Dowel Joint Design, Strength, and Precision

A long discussion on tweaking dowel and hole size to get good alignment and a strong joint. January 19, 2011

We presently use a 3/8 x 2 inch fluted dowel. I can't remember the size of the drill bit, but it is a reasonably snug fit to the dowel. I have often wondered about the mechanics of the dowel joint. I always assumed it did not act like a tenon as much as a membrane that provided a continuous link between the stile and rail.

What would be the difference in strength if I was to step a size down in dowel (5/16 instead of 3/8) and use more glue in the dowel hole? The dowel would serve as a bridge or conduit for the glue and the glue itself would produce the membrane. Assuming sufficient glue, would this produce a strong connection?

The advantage of the smaller dowel is that it would simplify alignment issues where rail meets the end of stile. You have to pay some attention to chip buildup at the boring machine stops and I was wondering if this would allow us to have a slight misalignment here from time to time, but still get the strength we were looking for.

Forum Responses
(Cabinetmaking Forum)
From contributor H:
If the surface area of the two pieces being joined is glued also, I see no reason why this would weaken the joint.

Glue two pieces of wood together with a butt joint, then break it. If it breaks at the glue joint, find another glue. If the wood breaks first, it's good glue!

From contributor J:
This sounds like a very bad idea to me. To be fair, I don't understand what you mean by "a dowel acts like a membrane," so maybe you've had some brilliant insight. But I doubt it.

My understanding is that most glues are strongest when their film is very thin. Using a dowel that is undersized for the hole will inevitably mean that the glue film is quite thick, so I'm betting that the joint would be weakened considerably. Also, most types of glue will take much longer to grab and, finally, dry in such a situation, which would dramatically slow assembly. Hope you have a whole lot of clamps...

From contributor K:
A loose joint with too much glue is a weak joint. It will fail.

From contributor N:
I agree with contributor J. The only way that dowel joint would have any strength was if you were to use a thickened epoxy for the glue. A dowel joint acts very much like a tenon (although not as strong) and must be a snug fit to be effective. Think of a repaired chair. If the mortise is worn and you don't make a tight fit, the joint will fail, very quickly.

From contributor I:
I agree with contributor J as well. I will add that a properly sized dowel hole should be too snug to push in with your fingers. The dowel is a mechanical fastener. The glue causes the dowel to swell and locks everything together tight. This usually takes about 5 minutes. After the dowel is inserted into the hole (with glue in the hole already), it will absorb the glue and swell to a tight fit.

Do not think of the glue as being the primary means of holding the parts together.

From contributor U:
Definitely a bad idea to use a smaller dowel. If you are talking about keeping the 3/8" hole and using a 5/16" dowel, a really bad idea. Think of the strength being measured this way - more glue surface area (the contact area between two surfaces) equals more strength. I guess I can see where you might think more glue means more strength, but not so.

We just increased our dowel size from on average 1/4 x 2" to 3/8 x 2" and the stronger joint is quite an improvement. If you could measure the surface of 5/16 dowel against the 3/8 dowel (there are formulas for this), it would make more sense, as there is a lot more area helping the glue to achieve a bond. Plus the mechanical strength of the joint is greatly improved.

Whenever we make a change in dowels, glue or our methods, we glue up some dummy panels for testing. Then we drop them and do things that simulate rough treatment. The best test is put the pipe clamps on them in reverse and press them apart. If the dowels slide out, it is a failure. If the wood starts popping and tearing and the dowels hold, we know we are on the right track. I suggest you do the same thing to several pieces and see for yourself.

I also have found that in our use, the fluted dowels do not hold as well as the ones where the groove spirals around the dowels.

From contributor G:
I also agree that this is a terrible idea. Contributor N's chair mortise is a perfect analogy.

A while back I talked to Franklin Glue, the manufacturers of Titebond. I asked what the thickest glue joint should be. Their response was .006". Assuming the dowel was perfectly centered in the hole, you could theoretically drill a .012" oversize hole. Keep in mind that the glue has a very high moisture content. It is not 100% solids. Once that moisture dries out of a thick joint, there will be a lot of air voids.

From contributor M:
I like to make things easy and simple. When joining two pieces of wood, the key is surface area. The more surface area, the stronger the joint. Decreasing a dowel from 3/8 to 5/16 decreases surface area, therefore decreasing the strength of the joint. Standard wood glue is all that is necessary, since it's stronger than the wood is. Changing the glue has different benefits, when standard glue will fail, or the drying time is unacceptable.

From contributor D:
Sounds like you are having an alignment issue, not strength issues. Try compressing the dowels with a pair of the right shaped pliers to slightly decrease the size temporarily until the glue swells it tight. A quick smack should line up the members with a little extra slack.

From the original questioner:
Thanks, everyone.

Contributor D, you hit it on the head. It's not a dowel-match-hole issue, it's a hole-match-hole issue. We use a retractable stop for indexing the stiles and rails to the drill center. Sometimes chips get between the lumber and back fence, which can cause a misalignment. We fixed this on the stick shaper by adding a continuous blast of air against the fence. We just need a better chip evacuation here.

From contributor U:
We have a good dust collection system, but found it was difficult to plumb for our dowel drilling. We use two shop vacs located by our two dowel drill setups. The smaller hoses and the available accessories make it simple to attach, to get maximum chip removal. Plus the vacs are available for general shop cleanup. Also keep dowels in sealed bags or containers to minimize swelling from the air moisture. Also check out downsized or properly sized bits for proper hole size. Excel dowels has some helpful information in their catalog about proper dowel sizes for different material thicknesses, and different types of dowels in a variety of woods. They even offer pre-glued (water mist activated), but I prefer to stay away from these.

From contributor O:
What kind of machine are you using to drill for your dowels? The reason I ask is that the holy grail I've always searched for was a drill bit that didn't wander. That in and of itself causes misalignment issues.

From contributor I:
If you want your doors strong and true to the old school ways, use a haunched mortise and tenon joint. I know a local cabinetmaker here in my town that does this. He has it set up so he can pump them out pretty fast. Nowhere near as fast as a simple cope and stick, but pretty fast when you consider the joinery.

He uses a single ended tenon machine and a mortising machine with a pneumatic drilling head (step on pedal and mortise machine comes down to drill). For doors with an inside profile, he will notch a miter in a sliding table saw.

From contributor U:
I big time agree with contributor M about the wandering drill bit problem. While we use a Ritter to do a lot of boring and also a two hole borer, we actually use in addition two Craftsman drill presses to drill dowel holes into the end grain of at least 1/3 of our rails. Now this is for shutter panels, but we have to be very accurate at drilling to center or our panels will look pretty bad. I always stay with brad point bits and get good results. With the more professional machinery the higher end bits we buy also have brad points.

From contributor D:
When I used dowels for frames, I noticed an alignment problem as well. Turns out after sharpening the bits a few times, according to a machinist, they were undersized. They were out 2 or 3 thousandths. I asked him to show me that on my tape measure.

From contributor O:
The closest I've come to finding an all in one solution was a drill that was made by Black & Decker called the Bullet Bit. The tip was ground with a center drill point then followed by the nominal size. I really need to find a way to resharpen it.

After that I had both Morris Wood Tool and Fuller grind an end grain bit for me. Yes, there is a different style for end grain and side grain. While it improved markedly there would still be a little wandering, at least with 3/8" diameter bits, of which I didn't have issues when drilling 3/4" holes for passage doors, which leads me to try one last thing. I think the last little bit of wandering can be taken out with a shorter bit that will take the flex out of the equation.

From contributor U:
I gave up on bit sharpening long ago. Over the counter bits last us a pretty long time and are not that expensive versus sharpening. I attribute 5% of drilling areas to the wood grain itself, especially end grain. The biggest discovery I made was realizing the drilled material itself will get hot and stick to the bit and cause a wobbling effect and enlarge the hole, messing up my dowel fit. Sometimes this moves the hole slightly off center. To allow for this, we do a lot of double drilling or whatever you want to call it. When drilling a 2" deep hole we will drill down until the bit loads up, maybe 3/4", back out and let it clear or clean itself and then re-drill until the proper depth is achieved.

As far as the wandering bit thing, another trick we do is at the very start, advance the drill bit very slowly and let the brad point and tip get about 1/8" or 3/16" into the wood and form the beginning of the hole. Once done, the hole itself will control the bit location while we then speed the plunging of the bit to drill the remaining depth. We also do this when drilling our louver ends for operable shutters, hundreds every day. It does not take that much longer to do, and really helps our accuracy.