Why dowels, not M&T for a Krenov cabinet

Re: Quite an assumption

MikeW

>Now Pam, I didn't say that. I do believe the topic is full of misinformation, though.

"I do know that having used dowels for casework such as Krenov (and some his other methods for casework) and having had some of these pieces hold together for over a decade or so (nothing I have built has a joint that has failed, regardless of the joint), someone will have to have something other than "Gee, I don't think it's a strong method" "short-grain vs. long-grain" theories that they themselves have not tested in order to consider the information."

The important part of the quote you edited out is until someone has something other than a mere opinion, *and I might add* would care to share it with us, it is just a theory--and that's putting it as politly as I can.

Without something more substantial than to say "However, the dowel-to-carcase-top (or dowel-to-carcase-bottom) is significantly weaker than the equivalent, tightly fitted M&T, because there is very little face-grain-to-face-grain surface in the hole in the top." and then go on to say "Note that I have built these carcase joints with biscuits, and I have even made mitered frames with biscuits. I know now that this sort of joinery has its problems, but so far, even these abysmal joints have held in these old projects. So maybe I'm just talking about archival processing, or maybe I'm just making a mountain out of a molehill."

Just what does that mean? At least someone is willing to admit maybe there isn't a problem?

But please, Pam, do you have something that would convince me it is a weak joint if done properly? I'm more than willing to listen and consider anything you, or anyone here, as to say. I certainly do not want to continue building things that may fall apart in the years to come.

Thank you, Mike

Re: Actually ..... more data

MikeW

>Hi Bill,

Thank you for taking the time to post all this. I for one am glad you are taking the time--and I would love to see the data--but won't bet. I've been making cabinets like this for too many years.

I had posted once before in an earlier thread or two. That case has 21 dowels per top/bottom joint. 42 dowels (I also use letter bits). I sold a display cabinet of similar dimensions as the one in the link way too many years ago. In that post I relate destroying one built that way too. Plenty strong. (Aside from the cross-grain joint. Bet someone is gonna like that one.)

Your test subject is gonna take a lot. If you use yellow glue, make sure to let it sit at least a week before stressing it, though.

Also, perhaps a joint made of say 1/2" box joint as a control (easier than the stopped m&t and comparable glue area).

MikeW

Re: Krenov Explained "Why" in His Book!

Norman (Ottawa, Ont.)

>True.

If there were any doubt about the inherent weakness of dowel joints vs. mortise and tenons. the multitude of dowels per carcase side

could easily dispel this. In my cabinet, if I recall correctly, and I should have made notes... I used 7 dowels per carcase corner.

Once the alignment and dowel orientation issues are overcome, doweled panels assemble quite easily actually.

I have another smaller cabinet built with the same Krenov-type doweled construction.. I built this 7 years ago. I carries much weight, namely a full load of music cd's.. No problems whatsoever.

I might have used 4-5 dowels per side corner on that one..

Norman

Re: Quite an assumption

William Duffield, on the Cohansey

>What I mean is that a decade is hardly "the test of time". And, BTW, I don't break them (at least not intentionally). It's the clients who break them, usually after many decades, and then I get to fix them, or to try to explain that they are not worth fixing. Much younger than that, and they're not worth fixing, because they're mostly mass produced factory furniture. Doweled joints break much more often, do a lot more damage to the piece, and are often much more difficult to repair than M&T'd joints. What is it that you don't understand about the lack of strength of glued end grain? That issue is only geometrical.

Where does the M & T come in?

Stephen W. Schoene

>I'm confused. Aren't the Krenov joints being discussed in carcasses? The traditional and most robust alternative to the dowel joint is a dovetailed joint not the M&T. And to avoid having the joint become part of the design, it would have to be a blind dovetail, or--as in much traditional furniture covered by a molding.

M&T joints are more generally the choice for face frames, not for the carcass bodies.

Re: Actually ..... more data

William Duffield, on the Cohansey

>I hae a couple of serious questions for you, Bill. I ask because I don't believe dovetailed carcases are appropriate for all pieces, and the other methods I have used leave me with a very uneasy feeling. So, if I could make a better doweled joint, I could very well be tempted to try building something using it.

My attempts at hand aligned dowel holes have been less than successful. Can you share some of the secrets you learned in the class, as illustrated in the picture you posted?

You note that M&T joints are well worn, and loose after several trial assemblies. I agree. [As an aside, I think most of the looseness is compressional, and just like a compressed biscuit, or an ironed-out dent in a surface, the application of wet yellow glue should cause the mating members to swell to their original size as the joint is assembled and the glue is absorbed into the fibers and sets up. But, that's not the question.] What I'm interested in is the sacrificial dowels. In order that they do not significantly compress and deform the sides of the holes, are they reduced a bit from spec'd diameter before you start the trial fittings?

Of course, the question most of us are probably asking is where do you get (or where do you recommend) your dowels and your drill bits, and is there a table of recommended size correspondences?

OK, that's three questions, not two.

Re: Quite an assumption, revisted

MikeW

>What I mean is that a decade is hardly "the test of time". And, BTW, I don't break them (at least not intentionally). It's the clients who break them, usually after many decades, and then I get to fix them, or to try to explain that they are not worth fixing. Much younger than that, and they're not worth fixing, because they're mostly mass produced factory furniture. Doweled joints break much more often, do a lot more damage to the piece, and are often much more difficult to repair than M&T'd joints. What is it that you don't understand about the lack of strength of glued end grain? That issue is only geometrical.

Mr. Duffield, I meant you no disrespect and I'll ignore the condescension.

I wasn't speaking to the issue of repair, to that, well, I don't care. If and when it occurs on a piece I am responsible for, I'll worry about it then. A decade truly is not a test of time. I don't believe that is an issue here, at least to me.

As for breaking them, I'm not certain I was speaking of that in this portion of the thread. I did comment on Bill's suggestion he would be willing to test the strength to a failure. I then commented that I did indeed do that, though not for any quasi-scientific purpose.

We also are likely talking about different types of furniture/cabinetry. A person who commissions a display cabinet may well have a catastrophic event happen to the piece, but I'm willing to bet that not only does this type of piece not see abuse, it is also not a piece that sees daily use such as the credenza I am building (with good old pinned M&T joints).

For the display pieces I finished a month or so ago, which will see much use and is of a rather large scale, it too doesn't use dowels. But it is a viable method for certain pieces.

Anecdotal "evidence" from either side of this issue is worthless. The figures Bill quotes are quite impressive as applied to a little wall cabinet I built two months ago that hangs via offset keyhole hangers from the sides and carries a whopping 10 lbs of weight, or the 11" x 20" cabinet on stand referenced by myself that uses 42 dowels precision drilled and glued with epoxy.

MikeW

Re: Actually ..... more data

MikeW

>I would imagine right about now you could care less to hear from me as to these issues.

But hey, if you change your mind I would be very happy to share this information as well.

I have no idea how Krenov nor a former student goes about it other than a few vague statements in Krenov's books. So thank you for asking Bill to share this. I too would like to know how they go about it.

MikeW

Re: Where does the M & T come in?

Todd Stock

>Krenov's stuff is designed to show in 3-D, so sliding dovetails for side-to-base or side-to-top joint would require the sort of slap on molding that he would tend to avoid at all costs. Further, Krenov's method for doweling works pretty well (brads, snipped off, with marks from dry fit defining center of dowel).

These days, I'd probably use those little football-shaped things rather than fuss with it.

Re: 4 more debatable reasons

Jeff Schmidt

>We'll see how Krenov furniture fares

It is likely to fare well since it is little used and mostly gazed upon. :-)

Glued Surface Areas- long

Patrick Gibbons, Houston

>Using dowels, one has the long grain of the dowel glued to the cross grain of the member, just as a tenon is long grain glued to the cross grain of the mortise. Imagine one wishes to glue a 2" x 2" leg to an apron of 3/4" stock by 3" wide (length is irrelevent in this case). Using three dowels of 5/8" diameter and two inch length where one inch is recessed into the leg and one inch into the rail, the surface area glued into the leg would be pi x diameter (circumference) x length would be

3.14 x .625 x 1 = 1.9625 times the quantity of three dowels would be 5.8875.

Assuming a M&T of 2" x 1/4" x1" with no haunch to make this easier to calculate the gluing surface area would be

1/4" x 1" x 2 + 2" x 1" x 2 = 1/2" + 4" = 4 1/2". The dowel joint has 31% more glue area than the M&T joint. One can drawbore the M&T providing mechanical strength that will hold without glue. I also wonder if the M&T joint has an added benefit of being together rather than in three different places. At first glance it appears that a well-made dowel and modern glue may in fact provide a stronger joint that a mortise and tenon.

Re: Glued Surface Areas- long

Patrick Gibbons, Houston

>It can be rightly argued that normally 5/8" dowels would not be used in 3/4" stock. The glued surface areas for alternate dowels in the same example above would be;

1/2" dowel- 4.71" surface

3/8" dowel- 3.53" surface

1/4" dowel- 2.36" surface

I would also like to state that I do not necessarily believe a doweled joint is preferable to a mortise and tenon. The tenon is a part of the piece while a dowel must rely on glue in both members being attached. As I stated before, the tenon can be mechanically attached to the mortise by drawboring without benefit of glue. A dowel joint is faster to make and therefor more economical. It is really the case that one should use the joint which is appropriate for the piece. Sir William has talked about repairing older furniture where the doweled joint has failed and that because of this joint being used, repair is much harder to do. Perhaps the better question to ask is why those joints failed. Bad glue? Bad dowel? The fact that both ends must rely on glue?

One More Correction

Patrick Gibbons, Houston

>In my example above, the M&T surface area should only be 4". The 1/4" x 1" faces would be attached to the end grain of the mortise and therefor not counted as they add no strength to the joint.

Yet Another Factor

Patrick Gibbons, Houston

>This is what happens when I don't ruminate before committing to paper. The dowels would also have a portion of their area mated to end grain. I'm not sure how much but I can intuit it being quite a bit as there would be six dowel areas as opposed to two tenon areas aligned with end grain. Further, the dowel areas exposed to end grain would be similar in area to that exposed to cross grain while the tenon's area exposed to end grain is substantially less. It is also easy to intuit that to make a dowel joint as strong as a M&T requires many dowels to be used. I still say the purpose should not be to make a dowel joint stronger than a M&T but to make a dowel joint that is strong enough for the purpose at hand.

Re: Yet Another Factor

Joel

>Awhile ago FWW ran what I thought at the time was a good article looking at joint strength. classic M+T was at the top, dowels IIRC were at the bottom. of course fit and sizing play an important point.

I do remember that a few years ago I saw some Frank Llyod wright chairs in a museum where the back slats had all been put together with dowels. The joints were all failing and the back was falling apart.

I don't think any joint is intrisically superior to another it depends on the application and a lot more factors then absolute strength. but unless you have really on the money dowelling equipment I'm just not a fan.

dowels had their hayday from the 1920-1980's now they are out of favor except for multispindle doweling machines and most people use biscuts. I would bet that if Krenov were starting again he would probably use biscuits.

had to look it up

Bob Lang

>Fine Woodworking, in issue #21, April 1980 ran an article by R. Bruce Hoadley titled "The Dowel Joint; Why Round Tenons Fall Out of Round Holes". I don't have it in front of me, but the gist of Hoadley's argument is that over time, the seasonal expansion and contraction of the surrounding wood and the dowel changes from perfect circles to ovals. Once this happens there is very little wood to wood contact, and the joint is subject to failure.

This doesn't mean that all dowel joints are doomed to failure, but given enough time the chances are pretty good that something will let go. My anecdotal experience is seeing lots of pieces around 90-100 years old with failed dowel joints. The number of other types of joints I have seen fail is far fewer.

I can see getting away with it under the right circumstances, particularly a lot of dowels, carefully placed, in a light-weight piece that won't likely be subject to extreme swings of temperature and humidity. Hoadley's article made enough of an impression on me that I still remember it 25 years later (probably one of the few things I remember from that long ago) and I haven't used a dowel by choice since.

While looking up the Hoadley article on the FWW archives, I also came across one by James Krenov from the Summer 1977 issue. I believe, but I'm not certain that this is an excerpt from his first book. For what it's worth, that article I didn't remember.

Bob Lang

maybe, but the students don't

bill tindall

>Of course I don't know what Krenov would use today but if you examine the reasons I stated for prefering dowels over M&T for the specific application under discussion (and that would be italics if I knew how to do Italics) one can see clear reasons for dowels over biscuits. Not surprisingly, it follows that the students of this school continue to use dowels for this specific application. Note that Krenov used floating tennons in many pieces, where necessary, for example leg atachment where there is stress.

Dowels (7! of them in the example I saw), aided by gravity, hold the top of the cabinet on and keep the sides aligned with the top and bottom. Will someone point out the load that will fatally stress this joint to where the admitted-by-all superior strength of M&T over-weighs the obvious disadvantages of the M&T in this application?

??? 

Luke Herzberg

>

Touche! I yield.

bill tindall

>

Re: Where does the M & T come in?

Stephen W. Schoene

>Wouldn't need a sliding dovetail. Could be made with a blind miter dovetail and that could even be modified to look like it was a butt joint. No dovetail would show on any side. But it is a PITA to make, and you don't get "credit" for with the customers it since it is invisible. Thus, since dowels, especially when so many are used, are fully capable of handling the modest expected loads, sheer economics suggests the dowels be used instead of the dovetails. My only point is that the conventional joint for joining carcass sides to top and bottom is the dovetail not the M&T, but that for Krenov to achieve the esthetics he targeted would have required the very, very challenging blind dovetail, not the almost as easy as dowels M&T joint.

Re: Anecdotal Evidence

William Duffield, on the Cohansey

>Sorry, Mike. I didn't really mean to be condescending. I was just trying to make it clear that square pegs in square holes make a stronger joint than round pegs in round holes, everything else being equal. Of course, everything else is never equal, so it comes down to how good is good enough for the application. I think we have reached a consensus that Krenov's techniques are more than good enough for the style of cabinets he and others with CoR training build using those techniques.

As far as the uselessness of anecdotal evidence, I must strongly disagree. Every data point taken in every scientifically designed experiment is an anecdote. It is only a whole lot of experimentation, more anecdotes, measurements, perspiration, statistics, good luck, education, research, inspiration, tweaking unreliable machines, work, work, work, free grad student labor, and a stable of lawyers making sure that the description of the results cannot be broadly interpreted beyond the scope of the experiment, that results in a defensible thesis.

Many times, it is a single anecdotal anomaly that leads a researcher to a new hypothesis, and to the testing of a new theory.

Demanding only exhaustive, scientific studies reminds me strongly of the parable of The Goose That Laid the Golden Egg. Where are you going to find someone to pay for the laboratory, the equipment, the salary of the PhD, the peer review, and the publication? Surely not on a free public forum? Surely, not even in a woodworking magazine that depends on advertizing revenue. Maybe the USDA FPL will pay for one? I think not. Even the USDA's studies are funded by private interests, and those interests are very careful with the allocation of their funds, and even more careful with the allocation of the advertizing dollars they use to publicise the unbiased Government studies. Needless to say, the studies that don't give the sponsors the results they want just kind of disappear. Be careful what you ask for. You might just end up with an empty goose.

Even the careful studies and analyses that are reported here are often attacked out of hand, but seldom, if ever, repeated or otherwise rigorously peer reviewed. This is not surprising, especially considering the effort and expense that would be involved in repeating an experiment in something as variable as wood woked by hand. Still, I encourage more data and more experiments. The more data the better. I'll take it one bit at a time, if necessary, as long as no one is intentionally fudging their results.

Re: Actually ..... more data

William Duffield, on the Cohansey

>Please do share your joinery experiences with us. I really would be most interested. I don't think it would be appropriate to ask why you were down, or why you decided to destroy the display case. OTOH, photographs of failure modes related to the force and direction of the impacts would be very useful. These would be even more interesting if they could be compared to similar failures using different joinery techniques, different woods, different dowel (or tenon, or biscuit) sizes, different glues, even different hammers, etc., but building and destroying that many display cases would be way beyond the call of duty.

I've read articles about testing different joinery techniques to failure, but the one I remember most clearly (in FWW, IIRC) was for typical door frame rail and style joints, and the destructive forces were only applied in one direction. While those tests appeared to demonstrate the superiority of M&T joinery, they were far from conclusive or exhaustive. They did nothing to "shed light on" the kind of joint we are discussing in these recent threads. Some expected results can be inferred by extrapolation, but nothing about what might happen under extreme twisting loads.

Re: Anecdotal Evidence

MikeW

>I defer concerning square is stronger. But there are also joints stronger than blind tenons.

I wasn't demanding--maybe being a bit pushy for something other than what was offered. I don't know.

I do feel that my anecdotal experience was being dismissed whereas some were operating from the point of view of being taught what is "right," and responding from there. For others, perhaps a different experience under different circumstances was being applied across the board.

If nothing else, that some do and some don't have problems with dowel construction at least as applied in certain situations should suggest a difference of some sort.

It is this difference--whatever that is--that takes two open minds and the willingness of both to communicate to each other in order to understand, analyze, and assess the benefits (if any) in order to assign value and application.

If you followed that you are a better person than I--I'm too tired to care at the moment.

MikeW

Re: Krenov is Krenov

HC Sakman

>Hi Ron,

You said; "I have not had the opportunity to talk with him about it to understand why he prefers dowels..."

My humble guess is... That's what he was thought (probably) and/or he just feels comfortable with that method of joinery.

Chico...