I’ve recently started using Miller Dowels and so far love the system. But while raving about them to a fellow woodworker, he asked me why bother using them, when a plugged screw yeilds a stronger joint and looks the same once complete. I was stumped. Anyone have a reason why Miller Dowels are better than plugged screws?
Chris @ www.flairwoodwork.spaces.live.com
– Success is not the key to happines. Happiness is the key to success. If you love what you are doing, you will be successful. – Albert Schweitzer
Replies
You seem to take for granted that a screw is stronger. How do you figure? Also, in some situations, the dowel is bound to be stronger than a plain screw, such as a joint where the screww would bed in end grain - there, the dowel would be catching long grain.
Samson,
Good point. I forgot about that. However, I think that in most situations, the screw in end-grain would have enough holding power. One plus about screws is that they are self-clamping.Chris @ http://www.flairwoodwork.spaces.live.com
- Success is not the key to happines. Happiness is the key to success. If you love what you are doing, you will be successful. - Albert Schweitzer
I got nothing against screws, and don't use Miller dowels. I was just curious if some test or something had shown screws to be stronger. I guess it might also depend on the details of the screw (length, metal type, etc) and wood in question.
Samson,
I've got absolutely nothing against wood joinery, but I've always viewed metal fasteners to be stronger and quicker, with their downside usually being their ugliness.Chris @ http://www.flairwoodwork.spaces.live.com
- Success is not the key to happines. Happiness is the key to success. If you love what you are doing, you will be successful. - Albert Schweitzer
I think that in most situations, the screw in end-grain would have enough holding power. One plus about screws is that they are self-clamping.
I find that screws in end grain are rarely strong enough joinery to be acceptable. Miller dowels are also self clamping and in fact the action of hammering them home imparts far greater tightening force than even screws in flat grain (the screws are FAR FAR less strong in end grain and frequently cannot snug the joint even softly without stripping out). Further the Miller dowels save time by producing a finished look similar to plugged screws in far less time than screwing a joint and then plugging it. They excel where side grain is to be joined to end grain (such as the joinery in drawer boxes or carcass sides). I find them nearly as strong and attractive as dovetails with a much reduced time investment.
My personal assessment would be that screws in end grain would generally average something like 10 per cent or less of the strength of miller dowels. Of course the details of the joinery and the skill of the joiner will have an influence but I am saying that IMO Miller dowels are inherently a much stronger system for this situation.
IMO nails are also superior to screws in end grain situations. The nails don't slice the wood fibers into tiny pieces like the screws DO (but Miller dowels are much better than nails too).
The reason I mention the screw's advantage of self-clamping is that on my last project with Miller Dowels, the butt joints opened up after driving them home. I screwed and plugged other joints (because I ran out of dowels) and they all ended up fitting perfectly. Yes, screwing and plugging does add another step, but as another poster mentioned, if the plug is cuts and aligned properly, it can virtually dissapear.
I'm surprised to hear that you see nails as being superior to screws in end-grain. I assume that you are referring to cut (square) nails, rather than round-shank nails. I think that cut nails could be comparable in strength when driven into the face, but I'm not so sure about end-grain. This is not based on any research however.Chris @ http://www.flairwoodwork.spaces.live.com
- Success is not the key to happines. Happiness is the key to success. If you love what you are doing, you will be successful. - Albert Schweitzer
Discussions of the holding power of dowels comes up regularly in woodworking forums. Whether ANY kind of dowel compares to the holding power of a screw is really beside the point. The point being, dowels, no matter how they are applied in ANY kind of joint, do not provide any increase in the strength of the joint. None. In many cases they decrease the strength.Other than providing a "locking function," as in a dowel through the cheeks and tenon of a M&T joint, they don't provide ANY increased holding power within the actual joint.If a joint IS correctly constructed, there is no benefit to putting a dowel in there, no matter what orientation the dowel is used. Dowels actually take away from potential gluing surface area.If a joint IS NOT properly constructed, a dowel (within the actual joint it self), will not materially improve the problem.Dowels have been used by the furniture industry since mass production began as an aid to assembly and alignment. Somehow, because they were used by large manufacturers in "quality" furniture, the term "doweled and glued joints" became synonymous with "quality." Such joints don't deserve that accord, just because there are dowels in there.The furniture MAY be of higher quality because dowels assured that all the pieces went together during manufacture the way they were designed, without anything moving out of place.The only real strength in a wood/glue joint comes from assuring that long-grain is glued to long grain. Two boards can be glued edge to edge, or they can be turned at 90 degrees such as in a lap joint or a M&T joint. As long as long grain fibers of each piece are in contact and the glue job is done correctly, with correct pressure, time, etc., etc., nothing can improve the quality of the joint. Not dowels, not tongue and groove (another alignment method, not a joint-strength increasing technique), nothing. A properly made glue joint will survive stress that breaks the wood surrounding the joint.A butt joint is inherently weak. Long grain contacts short (end) grain. There is no strength in such a joint. The glue can't adhere to the end grain. It's like trying to glue to the business end of a paint brush.The way to make a butt joint strong is to re-design it as a M&T joint, a lap joint or a saddle joint. Putting a screw through the cross member into the long grain of the "butting" piece is a weak solution, because the screw will pull out of the end grain, as has been mentioned. However, the screw will not pull through the cross piece because it's head prevents that (to a degree). So it is strong in the cross member but weak in the end grain member.Conversely, a dowel is strong in the end grain piece, but weak in the cross member. It is true that a dowel will hold well when glued into the end grain piece. The long grain of the dowel and long grain of that member are aligned and that glue joint will be stronger than the wood itself. But the dowel in the cross member is very weak. Its long grain contacts long grain in the cross piece only at a very tiny surface (tangent) in the hole drilled in the cross piece. All other contact is long grain to short (end) grain in that hole. The dowel will eventually pull out.There is one way to strengthen a butt joint, short of properly redesigning it. Screw the joint together, but put a dowel through either face of the butting (end grain) piece, so that the screw passes through the dowel's side, and bites into those fibers. At least the screw will have holding power. But the joint still will not be as strong as a properly-designed glue joint.Rich
"The point being, dowels, no matter how they are applied in ANY kind of joint, do not provide any increase in the strength of the joint."
That's ridiculous. Dowels most assuredly increase the strength of butt joints. Which is, of course, the only circumstance in which you would consider using a Miller dowel!
"Conversely, a dowel is strong in the end grain piece, but weak in the cross member."
Are you familiar with Miller dowels? Not just dowels, but Miller dowels? When properly sized for the joint, a Miller dowel is glued long grain to long grain in the on-end piece (thus providing plenty of glue strength in that half of the joint), and mechanically interlocked in the cross-grain piece (thus mooting the issue of a weak glue joint there). A Miller dowel is comparable to a wedged through-tenon in terms of its strength properties.
-Steve
Steve, I was about to write something similar. You said it all very well and concisely.
One thing though. In my experience, even regular (non-miller) dowels are decent in butt joint type senarios. I think of them as akin to floating tenons, except that they are "through" tenons and round instead of rectangular. I don't think the roundness - the only real difference - deprives them of all their strength.
Steve,It's a little hard trying to exchange ideas with someone who immediately labels the other's words as "ridiculous." So I doubt you'll have an open mind about any of my explanation. I would have had no problem with, "I disagree, here's why." But ridiculous is a little harsh.Anyway . . .Yes, I am familiar with "Not just dowels, but Miller dowels."They are an interesting idea, but are not a substitute for better design of the joint.In many decades of woodworking, I've come to respect the wisdom that has been accumulated over the ages regarding the structure necessary to achieve a strong glue joint in wood. The weakness inherent in a butt joint and the design flaw of a device such as the Miller dowel were known and described centuries ago.The Miller dowel is an attempt to utilize the "cap" of the dowel, much like a screw head, to keep the device captive in the cross piece of a butt joint, to permit a worker to construct a butt joint with a sense of confidence that it will be strong.I have 2 problems with the attempt to raise the status of the dowel to a fastening device. Again, I regard it as an alignment device, nothing more.First: there are many who have no knowledge of the inferior holding power of dowels, who believe that somehow doweling is a strong construction method. Here's a "high-tech" dowel that implies even greater performance, further misleading someone who needs to understand just how dowels actually perform in a glue joint.Second: for those who DO have understanding of dowels in a glue joint, the unique shape of the Miller seems to solve some of the problem. That, too is misleading.Yes the Miller, like ANY other dowel is very secure in the long grain piece. Properly glued, it wont come out of THAT joint. It's alleged advantage in the cross member, is that its "cap" locks in into that piece, preventing it from pulling through, much like the action of a screw head. The long grain of the Miller in the hole of the cross member has no greater holding power than an ordinary dowel as it contacts precious little long grain of the cross piece within the hole in that piece.Have you ever made a M&T joint, entirely by hand? Using a real mortise chisel? While chopping the mortise, some care must be exercised regarding the long grain section, the width of the mortise, just beyond the mortise and between the end of the piece.It is VERY easy to "pop" that section of tissue right out the end of the board while chopping the end the mortise there. There is very little strength in the surrounding tissue to hold that little "plug" in while chopping and levering out loosened mortise waste.In fact, many books advocate cutting the board several inches longer than the finished size, chopping the mortise, then cutting to size. The long "horn" out beyond the mortise end resists the tendency to lever out that long grain tissue.Even then, carelessly fitting the tenon into the finished mortise, can still pop out that area, if the tenon is angled in and levered about. Even assembling the joint straight in can pop out the area by piston action when there is glue in the joint and it is very close-fitting.The head of a Miller dowel is "connected" to its shaft with the same end grain tissue as the end "plug" of a mortised piece and can pop off the shaft in the same way when stress is applied. Additionally, the only real strength it confers to the joint, while it's intact, is the very small holding action of the narrow "shoulder" provided by the cap.I don't use butt joints at all in any kind of construction. I design every joint to have mating long grain to long grain contact. I believe that is the way to do it.Are there examples of commercial construction and custom woodworking in which long grain to short grain joint construction exist? Even abound? Yes! to a very large degree. I would not even argue that examples of incorrect joinery far exceed correct joinery.Will people who find butt joints acceptable change their practices by reading the kind of information I've presented here? Not likely.That doesn't make the information any less valid. Today, or 100 years ago. Or 100 years from now (unless some new adhesive is discovered or invented which actually "welds" wood fibers no matter their orientation). I know how to ensure that long grain is always contacting long grain in every joint I make, many others do, and we'll go on doing it that way. It is every bit as easy to do it right as to do it any other way. And I subscribe to the philosophy that excellence is its own reward.Some may find this interesting and put it to use. There are TONS of woodworking books which discuss this issue at great length.Rich
Edited 7/13/2008 12:55 pm ET by Rich14
Just out of curiosity, Rich, what joint do you use to attach shelves or dividers in a deep carcass? I know sliding dovetails are theoretically possible, but have you made any 24" long? I know wedged through tenons are also an option, but the designer might not want that visual element. Some might use a dado, but that's not too strong as it has no long to long contact. Would the use of dowels to add to a dado improve such a joint? There will be long to long in both parts.
Samson,First, let me explain that no less an icon than James Krenov would take me to task for the position I've expressed. But that's my story and I'm stickin' to it.In many of his small wall cabinets, the bottom overlaps the bottom end of the sides (hangs under them) and is attached to them with, yes, (gasp) . . . wait for it . . . dowels! He feels that the dowels have (I'm forgetting the exact quote) "more holding power than they're generally given credit for."He places about 10 dowels in a line of 12". There isn't much holding power in each of those dowel joints, but the combined strength of 10 dowel joints in 12 inches gets the job done. I doubt if any of his cabinets have come apart. (but then again, I have no information about that.) Expedient, efficient. Who's to argue with Mr. Krenov?Well, I would not do it that way. I don't have to design for anyone but myself. I would use dovetails, if they complemented the overall design, wedged tenons, finger joints (through or blind).So, to answer your question, many dowels would get the job done. Shelves are not necessarily a major part of the construction and may not need to be used for the strength of the piece.I would use through or blind finger joints, wedged tenons. Like that.Rich
Let's take those blind finger joints. If you made round mortises for the "fingers", then shaved the corners fo the fingers to match, would the joint be significantly weaker? Cause that's the same as dowels.
"If you made round mortises for the fingers"??Assuming shelves 3/4" thick. Make fingers of 1/4" to 3/8" width on the ends of the shelving. Rout the mortises with matching-sized bit. Round the ends of the fingers or square the mortises. No real difference in effort.The fingers could alternate with equal-sized spaces or could be made one to an inch or so. A template makes easy work of both fingers and mortises.There will be about 1/2" of long grain on each face of each finger/mortise - 1" contact per. That's astronomically more long-grain to long-grain contact than happens between a dowel and the walls of its round mortise.Rich
Edited 7/13/2008 4:35 pm ET by Rich14
Rich, you are acting as though long grain only happens at the perfectly perpendicular face - within that single degree of arc. This is not how grain works. But enough. Believe what you like and I'll do the same.
Samson,I never said that long grain "only happens at the perfectly perpendicular face - within that single degree of arc."I said that there is precious little long grain in the hole/mortise. I can't say how many degrees of arc are long grain, but most of the grain that the dowel encounters there is end grain. And when most of the grain that is available in a glue joint is end grain, that's a weak glue joint.Rich
Well, I would not do it that way. I don't have to design for anyone but myself. I would use dovetails, if they complemented the overall design, wedged tenons, finger joints (through or blind).
No offense, but it seems you'd be overbuilding for the sake of being 'proper'. The dowels fit Krenovs designs, and like you I've never heard of one of his cabinets falling apart.
Are dowel joints as strong as mortise and tennon? No, but they are strong enough in many situations. Specifically those that are not stressed on a regular basis.
If I say that a statement is ridiculous, it's because I believe that statement to be, in fact, ridiculous.
This statement:
"The point being, dowels, no matter how they are applied in ANY kind of joint, do not provide any increase in the strength of the joint."
certainly fits that description, since anyone who has reinforced a butt joint with dowels has ample evidence to the contrary. The idea that a butt joint with dowels is no stronger than a butt joint without dowels isn't just ridiculous, it's completely ridiculous.
I'm aware of the anisotropy of wood, by the way, and am familiar with the differences in shear strength in different shear planes. Next time you chop a mortise, try putting a clamp on the sides of the mortise at the end, and see how much doing so reduces the risk of blow-out. Now think of how you might apply that principle to the mode of action of the Miller dowel.
All you had to do was say why you didn't like to use dowels, and everything would be fine. But when you begin with a statement that is blantantly false, everything that follows is suspect.
-Steve
Since your method of communication is to immediately pass judgment and cast derision, rather than, at least, listen to the reasoning behind a statement that you find completely upsets your notion of how something is, there is no hope in exchanging anything with you.You don't want to listen, your mind is closed, you became instantly angry and essentially started off by telling me to shut up. You don't have to pay attention to the rest of this, either. I'll just post it for anyone else who might want to think about it.A butt joint is almost the weakest kind of joint in woodworking. (The only weaker joint is one that contains nothing but end grain to end grain contact) There is no strength in its union of long grain and end grain. A plain (non-reinforced) butt joint will fail 100% of the time.Epoxy glue and polyurethane glue manufacturers' claims, notwithstanding.How long it will take to fail is variable, but it will fail. The test of a properly-made joint is that it will NOT fail, that it will last as long as the wood itself. After all, a well made glue joint IS stronger than the surrounding wood.In a practical sense, or "real-world" test of the length of time a good wood-glue joint should last, in fine woodworking, the answer is at least an average human lifetime, and preferably more, much more. We have had glues that last that long for a very, very long time, and we certainly have modern formulations that are at least as good as the stuff that boiled down horse hide can provide.A butt joint assembled with dowels WILL last longer than one without them. That's true. If nothing else, just their mechanical holding power, without any glue at all can keep the joint together for some time. And with glue, the joint can last even longer.But the fact is, even with dowels and glue, a butt joint will fail. It will eventually fail very close to 100% of the time, in NORMAL use. It will fail in a much, much shorter time than that which we expect "good" joints to last. It is undependable and unpredictable, unlike other joints which we know to be sound and on which we stake our reputations when we build furniture and other things that we expect to last. It will not resist stress that other joints do. It will not withstand seasonal changes and wood movement. And the dowels will let go from the cross members of the joint. THAT you can take to the bank.Since doweling a butt joint will not prevent the fact that it will fail, most if not ALL the time, doweling does not increase its strength.What's failure? Completely coming apart. Loosening so that movement is apparent. Need for taking apart and regluing, or re-engineering. Internal movement that lets other parts of a piece lose their alignment or relationships. A good joint never allows any of that.Maybe I should have said, "in a properly constructed joint (and that excludes butt joints from consideration by definition) there is nothing that doweling can do to increase the strength. To the contrary, dowels can weaken the joint by reducing potential gluing surface."But then you would have started jumping up and down anyway, screaming your head off that I had the nerve to exclude butt joints by definition. Ideas aren't ridiculous, Steve. Just the way some people behave when they come across some idea they don't want to tolerate. What's ridiculous is your silly anger.RichRich
Your argument makes no sense. You're saying that since a plain butt joint and a dowel-reinforced butt joint both eventually fail, they both have the same strength. (All joints can be made to fail, of course, but we'll ignore that little detail for now.)
If I make a screwdriver out of basswood, it will fail as soon as I try to tighten the first screw. If instead I carve a screwdriver out of butter, it will also fail, probably before I even get it near a screw. So, since they both fail, they must both have the same strength, right?
That's ridiculous.
By the way, I'm not angry, and I'm not jumping up and down. All I'm doing is pointing out ridiculous assertions.
-Steve
Steve,You're just proving my point. If you make something out of the wrong material, or use the wrong technique to make it, it will fail under the very (normal) conditions that it must perform.The butt joint, with or without dowels will fail in NORMAL service. It doesn't have to be unusually stressed.Rich
Hi Rich ,
Good post , and you have pointed out some of the aspects and variations of dowel / joinery and butt joints as well .
Imo dowels are more for alignment then for strength and I do agree about butt joints with some exceptions , such as reinforcing corner blocks in doweled frames and chairs and furniture can be made with the use of dowels and have a successful outcome as well as many other works .
It's soo hard over the internet to tell when a person is mad or smiling or what.
I must admit I too was surprised by your words , what came to my mind about how any joint could not be made better by the use of dowels "not a quote" is a stile and rail door frame ,an entry door or cabinet type door . I've made a few entry doors and thousands of cabinet doors and done a fair amount of restoration work as well as reproduction of existing pieces including doors .
What I have found is even with a standard stub tenon joint a screen door or cabinet or entry door frame can be tremendously fortified by properly sized and placed dowels . The entry doors use like 1/2" - 3/4" dowels and imo they tend to stiffen across the joint to aid in the racking and slamming a door can be subjected to.
I hope you can understand how any of us can sound angry but , this is not about always agreeing , we share and teach and learn and participate in all types of conversations , and Steve is probably one of the smartest and most helpful member of this community . He can speak for himself for sure , but there are so many other real things in this life to be concerned with , it would be great if we could figure out how to avoid the negative energy .
dusty , who doweled for years but little more
Rich even well constructed joints will fail, given time and loading.
I am in the process of restoring a bedroom set. The dovetailed joints in 11 of 16 drawers are no longer holding tight.
And, I agree with Steve. Dowels do add considerably to the strength of a butt joint. To state categorically that they don't is, well just plane ignorant, in the purest sense of that word from it's root of ignore. I have seen several articles down over the years, that show substantive increases of strength of doweled joints.
Dusty and Jigs,But I DID define "failure" of the joint. Yes, any joint can be made to fail if it is stressed beyond the design parameters.My only real point is that a butt joint, reinforced or not, will fail under normal conditions. Normal wood movement will do it. It is never a reliable construction.That's not the kind of worry that any good craftsman would ever have to have, for instance, about a correctly made M&T joint. Even under severe use.My definition might not meet yours, but since the joint will fail, reinforced or no, the reinforcement adds nothing but a delay in the inevitable. And that places the joint in the unacceptable category.I'm not trying to be ignorant (in the "purest" sense). In fact I am deliberately NOT ignoring the ultimate fate of this kind of construction, as some may be accused of when they say that a reinforced butt joint IS stronger than a plain one. Delay in failure is interesting, but useless.We've all made our points, though. I'll just muddle along, avoiding butt joints and using dowels only for alignment purposes.Rich
Chris;
I never use cut nails in end grain. Hot galvanized nails are much better than screws. The rough texture created by the hot galvanizing will grab in the end grain much better than screws where their threads slice the grain to pieces. Sinkers are also good as they have a hot glue coating that melts as the nail is driven in and glues it to the wood. Ring shank nails are better than either hot galvanized or sinker nails though. Unlike the screws the ring shanks will slide into the end grain without damaging it but their ridged surfaces are very resistant to being pulled back out. In addition, where there is room to do it, driving nails in at opposing angles (sort of dovetailing) solves many problems with end grain fastening. I still prefer the Miller dowels to any of these options in most situations though.
I usually predrill for dovetailing the nails (through the edge grain piece that is being nailed on... it helps prevent splitting and guides the nails in the proper angle).
Regarding your last project; I have had excellent results with my Miller dowels self clamping. I use only a small amount of glue (three fine lines running lengthwise) and (due to the VERY tight fit of the dowels into their holes) I normally get an instant grab as the glue swells the dowel and dries as it contacts the dry hole surfaces. On occasion I have had to clamp warped pieces for a few minutes so that the warp didn't pull my structure apart before the dowels could grab. Perhaps you are using too much glue which would then act as a lubricant instead of grabbing? Glue applied into the hole can prevent the dowel from being driven in solidly (by building up in the bottom of the hole and creating a powerful hydraulic piston effect). It is also important to clear the drilling chips from the hole effectively... otherwise they can interfere with driving the dowel in properly. I like to tip the piece upside down and tap it where it is doable. Places where that is too hard to do, I will use a blast from my air hose to suck the chips out.
PS: The dovetailing strategy is useful with screws as well and vastly improves their peformance in end grain situations.
You know, now that I think about it, I don't know if I've ever used a nail in a piece of fine woodwork. Ring shank nails do seem like the way to go, though, as far as nails are concerned. Although the "dovetail" nail approach also seems like a good choice.
When drilling for Miller Dowels, I do make a point of trying to clear out the hole as best as I can. I usually run the drill bit in and out a couple times to remove the chips. I have never tried turning the hole upside-down and knocking any chips out. Also, I have been smearing my dowels with glue as I always do - covering the entire surface. Next time, I will try a couple beads as you suggest. When you say "instant glue", are you referring to a CA?
Thanks for your help,Chris @ http://www.flairwoodwork.spaces.live.com
- Success is not the key to happines. Happiness is the key to success. If you love what you are doing, you will be successful. - Albert Schweitzer
Chris;
I actually use "Titebond Extend", which is my standard shop glue. It is a bit thinner than regular Titebond and I apply it from a round salad bar type squeeze bottle with a fine tip. I squeeze out three fine beads running from the tip to the head of the Miller dowel and then hammer it down. I have tried several ways and this seems best. I started out trying to get plenty of glue on and make sure the whole dowel was covered... but I now have confidence in these dowels and realize that they won't let go if some surface isn't glue flooded. I've never had a joint made with these fail. I don't expect that I will. Ideally my glue beads are sized so that I get just a dab of squeeze out at the top when the dowel is fully seated... but I don't panic if I don't get any squeeze out at all... I just make the beads a tad larger on the next one. Softer woods take a bit more glue.
I think that the thinner glue (extend) spreads better in the joint and sets faster (even though it's designed to set slower) because it is a thinner coat and the moisture is wicked out of it very quickly.
I don't use nails in most cabinetry either but I make lots of kinds of things. I employed the dovetail nail system recently to assemble cedar 2X6's for raised bed planting boxes. The nails are more dependable where the wood will be in constant soak and dry cycles. You can easily use both ring shank nails AND the dovetail nailing system. For decking I use a nail gun with glued ring shank nails... you cannot remove a board once it's nailed this way without destroying it. The heads will pull clear through the plank most of the time (screws pull much easier... even in side grain).
I think that Miller dowels are a little more forgiving in the area of wood movement than screws.
One thing to consider with the miller dowels is that you end up with exposed end grain whereas with a screw and a plug the plug can be cut with a plug cutter and have face grain so it will not show as much. I have used the miller dowels and they work well but I doubt they have the shear strength of a steel screw.
Troy
I only consider introducing screws into a project if there is not enough glue surface for a strong wood to wood bond or if it needs to be disassembled later. The miller dowel would provide a connection into end grain that is stronger than the surrounding wood. I don't see the point of a fastener being any stronger than that.
In my opinion Miller Dowels are easier to use than screws. I've used them in several projects and like how the "system" works. For my projects I used a contrasting wood dowel so "grain match" WAS NOT A CONSIDERATION.
This forum post is now archived. Commenting has been disabled