Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

A55 Smoothing Plane

How many tails?

Posts

How many tails?

#1

How many tails?

Patrick Gibbons, Katy, TX

>I'm dovetailing a carcass with 10" wide boards. The dovetailing will not be seen once the project is complete. I'm going to use modern glue so it seems to me that there's a point after which more glue surface is redundant. Unless someone can give me a reason to the contrary I'm going to cut only three tails per corner. Will that suffice? Answer in complete sentences and show your work.

Re: How many tails?

#2

Re: How many tails?

Pam Niedermayer, Austin, TX

>I think it depends on how wide you want to cut them. Do you mean 3 plus the two halfs? If so, that should be sufficient. If not, I think I'd add 2 per, perhaps because it's easier to cut them freehand.

Sorry, no photos available.

Pam

Re: How many tails?

#3

Re: How many tails?

Warren in Lancaster, PA

>I think five tails and six pins (or six tails, seven pins) would be more like it. You have to realize that the dovetail joint is a mechanical as well as a glue joint. Each pin is a wedge that contributes strength to the joint, so economizing on pins limits strength. If you are set on depending on glue only, why bother with the dovetails at all?

Re: How many tails?

#4

Good point Warren

Patrick Gibbons, Katy, TX

>But I wonder if increasing the pins by a count of two will really add that much strength. I'm more concerned about the long-term effects of wood movement on the ends of the corner but once again with modern glue does that matter? I've used lap dovetails on some basic construction with only two pins and one tail and have had good success. What got me thinking about this was an article by Mike Dunbar on making a 17th c. cabinet in pine with 6-1/2" sides and he used only two tails per corner. This was in a Fine Woodworking article. He was copying an original, though his using only two tails seemed to have been his take on what was necessary. Thinking out loud, I don't think I'll live long enough to see if more tails makes a difference. I wonder if the size of the two end pins at each corner is more important than the number of pins in total? What do you think Warren?

Re: How many tails?

#5

Re: Good point Warren

Warren in Lancaster, PA

>Increasing the number of pins will add strength. Long term both the tail board and the pin board will shrink and swell seasonally; they will move in concert. Long term evidence (centuries) is that even dissimilar woods do fine.

Dovetailing boards was pretty primative in the 17th century. Many drawers had only one large tail and two pins, reinforced with large nails going through the tails. The nails are apparently original. With only one pin at each end they are vulnerable to twisting forces and I think they would have failed without the nails. When we make 17th century reproductions, we follow this same scheme (with nails) even though we know more tails is stronger. As the 18th century went on pins became finer and more numerous and perhaps they went overboard at the end.

Re: How many tails?

#6

Re: How many tails?

Bill Tindall, E.TN

>As I understand it these 10" pieces will define a box like for some sort of cabinet. A cabinet with an additional overhanging top endures little racking force so the main function of the dovetails for this "false" top is to retard cupping of the sides. I have seen many case projects where instead of a full width top the "false top" is simply a dovetailed "rail" front and back that provides something to screw the overhanging top to the case. That said, if this case has an overhanging top fastened to the box defined by the 10" pieces then 3 tails seems adequate to me. But I would probably have used 1 or 2 more because I need the practice cutting them.

Re: How many tails?

#7

two tails?

Ian (in Sydney, Australia)

>Patrick

apart from holding the sides to the top what purpose to the dovetails serve?

(let's set aside for the moment issues arround your own self esteem and knowing how well the hidden parts of the project are constructed)

On the basis that the joint is hidden, aesetics � which typically defines the number of DT used � is not a consideration.

If the hidden top's function is to only to hold the sides the correct distance appart I can't see the need for more than two (2 x 1) rails (one at the front, the other at the rear) dovetailed into the sides.

You might add a third rail in the centre if you were worried about the sides cupping.

On this basis, three DTs in the top (with pins about 1" wide) will be perfectly adequate, and there is no need to make 1/2 pins at the ends. Or have I confused tails with pins ?

Given the tests I've seen some "modern glues" pass, there's even an arguement that you don't need any DTs at all � but let's not go there.

cheers

Ian

Re: How many tails?

#8

Re: two tails?

Patrick Gibbons, Katy, TX

>This is a design of my own so it can go a lot of places. It will be a carcass 40" high by 30" wide with two shelves and two drawers below the bottom shelf. There will be 6" moldings below the drawers that hide two secret compartments. The top will have moldings about 3" wide that will hide the dovetails and create a shallow space on top of the carcass that I will probably add a door of sorts that will be hinged at the back. The top of the carcass provides the bottom of this shallow space. If it isn't obvious, this is for my two young sons who love the idea of secret compartments.

Re: How many tails?

#9

anvil specifications

Bill Tindall, E.TN

>My grandfather once remarked to a friend that I (grandson) could break an anvil. I don't remember what I had recently broken to deserve this remark. Presuming this ability is common among male youth, add more dovetails.

Re: How many tails?

#10

Depends on the thickness, Warren

Andrew F in Australia

>Hi Warren,

It depends upon the thickness. The distance between pins is approximately three times the thickness.

So, a 3/4" board (I'm guessing,) would have a pin spacing of a shade over 2" between pin centres.

That makes it about 4-5 pins and about 3-4 tails.

If it was 1", then the pin spacing is more like 3", so you'd have two tails and three pins.

For strength, use a slope of 1:6 on the dovetail.

Cheers,

Andrew

Re: How many tails?

#11

Oops - - -> "Patrick"

Andrew F in Australia

>Sorry about that Patrick, addressed to the wrong person by mistake.

CHeers,

Andrew

Re: How many tails?

#12

Question for Andrew

Patrick Gibbons, Katy, TX

>Andrew, what can you tell me about this formula? Where did it come from? I would guess it was standard practice taught to you but it fascinates me to consider the why of it.

Re: How many tails?

#13

Re: Question for Andrew

Andrew F in Australia

>Hi Patrick,

You're right, it was trade training. Steeped in history, I suppose. No idea of its origins.

But I do know that it works and looks about right.

As an aside, a dovetail slope of 1:6 has been historically used for carcase/strength dovetails and the 1:8 figure for show dovetails.

The only reason that I can surmise for the (not trade correct) 1:6 being suggested for softwood and the 1:8 being touted for hardwood is that, historically, the drawers were made from hardwood and the carcases from pine.

I've been wrong before. About 1000 times every weekend.

Cheers,

Andrew

Re: How many tails?

#14

The rest of the rules

Andrew F in Australia

>Hi Patrick.

I oversimplified things earlier.

Here's layout in a nutshell

Strength dovetails = 1:6, show dovetails = 1:8 pitch

(maximum) Width of pin = 1/2 -> 2/3 the thickness of the timber with the tails. For 3/4" timber, use 10-12mm (3/8 -> 1/2") wide pins at their thickest point. Show dovetail pins more like 3/8" thick, strength dovetail pins closer to the 1/2".

Distance from pin centre to pin centre is about 4 x AVERAGE width of pin for strength purposes, 5 x AVERAGE width of pin for show purposes.

These spacings are approximate and need to be tailored so that the dovetails are uniformly spread across the board.

Earlier I said that the spacing between pin centres is about three times the thickness of the pin. This is for strength purposes as you indicated. Correct distance AS A GUIDE is this. The more detailed layout description above gives you, for 3/4" stock, distance between pin centres of just under 2", and for 1" stock, distance between pin centres of a bit above 2".

This gives you the five pins/four tails for strength in 3/4" and four pins/three tails for strength in 1" stock.

As always, your eye is the best judge as to whether it's suitable at the end of the day.

Must away - work beckons and I need to be on the road, 7:27am Monday morning.

The wording above is slightly clumsy, but I haven't the time to reword it. Trust that it makes sense.

Cheers,

Andrew

Re: How many tails?

#15

Re: The real story

ThomD

>The point behind 1:8 was that they were just as strong as 1:6. That is what makes the show point. What is being shown is that the craftsman understands that for a given type of wood that is all the slope that is required to make a strong joint. Would not be as true with pine. So just as proportional spacing purports to show that one does not need as many dovetails in the middle run as the corners, 1:8 shows an understanding of wood hardness. And it is all done in angles that originally were not available in router made joints.

Then along came the Leigh jig which allowed much of that style to be done by machine, essentially a machine that in it's conception made itself obsolete, since who needs a machine that apes joints that only exist because a machine can't ape those joints.

At a deeper level this is the kind of conversation in craft that is at the art level, even if unimportant to the average person. The didactic element is present if not in retelling War and Peace in a piece of casework, at least to the extent of telling the story of the discussion of means that is underway in the craft.

Re: How many tails?

#16

Re: How many tails?

ThomD

>When cutting dovetails one always has to remove the equivalent of one board's endgrain area times the depth of the joint. Whether you cut twelve pins or 2. So to go to a decent number of pins only costs a few dividing sawcuts, it's the chopping/wasting that takes the time. Admittedly if one were to use too many pins it would eventually get constricted to work in very narrow spaces. The only further complication is whether the joint fits together, and incompetence hasn't been the main guiding principle in sizing joints since presumably the rule setters had it wired. I would probably fire in a pin at least every 2 inches.

I think the premise of modern versus ancient glues is false. Hide glue is stronger than most modern glues, and all glues are pretty darn strong anyway. But I have seen a lot of well designed furniture that held together with just a few nails in the corners so it is doubtless true you can cheapen a lot of structures to a degree without paying much of a price for it. Built in structures gain so much strength from their surroundings that little else is often required. In most cases pieces for kids should be secured to the walls.

Re: How many tails?

#17

Re: The rest of the rules

Warren in Lancaster, PA

>I don't think you did your math right, Andrew. The boards are 10 inches wide. Using your new rules, I get 5 or 6 tails for 3/4 stock and 4 or 5 tails for 1' stock. This is pretty close to what I suggested three days ago.

1/2 of 3/4 is 3/8. times 4 is 1 1/2. goes into 10 6.67 tails.

2/3 of 3/4 is 1/2. times 4 is 2. goes into 10, 5 tails.

!/2 of 1 is 1/2. times 4 is 2. goes into 10, 5 tails.

2/3 of 1 is 2/3. times 4 is 2 2/3. goes into 10, 3.75 tails.

Re: How many tails?

#18

Re: How many tails?

Wiley Horne--So. Calif.

>Thom,

Great point about the waste! I really hate those long expanses of wood to waste on the pins board (when you have few pins). As you say, there's the same amount of wood to waste no matter how many pins you use, but I much prefer to saw more pins, in part because the waste job then comes in smaller packets and is more chisel sized.

Wiley

Re: How many tails?

#19

Re: The rest of the rules

Andrew F in Australia

>Thanks Warren.

As I said, I was in a rip-tear-bust rush this morning.

Tell you what - I'll draw them out and post them. I've been working in 30�C ambient heat and full humidity over a furnace today. Maths wont be the strong point.

Looking to the attached graphic, you'll see that a 1:8 slope and a maximum width of 3/8" might not be workable.

I'd personally go the route with the smaller amount of tails, just for time's sake.

Cheers,

Andrew

Note: Dimensions in mm. 25.4mm = 1 inch


img

Re: How many tails?

#20

Now we're getting somewhere

Patrick Gibbons, Katy, TX

>Thanks Andrew and everyone else who commented on this question. Nowhere have I ever seen any discussion concerning why the number of tails and pins. It seems to me that by understanding why we get a better grasp on what a good dovetail joint is. Now if I was only any good at making them...

Re: How many tails?

#21

Pic!?! Nice movie! Thanks for posting!

Scott Burr in Ben Lomond, CA

>Andrew, that was fun to watch, thanks! Funny about the spacing, guess we all have our own rules. As a Polish engineer I use to work with use to say "Rules are made to be bended" I still live by those words daily...

Re: How many tails?

#22

Here is a current example for assessment

Derek Cohen (in Perth, Australia)

>Some of you will recall the brace I recently posted to the forum. Well I am in the process of building a box for it. A box is just a drawer with an enclosed top.

My intention was all along was to cut the box open, but somehow I forgot it was a box and not a drawer, and so wound up with fewer dovetails than I originally planned!

The board thickness is a little under 3/8" and the height is 5". There is a double thickness in the lower half that houses a drawer, so this area has more reinforcing. The dovetails are 7:1. The woods are Jarrah and Tasmanian Oak.



Having screwed up with the number of dovetails, and not wanting to do it again (!), I decided to press on and see how it turned out - essentially hoping that the current mechanical and glue setup will surfice. I do have an option that I can add, if needed. Hopefully some here will provide their opinion as to whether that will be necessary.

The box was sawn through its centre on my bandsaw, and the edges planed flat (this where a set of edge planes are really useful). In the picture you can see the cut. In the "real world" it is almost invisible. I am very pleased with the fit!





The potential problem now lies with the unsupported (?) corners of the box opening. Usually I would have added in a larger dovetail and split this when cutting open the box. Of course this is missing completely. At present the box feels very strong. The question is, will the glue be enough to bond the sides, or shall I need to add a triangular section brace the length of each corner. I'd like not to add the brace if possible.



Apologies for the fuzzy photos.

Regards from Perth

Derek

Re: How many tails?

#23

Re: How many tails?

Pam Niedermayer, Austin, TX

>Oh, I was just talking about the ease of bisecting bits of the board for freehanding. On 10", 5 tails are easy to lay out by eye, perhaps easier than 3, and 4 is inadvisable.

I agree with you about hide glue.

Pam

Re: How many tails?

#24

Show off

Patrick Gibbons, Katy, TX

>Just beautiful. Don't know about your questions, but the wood, the box, the brace... beautiful.

Re: How many tails?

#25

question

Dan Donaldson

>First, nice looking box. The question that I have is that I cannot see from the pictures what holds the lid in place. What keeps it from just sliding off the lower part?