Friday, February 8, 2013

French Connection 10

Continuing on with the drawing work on the 19th century layout exercise of a three-legged bench with various arrangements of scissor braces - or Saltaire crosses as they are also termed.

I started with brace pair 1, in which the faces of the two brace sticks are in the same plane as one another:

There are some tricky bits to the layout of these, especially in regards to developing the end cuts of the sticks to fit around the post and beams to which they fit at a compound angle. The problems can be solved for you with a 3D drawing program, however I prefer to exercise my powers of visualization and work the drawing in a 2D format, as if it were a large sheet of paper, developing the angles and then transferring the information directly to the sticks:

Satisfied with side #1, I moved to what I am calling side #2:

With side two, the two brace sticks are rotated so that their front and back sides are plumb. Certain aspects of side #2's layout are simpler than side #1, however there are complications as well.

One of the complications with this project involves tenons. The drawings in the text show tenons on many - but not all - of the sticks, however I learned well from my previous 4-legged Mazerolle sawhorse build to take such drawings with a grain of salt. While a connection with a mortise and tenon might be shown, that doesn't mean it is possible to actually assemble the piece with joinery in such a configuration! indeed, in terms of the French carpentry models I have looked at, the vast majority of the connections in a piece like this do not employ tenons but are spiked together. Still, I want to see if I can design the piece in such a way that it can go together without recourse to metal fasteners. This may be a case of if wishes were horses.

The first tenon complication i am looking at involves how the lower brace ends connect to the lower ends of the legs. The perspective drawing in the text shows a through tenon:

Let's look a little closer at that connection:

Indeed, if considering only the tenon on brace set #1, then the mortise would be configured as a compound angled parallelogram, and this is exactly what I had after detailing that tenon and mortise:

It's not an easy connection to fabricate, but this is how the French would seemingly prefer to do things, so I'll give it a shot.

There is the lower end of the the right side brace from set #2 to connect to this same area though, and the plan drawing in the text indicates that the tenon ends from the two braces are to pass through the post and meet in line on the exit face:

It's a little unclear whether the tenon from brace set #1 is cut short to let the tenon from brace set #2 pass in front, or vice-versa. Both seem to be indicated as 'possibilities'.

If you look at another corner of the plan drawing, where the brace sets #2 and #3's lower ends will meet at the post, the same sort of connection is indicated:

Sounds like a cool plan to bring two tenons through to meet in one location, however in reality, the shapes of the tenons are completely opposite to one another in this location where brace set #1 and #2 lower ends meet the post:

The parallelogram-shaped tracing on the face of the leg is the outline of the brace from pair #2, which is gray and enters from the left side in this view. You can see that it's form is completely opposite to the parallelogram shaped mortise produced by the yellowish brace from pair #1, and they are at different heights. If I formed a tenon on the gray brace's end and ran it through the post, then it surely would not look the same as what is shown in the perspective drawing shown a few images above.

The book shows that one of the tenons could be held back so as to let one exit out on the face, and it would appear the book means that to be the tenon from brace pair #1. So, following that train of thought, let's trim that gray tenon back so that only the yellow tenon comes through:

 With the front face of the post removed, you can see how I've trimmed the grey tenon back so as to allow the yellow one to pass through to the exit.

Now with the face put back on, the appearance would be fairly similar to the perspective drawing shown in the book, however the gray brace would leave a slit down below the yellow tenon:

That grey brace end tenon could be trimmed a little differently to hold it off the face, but what is a far bigger problem in my view is the sort of mortise one would need to cut for the grey brace:

Compound angled parallelogram shaped mortises are annoying enough to contemplate, but to do one which is partially blind as well does not appeal to me so much.

A look down the brace mortise for the yellow stick reveals an easier path to cut:

Anytime you can cut right through it will be easier to execute. I'm not exactly enthused about the prospect of cutting a blind compound mortise, though I've done them before. It's just very hard to get a clean and accurate result.

So, with that apparent roadblock staring me in the face, I decided to look in some of my other French drawing materials to see if I could find some other approaches to the same problem. Delataille's work Art Du trait Pratique de Charpente does show several iterations of the three-legged bench. On on the cover of volume I (and not found otherwise in the text) is shown a through-tenoned connection at that lower post location in a bench done with each brace pair rotated in the same manner:

Here he shows a tenon for each stick poking though the face of the post.

That approach seemed like it was worth a try at least:

That looks like it would work, however when you remove the braces, the mortises that must be done in the post are structurally weak. here's the mortise for the yellow brace:

 And here's the mortise for the grey brace:

So, that doesn't seem like much of a solution from my point of view at least. One could squeeze the tenons in a little closer to one another, and thicken up the leg a bit to improve the situation, but I'd rather see if there might be a better way to go.

What I have come up with at this stage is to have the tenons emerge on the leg's front face aligned in the same central location, and lightly lapped over one another, the grey brace's tenon trimmed back partially:

It's not what Mazerolle showed in his perspective sketch, but it would work.

Remove the post and you can see what needs to go on with the tenons:

And with the yellow brace out of the way, the bizarro cut on the bottom of the grey brace's tenon is apparent:

What this means in the bigger picture is that the grey brace's tenon must insert fully before the yellow brace tenon can be inserted more than about halfway in. Not sure at this point whether such an assembly sequence will be possible. In fact, considering the entire assembly sequence, as far as what is going to be possible, where there can be tenons and where there can't be tenons, is going to be a tricky matter on this piece. You certainly can't go faithfully off of what the book shows, and that is one of the major challenges when working with such material. Delataille is no better, as some of his illustrations of three-legged benches are clearly showing impossible assemblies.

All for today. Thanks for coming by and if you have any good suggestions as far as connecting this beast together, I'm all ears. On to post 11


  1. Wow Chris, I got dizzy just looking at your drawings of this joinery. Is this something that was actually used? Where do you find something like that? I like to challenge myself but not so sure I could handle that level. I'm afraid I can't offer a thing to help you out on this but it's fascinating!

    1. John,

      thanks for the comment. This project is a layout study in a French carpentry text from the 1850's, so it's primary purpose isn't necessarily a practical one, any more than learning the multiplication tables is directly practical. It's about learning a method which you can apply to practical things, a study that can be realized by constructing this bench rather than having to build an entire roof. I've seen several examples of this piece in constructed form. Actually, this three-legged sawhorse is an exercise which related to placing rotated braces in a timber frame structure between beams and purlins, called "Liens de pente".


  2. I am befuddled but also enthused by your passion for this. Photos please as you build it.

    1. Bruce,

      well I'm equally befuddled at times myself, and that's part of the fun I suppose. Will definitely photo-document as I build it, however the core part of the puzzle - the truly difficult part - is the drawing. Once you can draw such a piece, constructing it is more or less 'just' a matter of carpentry.

      Mind you, it is perfectly possible to come to learn the drawing method without having any carpentry skills at all, however the shortcoming with that approach is that issues such as the tenon problem detailed in the post above will not be apparent unless you have done a fair amount of joinery cutting and have a grasp of those sort of issues.

      The best, I feel, is to combine the intellectual understanding of the piece AND engage the hands in the making of it, to complete the circle of understanding. A large portion of what informs my approach to this piece was the go-round with the four-legged Mazerolle sawhorse, detailed on this blog over 50 posts a few years back.


  3. Wow that does look pretty intense! Have you considered not mortising all the way through the post? Maybe only go about half way through, it might make it easier to assemble if the tenons were a little shorter. You might also consider beveling back the tenon face of the uphill or non bearing side face of the tenon so it is diminishing in thickness as it gets longer. This would allow you a little play as you engage all the pieces as every little bit helps. I would love to see a picture when you finish this.

    1. Tait,

      thanks for your comments and suggestions. Blind compound mortises were already mentioned in the post, and I would rather avoid them.

      As for the beveling of the 'non-bearing' face of the tenon to allow more wiggle room, that's just not good joinery practice as far as I'm concerned, and would result in a loose interior fit for the tenon. The barbes on the pieces make these connections problematic to peg, and I would generally be thinking to wedge them on the exit face of the leg. So the diminished tenon idea wouldn't work so well with that I don't think.

      I'm not looking for just any old way to do it, I'm looking for a structurally sound and aesthetically clean connection between parts. I also realize this may not be possible given the parameters of the piece, but I'll swim upstream as far as I can before that surrender.


    2. When I talk about a diminishing tenon/mortise I mean one that is systematically laid out to the same tolerance as any square cut mortise not just randomly tapered. For example offset nominalº over the depth of the tenon and the mortise. The wiggle room comes into play and is advantageous only during assembly as the tenon is entering the mortise as it does not have to travel directly into the mortise. This is a structurally sound and common way to avoid what you call earlier "impossible assemblies"
      Assuming it is a compression joint the the strength of the joint would be the same.
      I do see there may be a pegging issue perhaps because of the "barbs". but it could be done.
      Can't wait to check out your progress!
      It does look like a great project.

      What are the dimensions of these pieces?

    3. Tait,

      ah, I see what you mean - a tapered tenon in a tapered mortise. something to think about. Objections that come to mind immediately are:

      -such tapered mortises and tenons are never shown in any of the French drawings (not that the lack of that method is a huge disqualification mind you)
      -it means taking an already difficult mortise, non-square and compound angled, passing obliquely through a member, and adding a tapering to it. Not a tasty proposition particularly

      As for dimensions - please take a look at the next post - #11. There, I mention the brace dimension of 1.125", and the top beam dimension of 1.5". I'm making this as a coffee table, as mentioned in post 9 in this series, so the parts are furniture size.


  4. Also you could level the tenon off from the point it vanishes into the post. This would make the mortise 90º to the post which is much easier to clean out than a negative angled mortise. Also all that extra tenon just makes it hard to assemble and doesn't add any strength.

    1. Tait,

      thanks again. I am looking to make some of the mortises 90˚ to the connecting member, and have already done this (as of today's drawing work) with the connections between the three top beams, which can be pegged. It also makes sense i think between the post tenons and the beams.

      It's a good idea and I'll see about incorporating it where it makes sense, which, it seems to me, isn't in every location. It might help out in enough places though to make this project possible without metal fasteners.



All comments are moderated:

Anonymous comments will be deleted

Spam will be deleted

Comments containing links unrelated to blog content will be deleted