If you have a dial caliper,or even a plain caliper,roll some around between the jaws and see. Orienting the grain direction of the dowel ACROSS the grain direction of the hole it's going into MIGHT help some,since the dowels shrink with their grain much more than across it(Average is supposed to be 19X more across the grain than with.) I am sure that figure varies with species.

[Phil Thien]

How much is there to this "dowels become ovals over time" thing?

George Wilson is right, a lot of dowels have an oval cross-section when new.

And I understand that flatsawn wood shrinks more than radialsawn. But it also EXPANDS more, right?

It has been a while since I read the incredible shrinking dowel article, I will have to dig it up, but...

It seems to me that not only would the dowel be shrinking, but the wood around it would be shrinking, too.

Furthermore, it would seem to me than M&T would suffer the same fate.

BTW, if you calculate the shrinkage of a 3/8" dowel, you will probably come up with about .003 to .006" depending on species and upper/lower moisture contents. But wouldn't the HOLE shrink a nearly identical amount?

And, if you look at expansible dowels, I imagine they are compressed more than that when new.

I guess I have to read that article again. This would have been a neat science fair project in grade school.

[Jon van der Linden]

Thanks, Jon.

But with regards to his hanging cabinet design featuring sides doweled to a bottom that's often proud of the case sides, I'd be very nervous, over the long term, putting anything heavier than a sake cup or maybe a couple of dried flowers on that bottom. Bottom doweled to the sides, not my preference, but at least the joinery method is working with the dowels' sheer resistance against any reasonable load.

Did he ever do through dowels on alternating angles (sometimes called dovetailed doweling)?

I think it's very important to understand what we're really looking at. Photographs of an object often don't tell you some of the most important things about it. In the case of these cabinets, what people seldom realize are what the dimensions are (many of the dimensions given in his books are incorrect). The range for depth is about 4-1/2" to 6-5/16" with most around the 5-1/2" mark (for the type of cabinet under discussion). The amount of stress that any weight can place on 7 dowels per joint in that depth is negligible. The cabinets are quite small and delicate in appearance, but more than robust enough to withstand use.

In keeping with the small dimensions, the sides, top, bottom, and door are actually quite a bit thinner than one might expect from just looking at photographs. The cabinets give a feeling quite a bit different than the heavy examples most people make when they "copy" his work without thinking or understanding. I think your description of holding a sake cup is an accurate one, they clearly aren't meant to hold books or stacks of dishes, they are for special items.

As to the construction questions, the dowels are not through dowels, so they can't be angled. The sliding dovetail suggestion also doesn't work... simply not enough material there for that to be a viable option for Krenov's designs.

[george wilson]

How much dowels shrink no doubt relates on how dry the wood was when they were made. Hopefully,I'd like to use wood that is good and dry,but I can't vouch for who(probably imported) made the dowels. They are most often not round,though.

IF you use properly dried wood, your mortise and tenons cannot be compared apples to apples with dowels. The only way to get round dowels is to make them yourself from well dried wood,I guess.

[Phil Thien]

In the case of these cabinets, what people seldom realize are what the dimensions are (many of the dimensions given in his books are incorrect). The range for depth is about 4-1/2" to 6-5/16" with most around the 5-1/2" mark (for the type of cabinet under discussion).

Are you sure about that? I knew they were small, but not THAT small?

[Jon van der Linden]

Are you sure about that? I knew they were small, but not THAT small?

Talking about the wall hung cabinets with the dowel assembly... yes, they are that small. I think there are depth dimensions in his first book. (Not all the dimensions are correct, so you do have to think a little just to verify that it's not way off.) The width is generally within the size of a standard sheet of paper, 8-1/2" to about 11" although there are some narrower and wider. Their scale and lightness reminds me a lot of some 18th c. French work.

[Trevor Walsh]

I'm building a small wall cabinet at the moment and am after the overhang you're interested in. I too had some doubts about the dowel, but in a case piece like this (mine is 7" deep, 16" tall and 12" wide) they are okay, and plenty sound enough for the keys that will be going into it. If however you were making a case for your brick collection to hang on the wall, I'd say you have to go with dovetails, and apply molding or something similar to get the needed strength.

[David Keller NC]

I am just curious what the general consensus is here on the merit of dowel carcass joinery. Would it be worth the extra effort to go through with M/T joinery in place of the dowels or would dowels suffice in a fine cabinet that is meant to last 100 or more years?

One aspect of your question may be self-answered. If you are interested in making a piece (any piece, not just a cabinet) to last the ages, then picking from the list of standard techniques based on saving effort may not be the best choice. That doesn't necessarily mean that one would want to use a non-standard method that is lots of extra effort just for the purpose of spending extra effort (think claw-type through dovetails that have to be carved). But in the plethora of routine methods available to a woodworker, more effort mostly means "better".

An example - there is nothing structurally wrong with biscuits or pocket-screw joinery, but that doesn't mean that I would use it on a signed piece of furniture. Shop cabinets, yes, but queen anne highboy, no.

[Jon van der Linden]

One aspect of your question may be self-answered. If you are interested in making a piece (any piece, not just a cabinet) to last the ages, then picking from the list of standard techniques based on saving effort may not be the best choice. That doesn't necessarily mean that one would want to use a non-standard method that is lots of extra effort just for the purpose of spending extra effort (think claw-type through dovetails that have to be carved). But in the plethora of routine methods available to a woodworker, more effort mostly means "better".

An example - there is nothing structurally wrong with biscuits or pocket-screw joinery, but that doesn't mean that I would use it on a signed piece of furniture. Shop cabinets, yes, but queen anne highboy, no.

To quote Ruhlmann, "your craftsmanship is holding you back!"

You've really got it backwards here. It's about design. It's about a specific stylistic choice which does something that's the opposite of what's structurally sound. It's about some very thin light pieces of wood. It may actually be that dowels are stronger in this particular application, something you don't seem to consider.

Keep in mind that the "overhang" of the top and bottom is probably only 1/16" or so, just enough to exist. Making a through mortise will change the aesthetics of the piece, and because of the very thin amount of material on one side of it, a mortise may actually be weaker than a dowel. Whether you like his aesthetics or not, Krenov was thoughtful about his choices and certainly capable of making mortises if he thought it was appropriate.

[David Keller NC]

You've really got it backwards here. It's about design.

That's actually precisely what I meant - there is nothing structurally wrong with biscuits or pocket screws, and in some cases they would be the best method. But I would not use them for aesthetic reasons.

Though if I chose to make a cabinet with so little overhang and out of thin stock, I would personally do one of two options - dovetail the case together an apply a small covering molding (the most probable route), or if it was intended to hang from a wall so that the top and bottom would not typically be highly visible, I would screw the case together and plug the holes. If done carefully so that the grain lined up, they would be difficult to see.

But I would know they were there, so option #1 would be the high likelyhood outcome.

[Chris Fournier]

I don't use dowels for much in my work, I've never really thought much of them. I have used them for registration pins on furniture that was intended to be assembled and dissassembled during it's lifetime.

To join a carcass like a Krenov cabinet I would reach for a biscuit joiner. You can use the BJ on the overhanging top no sweat.

[Jon van der Linden]

I don't use dowels for much in my work, I've never really thought much of them. I have used them for registration pins on furniture that was intended to be assembled and dissassembled during it's lifetime.

To join a carcass like a Krenov cabinet I would reach for a biscuit joiner. You can use the BJ on the overhanging top no sweat.

Riiiiiight... so you would use a biscuit joiner "no sweat" on a 3/8" thick piece of wood??? :eek:

[Chris Fournier]

Riiiiiight... so you would use a biscuit joiner "no sweat" on a 3/8" thick piece of wood??? :eek:

Yeesssss, you said 3/8" Jon, I've yet to see a Krenov carcass made with 3/8" thick stock except for perhaps a backer panel and I'm sure that at 3/8" that would be in a rabbeted channel. I'm not sure that you have a point except that you don't like biscuit joiners perhaps?

[Jon van der Linden]

That's actually precisely what I meant - there is nothing structurally wrong with biscuits or pocket screws, and in some cases they would be the best method. But I would not use them for aesthetic reasons.

Though if I chose to make a cabinet with so little overhang and out of thin stock, I would personally do one of two options - dovetail the case together an apply a small covering molding (the most probable route), or if it was intended to hang from a wall so that the top and bottom would not typically be highly visible, I would screw the case together and plug the holes. If done carefully so that the grain lined up, they would be difficult to see.

But I would know they were there, so option #1 would be the high likelyhood outcome.

You're still placing construction first and aesthetics of the piece second.

Adding applied moldings would give an entirely different feel to the work, as would "difficult to see" holes, whether for screws or mortises. If you make a case for these solutions improving the aesthetics of the piece as a whole, then you might have a point.

As far as mechanics, the dowels work very well. Not sure why you insist on finding a different solution, especially one that fundamentally affects the aesthetics.

[Tom McMahon]

There are two separate discussions going on here. One is about the engineering of a cabinet or how to build the strongest cabinet. The other is about the visual design of a cabinet or how to execute a visual statment with engineering that is adequate and does not affect the statement.

[Tony Shea]

Trevor,

Have you actually started the case construction yet? I'd be interested in the method you use for your dowel registration. I was probably just going to make my own shop made jig such as Krenov shows in his writings with a registration tab on the end of it. But I've seen a jig on the market for dowelling called the Joint Genie that looked very similar to the design Krenov uses on his shop made version. The Joint Genie might prove to be slightly more accurate. But the last jig I purchased has yet to see the light of day and probably so would this thing. Just curious how you plan to execute or how you have executed your joinery for this case.

[Jon van der Linden]

Yeesssss, you said 3/8" Jon, I've yet to see a Krenov carcass made with 3/8" thick stock except for perhaps a backer panel and I'm sure that at 3/8" that would be in a rabbeted channel. I'm not sure that you have a point except that you don't like biscuit joiners perhaps?

So I guess you've seen the big sign that says no biscuits at the College of the Redwoods (aka the Krenov school)? But I'm a rebel, so I do have a biscuit joiner. In general though, if the stock is thick enough for a biscuit there are better ways, although few are faster.

Regarding my math and your experience with Krenov's work... obviously I have no way to judge your experience, but you do imply that you've at least seen his personal work being restored, because how else would you know how thick his backer panels are? I don't have that kind of experience, but I can show a photo.

3/8" is roughly 1 cm, in the photo it's clear that the piece being doweled is thinner than the guide block which is about 1 cm. So yes, I did say 3/8" because that's what it is.

197716

[Chris Fournier]

So I guess you've seen the big sign that says no biscuits at the College of the Redwoods (aka the Krenov school)? But I'm a rebel, so I do have a biscuit joiner. In general though, if the stock is thick enough for a biscuit there are better ways, although few are faster.

Regarding my math and your experience with Krenov's work... obviously I have no way to judge your experience, but you do imply that you've at least seen his personal work being restored, because how else would you know how thick his backer panels are? I don't have that kind of experience, but I can show a photo.

3/8" is roughly 1 cm, in the photo it's clear that the piece being doweled is thinner than the guide block which is about 1 cm. So yes, I did say 3/8" because that's what it is.

197716

Wow, you play pretty fast and loose with facts and logic Jon.

If stock is thick enough to take a biscuit, and there are very small biscuits available, well below the 0 size then what are the better methods that are slower? Part of the efficacy of the biscuit joint is the speed of the process. If it holds, its fast and it suits the project construction what would you recommend that would be better? And by better I mean what would hold as well as the biscuits and be easier to accomplish?

Regarding math and experience I'm at a loss regarding your conclusions. I have indeed seen several Krenov pieces in the hands of friends who collect furniture. I have had time to pour over them and - gasp - even touch them! While I am not a Redwoods grad, I have met and talked with Krenov and I asked him about the dowel joinery and he stated that it let him build the look that he wanted with a minimum of fuss. I should add that this was before biscuits were about. I can't agree with your extrapolations taken from the photo you attached either. While the photo is pretty fuzzy I'd say that you're looking at 5mm holes, the bottom of their circumference being again about 5mm from the table surface and the top of the guide being about 8mm above the holes for a total height of the guide of 18mm or 5/8". The piece being doweled is harder to make out but I'd say that it's at least 1/2 thick.

Nonetheless it seems that the sum of your experience with this matter comes down to having read the same books that I did. Attitude is no attidote for a lack of facts.

[Phil Thien]

3/8" is roughly 1 cm, in the photo it's clear that the piece being doweled is thinner than the guide block which is about 1 cm. So yes, I did say 3/8" because that's what it is.

197716

Something doesn't make sense.

Krenov is a big fan of knife hinges.

He mortises the hinges into the cabinets and doors.

If you look closely at the proportions of the knife hinges to the stock, it is clear that the typical stock thickness has to be 5/8" to 3/4" thick.

If we were to insist that the stock were 3/8", that would be the knife hinges were 3/16" wide.

Furthermore, the drawers and inside doors are approx. three quarters to one-half the thickness of of the carcase. So you'd be talking about 3/16" to 1/4" thick components there. That doesn't make sense.

What do you think?

[David Keller NC]

You're still placing construction first and aesthetics of the piece second.

To me, "aesthetics" is a lot more complex than the look, surface ornamentation (or the lack thereof) and proportions of a piece. How it's constructed has at least as much contribution.

[Trevor Walsh]

Tony,

Yes, the carcass is done for the most part, I have the outsides to smooth the knife hinges to install and that's it, of course I can't do any of those things until the door is made so at that point the whole thing is about done.

I used a shop made poplar slip planed to match the thickness of the sides which are about 1/2" inch. I'm using seven 1/4" dowels on each edge.

I looked at the joint genie too, when planning the build, but after thinking about it, I'm going to pay $200 for a bar with some holes in it (albeit a nice bar with holes in it) when the master uses a two pieces of wood and a few screws/glue and nails? Am I a woodworker or what? I've since spent that $200 on hardware, wood, chisels and other hard to make goodies.

The alignment isn't all that hard either, judging by your writings I'm going to guess you've read his books or at least seen a lot of pictures of the process? If you follow his explanation from pages 132-140 in The Fine Art of Cabinetmaking diligently you should be good, now I used a hand drill rather than a drill press and horizontal boring machine, and my accuracy is pretty good.

I would make the following suggestions and some loose rules after going through this...

1. Make absolutely sure you have flat, square stock to work with
2. Find a drill and dowel combo that are a good fit in both face and end grain. I tried three different 1/4" drills until I found the right one.
3. Probably use something like maple for the jig, drill it in the drill press (I did mine in the DP and drill the holes in the carcass with a hand drill) Everything worked out, but I was worried that the poplar was wearing too much, it's a small amount of wood, go for the maple.
4. DO NOT, I mean really, completely dowel the joints for the test fit. It's a bear to get apart, you shear off dowels then have to drill them out (scary) or risk splitting the sides. I use 8 one in each corner front and back to test the case and fit the back/door. Don't worry, if you took care with the jig, drilling, and matching the orientation of everything the holes will line up.

I think that's it, it sounds worse than it is, just mark everything, do it the right way and it'll go together like butter and toast.

[Tony Shea]

Thanks Trevor, very helpful. And yes I am very familiar with all Krenov books. I agree that the $200 price tag does not settle well for such a simpley made jig that seems like I can make myself easy enough.

Also, what is a good source for good quality dowels? I am not about to use those terrible excuses for dowels at the local Borg. I've made a couple planes useing those dowels for allignment purposes and they really aren't close to round and not close enough to the matching drill bit size. I tried every drill bit I had plus I bought a couple others and still had loose fitting dowels. I was able to get past this issue with my planes though but do not feel the same about carcass construction. Hence my original question about this style of joinery. I unfortunatly do not have the option to turn my own dowels. And its too hard to get a straight dowel out of a dowel plate due to variation in grain. Although those dowels sure are strong compared to turned dowels.

[Trevor Walsh]

I got my dowels from McMaster.com oddly enough, search for "wood dowel pins" and they come right up. Anything else and you'll never find them. they were cheap, and in my case round and well made. They have the straight for particle board and spiral for solid wood.

Glad to help, I can't wait to see what you've got cooking there. What wood are you thinking for the cabinet?

[Adam Cherubini]

How much is there to this "dowels become ovals over time" thing?
And I understand that flatsawn wood shrinks more than radialsawn. But it also EXPANDS more, right?

I don't think so. Wood shrinks over time. The shrink swell cycles aren't dimensionally equal.



BTW, if you calculate the shrinkage of a 3/8" dowel, you will probably come up with about .003 to .006" depending on species and upper/lower moisture contents. But wouldn't the HOLE shrink a nearly identical amount?

The depends on the grain orientation. When joint woods that are oriented perpendicular to one another, its likely one board won't have favorable orientation.

There was discussion about 18th c dowels on the SAPFM forum some time ago. People look at old pegs and assume they were rough shaped or square pegs (in round holes). The Conservators at the Philadelphia Museum of Art believe the squarish shapes are the result of round pegs shrinking. Remember, according to Hoadley dowels don't just go oval. The skinny top of the oval will square off as the rings seeks to straighten themselves out. People thinking they are getting 360 degrees of gluing surface are just kidding themselves.

Adam

[Tom McMahon]

Here is a picture of what Adam is referring to. You can see that the dowel has also begun to work it's way out of the hole. This is from a chest over 200 years old, but you can begin to notice dowels moving in as little as a couple of years in some cases. I doubt that you would see any significant problems in your own lifetime but it is possible. The Krenov pieces arn't old enough to see the problems yet.
http://i44.photobucket.com/albums/f14/millcrek/013.jpg

[Jon van der Linden]

Wow, you play pretty fast and loose with facts and logic Jon.

If stock is thick enough to take a biscuit, and there are very small biscuits available, well below the 0 size then what are the better methods that are slower? Part of the efficacy of the biscuit joint is the speed of the process. If it holds, its fast and it suits the project construction what would you recommend that would be better? And by better I mean what would hold as well as the biscuits and be easier to accomplish?

Regarding math and experience I'm at a loss regarding your conclusions. I have indeed seen several Krenov pieces in the hands of friends who collect furniture. I have had time to pour over them and - gasp - even touch them! While I am not a Redwoods grad, I have met and talked with Krenov and I asked him about the dowel joinery and he stated that it let him build the look that he wanted with a minimum of fuss. I should add that this was before biscuits were about. I can't agree with your extrapolations taken from the photo you attached either. While the photo is pretty fuzzy I'd say that you're looking at 5mm holes, the bottom of their circumference being again about 5mm from the table surface and the top of the guide being about 8mm above the holes for a total height of the guide of 18mm or 5/8". The piece being doweled is harder to make out but I'd say that it's at least 1/2 thick.

Nonetheless it seems that the sum of your experience with this matter comes down to having read the same books that I did. Attitude is no attidote for a lack of facts.

I’m not sure why you didn’t respond with something like “Lamello has had a setting for 8 mm stock since 1957, it’s German precision, must be user error if you can’t get it to work.”

The simple answer is that biscuits are inherently less accurate than dowels, which is one reason not to use them. The primary reason for not using them in solid wood is that they can telegraph through to the surface. This is due to the same humidity cycle that many have mentioned in this thread causing the dowels becoming oval, only it shows on the surface.

[Jon van der Linden]

Something doesn't make sense.

Krenov is a big fan of knife hinges.

He mortises the hinges into the cabinets and doors.

If you look closely at the proportions of the knife hinges to the stock, it is clear that the typical stock thickness has to be 5/8" to 3/4" thick.

If we were to insist that the stock were 3/8", that would be the knife hinges were 3/16" wide.

Furthermore, the drawers and inside doors are approx. three quarters to one-half the thickness of of the carcase. So you'd be talking about 3/16" to 1/4" thick components there. That doesn't make sense.

What do you think?

The commonly available hinges are 3/16" and 1/4". It's common practice for the hinge to be half the thickness of the door and you can see photos of this in Krenov's books, so his work is obviously not an exception. There would be no need to go to 3/16" hinges if the door wasn't well under 1/2".

The idea that the sides have to be thicker than the door is incorrect, although they can be. Some of Krenov's pieces have sides equal to the door, some don't. I think people are reading too much into the 3/8" number I stated. It's a lower limit for the smallest and lightest pieces.

My real point in mentioning that limit is that some of the pieces are very very small and thin. It's important to know and understand that he wasn't using the same thickness as we see in kitchen cabinets. He used both much thicker and much thinner stock. There's also a lot going on with his treatment of edges and transitions that call for different thicknesses in works of the same size and proportion because of their visual weight.

Regarding the thickness of components, some are very thin, possibly as thin as 3/16". He wouldn't be the first to use thin components, I know of many French 18th c. pieces that have extraordinarily thin shelves, drawer sides, etc. I've seen 18th c. drawers so thin the bottoms had to be glued on because there's no room for a dado. Krenov used thin shelves in his tool cabinet and many other places. Sometimes he would chamfer the front edge of a shelf to make it look lighter than it was. There are many factors for these dimensions, but they were all for work that was to be treated with respect, not like a steamer trunk that would be tossed around.

I always get the feeling that people are looking for some kind of recipe book and absolute answers. I'll bet that if everyone from this forum cooked the same recipe they'd all taste different. That's the "art" side of things, which doesn't translate into numbers. While it's possible to go around and measure all of Krenov's work, I don't think it's desirable. If you're creating a work it should reflect your personality and needs, not someone else's.

[Jon van der Linden]

To me, "aesthetics" is a lot more complex than the look, surface ornamentation (or the lack thereof) and proportions of a piece. How it's constructed has at least as much contribution.

This looks like a fun topic, but somehow I suspect it's like debating different religions!:D

[Tony Shea]

The cabinet in mind is going to be of some slightly curly pear wood I've accumilated at Gilmer. It is steamed but still incredibly nice stuff. A couple peices are very curly but most are slight. I've done some planing just to see what the surface is like and all I can say is wow. Of all the woods I've worked with I've never produced such a silky glossy finish with a plane. Just great stuff to work with.

Read a little of your blog and really liked the mockup idea with cardboard. Is a very showing method of life size proportions. I will def be using this method in my future projects. So simple yet so telling. Thanks again.

[Gary Curtis]

When I was growing up in the '40s, like all kids my memories are of crawling around on floors and looking up at the underside of furniture. And I remember so many exposed dowels, which were a curiosity to my eyes. Why would they be exposed?

In the FWW book on Joinery there is a damning, scientific commentary on Dowels.. The essay promotes loose Loose Tenons. The Tennons are great because they form a laminate with the carcass or frame member. Several drawings compare them to Dowels. And that is where the damnation comes in. A cross-section reveals that virtually all the grain on a dowel — save for a narrow band on each side — is end grain. And through seasonal changes, the dowel is anything but round after a short while. Though the bore hole it's glued into is definitely round. So cavities form, fracturing the glue.

A California architect named Singer, wrote extensively here on this forum on the advantages of joints which weld the elements together in a laminate. I use a lot of Festool Dominoes.

[Trevor Walsh]

Tony, I'd love to see a picture of that pear, I've been hunting for some, but my usual haunts are all domestic woods, I'll have to find a special place I guess.

I'm glad you liked some of the blog, good to know some people actually read it. I may or may not have stated elsewhere, I went to school for product design. Everything we did involved mock ups, for function or to do human factors testing, different graphic styles, drawing after drawing. Many dozens of drawings became several cardboard models, became a few better models became the thing that got tweaked in a few more models or revisions. Iteration, the first 10 of something were warmup then you really started looking at what you were drawing.

Wood is (the good stuff) expensive compared to cardboard or pine, and it builds fast. To me there is no reason to break into the good stuff until you've seen a bit about how it fits together first.

[Phil Thien]

A cross-section reveals that virtually all the grain on a dowel — save for a narrow band on each side — is end grain.


It would seem that dowels are all long grain to my way of thinking. But, the hole they go into may have considerable end grain.

Again, though, I'm just not certain of some of the assertions by the authors of these articles.

First, when examining doweled joints that are 100+ years old, we have no way of knowing whether the dowels were truly round to begin with.

Second, shrinkage calculators seem to indicate we'd only be looking at (typically) .003 to (max) .006 (depending on species and size of dowel) of shrinkage.

Third, the wood forming the project we're holding together with dowels is also shrinking.

Fourth, end-grain glue joints are stronger than most people give them credit.

Fifth, modern glues can accommodate seasonal movement.

Sixth, we have newer compressed dowels which expand when they come into contact with the water in PVA glues.

Someone mentioned that Krenov's work isn't old enough for us to know whether his doweled joints will stand up to time. I would submit no work glued with modern adhesives is old enough for us to know whether ANY of it will last.

Finally, I agree with the poster that said they'd use biscuits on that joint. I would, too.

Hey, at least I wouldn't use pocket screws!

[Gary Curtis]

Wish I had access to that illustration from the Taunton Press book on joints showing grain patterns. My books are all in storage because of a move. But I trained in architecture and worked in aviation for most of my career, so I understand strength of materials issues bit.

I believe it said that the way a dowel is milled from a blank leaves only a small fraction of the circumference that is not end grain. And this I'm sure of. End grain glued to anything is relatively weak.

Those studies done in Finland after WW-II did destructive testing on various wood bonds. But tradesmen from olden times didn't need to be told that.

[bill tindall]

I haven't ever posted here before so expect some errors. Before criticizing Krenov's means of holding his cabinets together it would be well to understand the details of the complete process, which I have learned from his students. I am no hero worshiper of Krenov. I'm too practical and impatient to subscribe to a lot of his philosophy. But, as I studied his techniques for making stuff I have come to appreciate he was an excellent engineer. You will come to know that if you build in his style and begin to face the numerous engineer challenges for making these pieces.

I have built Krenov style cabinets and indeed the first was made with "approved" mortise and tenon joints. That was the only one I built that way. What isn't obvious till you get into building one of his cabinets with a curved or faceted door is that you will be dry assembling and taking it back apart a multitude of times before finally gluing it. In these cabinets the door is made first and the cabinet fit to it. What is done is that all the dowels holes are drilled with a clever "Krenov doweling jib" but only 3 on each side are used for dry fitting. When it comes time to do glue up, the by now exhausted sacrificial dowels are tossed and new dowels installed all around. The miltitude of dry fits destroys M&T joints.

A Krenov cabinet will never fail because the dowels are only part of what is holding it together. Krenov backs are homemade, or otherwise, plywood that is tightly fit and glued in place. It is the back of the cabinet that provides the strength to his otherwise frail looking wall cabinets.

I have begged David Finch, or another of his students, to write a book on Krenov construction techniques. You will not find enough detail in Krenov's books to successfully emulate his cabinetry. maybe it will happen one day.

[Tom McMahon]

It would seem that dowels are all long grain to my way of thinking. But, the hole they go into may have considerable end grain.

Again, though, I'm just not certain of some of the assertions by the authors of these articles.

First, when examining doweled joints that are 100+ years old, we have no way of knowing whether the dowels were truly round to begin with.

This is true, how ever when you examine large numbers of them and they all seem to be taking on the same shape there has to be an explanation .

Second, shrinkage calculators seem to indicate we'd only be looking at (typically) .003 to (max) .006 (depending on species and size of dowel) of shrinkage.

That may be enough when repeated twice a year many years years.

Third, the wood forming the project we're holding together with dowels is also shrinking.

Again this is true but it just makes the problem worse because the dowel and the project expand at different rates due to grain orientation and this causes the cell structure of the wood to begin to be crushed loosening the joint.

Fourth, end-grain glue joints are stronger than most people give them credit.

I will give you this one.

Fifth, modern glues can accommodate seasonal movement.

PVA was invented in 1912, epoxy in 1936, cyanoacrylate 1942 these are only modern when compared to casin or hide glue all of which are stronger than the wood.

Sixth, we have newer compressed dowels which expand when they come into contact with the water in PVA glues.

Again this only makes the problem worse.

Someone mentioned that Krenov's work isn't old enough for us to know whether his doweled joints will stand up to time. I would submit no work glued with modern adhesives is old enough for us to know whether ANY of it will last.

Finally, I agree with the poster that said they'd use biscuits on that joint. I would, too.

Biscuits are just flat dowels.

Hey, at least I wouldn't use pocket screws!

Same problems as dowels.

I am not against using dowels as long as we understand what we are doing and that we actually don't know the long term reliability of this type of construction. As I said before it will probably be fine in our lifetimes.

[Phil Thien]

I am not against using dowels as long as we understand what we are doing and that we actually don't know the long term reliability of this type of construction. As I said before it will probably be fine in our lifetimes.

I just can't agree with statements like that. Mills used dowels for everything from entry doors to window components to whatever. Tens of millions of doors and windows have been (and continue to be) subjected to the most brutal conditions, and hold up very nicely, thank you.

These types of arguments kind of remind me of audiophiles arguing about interconnects and speaker wire. A sort of purist mentality that doesn't result in much enjoyment of music.

[Tom McMahon]

Phil, I agree with what you said about doors and windows. I was referring to using dowels to connect the bottom of a cabinet to the bottom of the sides of the cabinet, where gravity is working against the joint.

[Frank Drew]

[QUOTE=Phil, I agree with what you said about doors and windows. I was referring to using dowels to connect the bottom of a cabinet to the bottom of the sides of the cabinet, where gravity is working against the joint.[/QUOTE]

Unless the cabinet is as small as Krenov's seem to be, that's my objection, too. Gravity, plus any live load inside the cabinet.

I'm not sure when dowels became the joinery method of choice for doors and windows, but most millworks at one time, and not all that long ago, had single and double end tenoners and mortise machines (hollow chisel, chain, whatever). My first job in woodworking, in the early Seventies, was at an architectural millwork and virtually all the joinery we did was mortise and tenon.

[Jon van der Linden]

Unless the cabinet is as small as Krenov's seem to be, that's my objection, too. Gravity, plus any live load inside the cabinet.

I'm not sure when dowels became the joinery method of choice for doors and windows, but most millworks at one time, and not all that long ago, had single and double end tenoners and mortise machines (hollow chisel, chain, whatever). My first job in woodworking, in the early Seventies, was at an architectural millwork and virtually all the joinery we did was mortise and tenon.

This idea that stuff won't hold together is getting a little silly. First of all, we're not talking about large loads or roughly treated items exposed to weather. Second, we're not talking about poorly drilled tapered holes with out of round dowels held in with hide glue.

If we look at modern applications under stress, I can give a personal example where I glued a 16' 2x6 to two 4x4 posts without any mechanical connections as a test using a common adhesive. It was outdoors (a trellis with plants on it) and exposed to all weather conditions. After 12 years it was still holding just fine. In boats people glue winches in - epoxy in a round hole - and it holds just fine. The loads in these situations are much higher than we're talking about, and it's only the glue, no mechanical connection... and now there are even cars that are glued together, no welding or mechanical connections.

I'd bet that with a moderately good adhesive I could glue end grain to face grain and have it hold under a moderate load. The concern about dowels is way overblown.

[george wilson]

Airliners are also glued together,though that makes me a little nervous!! At least 1 roof section did pop off!! They must be doing incredible things with adhesives,gluing planes that will see exceeding low temperatures at 30,000 feet,and scorching heat on the ground. And,they glue on space shuttle tiles,too! In space,even!!

[Tom McMahon]

George that's the point. If any one has great glue it would be NASA and they have to re glue tiles in space that means that the glue failed. I can't even believe that this has turned into a controversy. Part of Krenov's genious is that he did thing that should not work and made them work, it does not make it good common practice and to pretend it is does a disservice to beginners. .

[Jon van der Linden]

George that's the point. If any one has great glue it would be NASA and they have to re glue tiles in space that means that the glue failed. I can't even believe that this has turned into a controversy. Part of Krenov's genious is that he did thing that should not work and made them work, it does not make it good common practice and to pretend it is does a disservice to beginners. .

You are kidding right??? Because adhesives fail over time with a particular material under particular circumstances i.e. 17,000+ miles per hour on reentry, and parts over 2,300 degrees fahrenheit. Comparing that to a the stresses on a cabinet is exactly the kind of silliness I was talking about... for one thing, I'd be more worried about the wood surviving the 2,300+ degree temperature.

Krenov did not do things that "should not work." His methods are very sound. Compared to 18th c. French furniture, or methods employed for centuries in instrument making, his techniques are simple and rock solid.

[Frank Drew]

This idea that stuff won't hold together is getting a little silly. First of all, we're not talking about large loads ...

Krenov-style slightly scaled-up doll house furniture (essentially, your description) isn't the only furniture being built; we are, in fact, talking about large loads some of the time.

Jon, you're free to build furniture or cabinetry any way you want; I like to have mechanical strength in addition to any glue strength in the joints I use.

[Tom McMahon]

Jon, I must apologize again for not expressing myself clearly. The point I was trying to makes is that I don't believe there is a magic permanent glue that doesn't fail over time, due to the seasonal moisture cycle or stress. I base this belief on 40+ years of repairing furniture, it's what I do for a living, I have worked on 300 year old pieces and pieces still in the furniture store that were not sold yet. All glue has the ability to fail, it may not fail but it may, for this reason I believe one should use joinery that will hold a piece together. If the glue was as reliable as you imply you would not need the dowels.

[Jon van der Linden]

Jon, I must apologize again for not expressing myself clearly. The point I was trying to makes is that I don't believe there is a magic permanent glue that doesn't fail over time, due to the seasonal moisture cycle or stress. I base this belief on 40+ years of repairing furniture, it's what I do for a living, I have worked on 300 year old pieces and pieces still in the furniture store that were not sold yet. All glue has the ability to fail, it may not fail but it may, for this reason I believe one should use joinery that will hold a piece together. If the glue was as reliable as you imply you would not need the dowels.

Agreed. All glues fail, and the wood does too at some point, even with the best of care under near ideal conditions. Sounds like a guarantee for future employment for you! With modern adhesives the problem seems to be a lack of reliable information on what will happen with a specific product over time. Obviously the failure of a winch being ripped out of the deck by a 70' tall sail is going to be a little more problematic than a shelf coming loose, but even there there seems to be little data - maybe the boats get junked too fast for the glue to fail!

Since we know that everything fails the real question seems to be how long is long enough? I may have mentioned earlier in this thread how I saw the front of a Riesener desk being destroyed. It seems that at a certain point fate has a lot more to do with what happens to something rather than how well it's made.

The moisture cycle is an interesting problem. Some glues are rigid and some remain flexible (at least in the near term), I think this would have some effect on the success of a joint that has movement in it. The long term is still the big question.

I might do a test just for laughs with some scrap wood to test a glue only butt joint. Probably a few days away before I can get to it, but it will be fun to see what happens.

[Phil Thien]

I'm not sure when dowels became the joinery method of choice for doors and windows, but most millworks at one time, and not all that long ago, had single and double end tenoners and mortise machines (hollow chisel, chain, whatever). My first job in woodworking, in the early Seventies, was at an architectural millwork and virtually all the joinery we did was mortise and tenon.

If you look at this site:
http://www.emeralddoors.com/Front%20Entry%20Door%20Construction%20Methods/custom_exterior_doors_questions.htm

You'll see them comparing their door construction to what is pretty typical in the mass-production entry door model. The mass-produced doors do typically use a simple cope and stick joint that is reinforced with dowels.

Here is a site where they show the method for making windows:

http://www.dynamicwindows.com/?action=d7_article_viewer_view_article&Join_ID=132870

I haven't demolished a ton of doors, but the few I have ripped apart have been held together with dowels. And they (dowels) haven't given up easily, even after 50+ years of service in a hostile environment.

I think dowels in doors and sashes became pretty popular with the post WWII construction boom. There are many benefits to this type of construction.

[Dave Anderson NH]

Great discussion folks. On the subject of adhesives there are many applications in the transportation industry including trucks, buses, and aircraft where structural and trim components are both held by adhesives without any mechanical fastners involved. Examples of this incllude the body panels of Greyhound buses, Navistar stucks, and the stainless steel rub strips on the flaps of the Airbus 320. All use a special acrylic pressure sensetive adhesive for speed of assembly and cost reduction. A further and more important reason for the use on large parts like these is the ability of the adhesive modulus to absorb the size changes of the parts caused by wide variations of temperature. Other than the space shuttle, I can't think of a more unforgiving environment than an airliner sitting on the tarmac in Phoenix on a hot summer day when the surface temperaturee of the wing reaches 130+ degrees and then having the same plane 5- 10 minutes later at 30K feet where the air temp is -40F. Properly engineered, specificed , and applied adhesive have amazing properties. To absorb temperature and humidity variation the adhesive chosen must have a high flexural modulus.

[Kent A Bathurst]

........an airliner sitting on the tarmac in Phoenix on a hot summer day when the surface temperaturee of the wing reaches 130+ degrees and then having the same plane 5- 10 minutes later at 30K feet where the air temp is -40F. Properly engineered, specificed , and applied adhesive have amazing properties..........

I recall seiing - this goes back a ways - a video clip if the SR71 Blackbird. Still has the record as world's fastest aircraft > mach 3. If they sensed a missile fired at them, they just stepped on the gas and outran it. Ceiling at 80,000 feet.

Anyway - the point - the thermal expansion of the components was so great that they had to be very loose-fitting on the ground - like inches [?], and the plane just sat there and fuel ran out of it. Not a drip, but not a gusher either - somewhere like a moderate, steady leak. They'd launch, climb fast to heat up the skin, and then refuel mid-air. I thought those guys had to be over-the-top nuts to fly something like that. But, what do I know?

[Jon van der Linden]

Of course there's the argument that the adhesives used for airplanes and space vehicles are both unavailable and prohibitively expensive. Bringing this back down to earth... I glued up a test block with a common epoxy to simulate the load on the edge of a cabinet with 3/8" sides.

Test piece: red oak scraps, surfaces planed and cut over two years ago, surface prep was two passes with 60 grit sandpaper to clean off any dirt. Dimensions: 3/4" x 1" end grain. The gluing is a little sloppy and I didn't use clamps (hey, I was busy doing other things). The adhesive cures overnight, but I tested it after two hours (full cure in a week).

198128198129

You can see a volunteer hanging from the 3/4" x 1" block glued end grain. I was able to get both of us on there, but no one to photograph it. The point here is that with no connection other than the adhesive an end grain connection (with 3/8" sides) can easily hold over 150 lbs. I think that's more than enough for fine furniture. The adhesive forms a rigid glue line (no creep) but it is also flexible, so it's doubtful that any amount of wood movement will cause failure. In essence it proves that a dowel connection should not be a major concern, and if it is, there are ways to make up for it.

Regarding failure from aging, I have contacted several major manufacturers but none have any aging concern except when the product is exposed to UV. The product that I use happens to have a UV inhibitor in it. Anecdotally, I was informed that there were boats over 40 years old constructed with a glued lamination technique without problems.

[Phil Thien]

The commonly available hinges are 3/16" and 1/4". It's common practice for the hinge to be half the thickness of the door and you can see photos of this in Krenov's books, so his work is obviously not an exception. There would be no need to go to 3/16" hinges if the door wasn't well under 1/2".

Where are you seeing 3/16" and 1/4" wide knife hinges? The smallest I can find are 5/16":

https://www.brusso.com/Product/instructions/install_st_pivots.pdf

I don't believe you could use any knife hinge in 3/8" thick material. By the time you account for the depth of the hinge mortise (1/8"), and screw bite (1/4"), I think the minimum side top/bottom would be about 1/2".

What do you think?

Not trying to belabor the point, I'm actually curious how thin those cabinet components are.

Have you personally seen any Krenov (made by Krenov, not in the style) cabinets with knife hinges where the top/bottom/sides are 3/8?

I seem to remember seeing some units (pictures, not in person) where the top/bottom were thicker than the sides. Perhaps that is how he handled it.

[jamie shard]

Great real world testing Jon! :)

[george wilson]

In the late 50's,when I was young and relatively inexperienced,I glued a classical guitar bridge on one of my guitars with epoxy. Big mistake!! Months later,I was on the phone with the guitar behind me. There was a loud pop,and the classical bridge(about 1 1/8" X 7" ,holding nylon strings,popped off and slapped me in the back! It was a perfectly clean break. I re glued it with epoxy,and months later,same thing happened. I think the epoxy (back then) had continued to get harder and harder until it got too brittle,and lost its mechanical strength. Last time I ever used epoxy on a bridge!!! I hope epoxy has gotten better by now. Even that long ago,I had been building for several years,and I'm sure I had mixed it properly.

[Jon van der Linden]

Where are you seeing 3/16" and 1/4" wide knife hinges? The smallest I can find are 5/16":

https://www.brusso.com/Product/instructions/install_st_pivots.pdf

I don't believe you could use any knife hinge in 3/8" thick material. By the time you account for the depth of the hinge mortise (1/8"), and screw bite (1/4"), I think the minimum side top/bottom would be about 1/2".

What do you think?

Not trying to belabor the point, I'm actually curious how thin those cabinet components are.

Have you personally seen any Krenov (made by Krenov, not in the style) cabinets with knife hinges where the top/bottom/sides are 3/8?

I seem to remember seeing some units (pictures, not in person) where the top/bottom were thicker than the sides. Perhaps that is how he handled it.

A lot of people make their own hardware, it's pretty easy. You can also buy hardware including 3/16" and 1/4" knife hinges made by graduates of the College of the Redwoods (Sanderson Hardware):

http://www.sandersonhardware.com/blog/store#ecwid:category=132832&mode=product&product=261164

You're right about the top and bottom often being thicker than the sides. I don't think I've seen a cabinet where the top and bottom were 3/8". But I don't make copies, so measurements in that way aren't something that I'm usually looking at.

[Jon van der Linden]

The range of epoxy resin formulations is pretty big. I know that the kind of stuff I use wasn't available then (water clear, no yellowing, with UV inhibitors, etc.). It would be interesting to know the actual cause of the failure in your guitar, my first suspicion would be that your surfaces were too smooth... epoxy needs some tooth to hold on to, much rougher than normal surface preparation in lutherie.

Most of the epoxy use I see is in cold molded lamination (boats) or sporting goods (tennis rackets, snow boards, etc.). Lots of stress in those applications, and failure is usually from a fairly obvious source like an extreme impact. Asked a friend of my brother in law about how his kitesurf boards held up and he pointed to a scratch on the bottom of his board - apparently it was what happened when he tried to jump the breakwater at Waimea Bay and skipped off the rocks instead. It's amazing how strong it can be, which makes it even more interesting when it fails.

[Gordon Eyre]

A couple of weeks ago my neighbor brought over a silverware drawer from his cabinets and the joint on the back of the drawer had come loose. It was a dowel joint and the glue had failed. I fixed it for him but thought at the time that dowel's are not meant for areas where there is a lot of stress. Personally a dowel joint would be down the list of proper joints IMHO.