Sunday, November 22, 2009

Tréteau VII

Did I actually end the last post with the comment "the end of the drawing is nigh" or something like that? Silly me. Such optimism and naiveté.

This is the 7th post of a thread devoted to the construction of a 19th century French carpenter's sawhorse, of, uh, a somewhat complex design. This sawhorse is a vehicle for roof carpentry study, and that's why I'm tackling it.

I thought the legs would be relatively straightforward to lay out, however they have proved to be rather difficult - even brain-melting to deal with. It doesn't help that once again I find the original illustration in Mazerolle's book to have further errors. Now, I'm with the reader who suggested that the errors are likely due to the work of the engraver of the printing plates, however I'm not so sure that is the issue in all cases. It's one thing to connect a line, especially when it is clustered amongst many lines, to the wrong spot. It's quite another to depict a part as a mirror image (and an impossibility), as I mentioned several posts back in the case of the development of one of the long side braces.

Now I have found a similar 'mis-drawing' as I work on the leg development. This was a part that looked wrong months ago when I was working on it in 2D only:

This is a view of the narrow end of the sawhorse, showing the development of the left-side leg and one of the Saint André cross pieces. Note the orientation of the tenon on the top, in both views, and consider how the tenon is supposed to be cut so as to be in line with the top beam. Note the single mortise on the face of the upturned piece at left, which corresponds to the tenon of the narrow side Saint André's Cross member. The illustration omits the mortise for the other Saint André's Cross member which connects further up the leg. The undeveloped leg to the right of the pair shows the mortise for the long side brace illustrated lower down.

Anyway, the illustration seems to show an impossible tenon on top of the post. It shows the tenon rotated 90˚ out from where it should be. Hmm... initially I thought I was merely seeing things wrong and had left the issue aside while I drew other parts of the sawhorse.

Now that I am back to working the details of the leg layout, in round 17, punch-drunk as I am, this issue arose once again. I was partway through drawing the tenon, 2D into 3D, when I realized that the tenon orientation was simply wrong. NFG. I dragged an existing 3D rendered leg out of the horse and laid it down next to the one I was constructing, just to double-check:

The wood-colored leg at the left is the one pulled out of the 3D drawing of the horse, while the one in white is the one I'm developing from 2D. You could mentally superimpose the white leg right on top of the plan if that helps. Note the location of the mortise again lower down on the leg (for the long side brace tenon) and the relative orientation of the tenon on top. Like the mirror-image long brace I found earlier, this is a curious 'mistake' -- was the engraver a heavy drinker, or perhaps thought it would be humorous to mess with the illustrations?

I moved past that issue soon enough when I accepted that there was simply another 'mistake' in the text - the legs which are oriented to the axes of the plan are reasonably straightforward to draw, and the mortises not unduly traumatizing to lay out.

The same cannot be said of the legs with are turned (in French: pied faces aplomb) in relation to the x-y axes of the floor plan. Here's the drawing supplied in the book for developing the mortises on the rotated leg:

Right near where 'Fig. 2' is written there is a shaded area that is supposed to be the mortise for the long brace in the lower portion of the leg. The vast majority of the lines necessary to construct the view of the full mortise, entry and exit, are missing from the sketch - just the skeletal outlines are provided to give the hapless victim the general idea. Trouble is, the mortise depicted in the illustration is the wrong shape, so the lines lead you into a quagmire instead of the hoped-for session of clear sailing. The tenon width is fine, however the top and bottom walls do not angle so sharply as the text illustrates, and thank god as the mortising would be a right pain if they were that shape.

Here's what the mortises should actually look like:

The smaller half-size mortise is the exit hole, as the tenon diminishes by about half as it passes through the leg.

I just about pulled my hair out trying to figure that one out, as the drawing in the text is quite misleading and it doesn't become apparent until you have drawn for a while, getting more and more puzzled. The problem is that the wrong lines are being connected to form the upper and lower walls of the mortise.

Thank goodness for the 3D, because it allows me to compare the piece, as it should look, with the 2D development. Without the 3D, I would have gone off the text's example, drawing the piece in a similar fashion and started cutting, only to find that the mortise would have been entirely wrong and the leg ruined after hours of work. Hopefully I can avoid such pitfalls, and the 3D is a big part of that.

Still, as I've mentioned, the essential skill for the carpenter to develop is the 2D drawing skill, which can be field-applied on a clean flat surface with ruler, pencil, compass, and string. Here's a short series showing some of the steps, going a little further from the 2D, in developing those mortises in the above picture into the 3D mortise. First, I raise lines to the thickness of the post on the uphill mortise, and connect lines down from that mortise opening to the half-size mortise.:

A peek around the back:

I raise more lines from the foot of the leg at the baseline, and construct a short section of the post, now with the mortise contained:

Next step is to deal with the mortise for the Saint André's Cross member. By the time that's done, the plan is a mess 'o lines. I'll spare you the horrow show.

Another curiosity I discovered in the past few days is that the two legs which are oriented to the plan axes are not actually square in shape. They are slightly rectangular by about 1/64". At first I thought I'd found another drawing goof-up on my part, but after checking carefully I found I'd connected all the dots correctly. One leg was rectangular in one direction, and the other rectangular in the opposite direction. Hmm...

So, I started a new drawing to see, experimentally, what would happen if I controlled the dimensions of the posts so they would be exactly 2" x 2". Well, you can do that, but the cost is that a bunch of other parts will end up being odd sizes. For example, one Saint André's Cross piece pair will be 2" wide, while the unit on the opposite end of the horse will be 1.98" wide. Same thing happens with the long side braces - 2" for one pair, and 1.98" for the other pair. I did a few rotations with legs to see how the geometry changed and discovered that the issue lies with the fact that the slopes are irregular in this sawhorse. I'm not going to try and explain in detail, because it is a bit complicated, but it was news to me when I discovered it. It's not possible with this method, using square-section legs in differing rotations, to specify the sizes of all the parts - most of them, but not all. so, one keeps the sizes at a set point in the elevation views, and the result is slightly rectangular legs and the interior x-braces end up being odd sizes in both dimensions (mine are 1.16" x 1.79").

Very fortunately, I left the legs just slightly oversize enough - about 0.02" - from the jointing and planing earlier that I have just enough wood, by just a couple of thou, to make the rectangular legs I now realize I need. Lucky. If the legs had been scant, I would have had to re-draw the entire sawhorse, and planed a bunch more off the braces. No thanks.

So, after experimenting with holding the leg sizes to 2"x2", I found it caused more problems than it was worth and returned to the original, er, 17th, drawing, and kept with the slightly rectangular legs, as it allows all the brace pieces to be exactly the same width. Reflecting further, and considering the difficulty I foresee of the cutting of the mortises and tenons with the French construction, I have to say I vastly prefer the relative simplicity and elegance of the Japanese method of re-shaping the legs to bring them into plane with the prismatic splayed form. While the hassle is there initially with the re-shaping, once done, and done accurately, the rest of the parts are relatively un-troublesome to cut, and the mortises and tenons free of extreme angles and barbes.

In a related vein, I wonder why the French have come up with all these methods for dealing with a rotated hip rafter when the backing cuts are not all that much of a hassle to make? Sometimes they do use a backed hip rafter, from what I've seen in the various books, but not consistently and I'm curious why, given the extra time and hassle of the tenons with their barbes and the often angle mortises that result. It doesn't seem worth the trouble, so I am curious to know what advantage was perceived by the other method.

Perhaps for the next sawhorse, a year or two down the line, I will breed, so to speak, the Mazerolle tréteau with the backed leg technique I figured out for the irregular splay Japanese sawhorse I already have made. That would be a pretty cool cucumber. For now, I'll forge on, and finish up the drawing in the next day or two and then can move on to layout and cutting by mid- to late-week hopefully.

Thanks for your patience and for dropping by. --> Go to part VIII


  1. All I can say is you sure have a lot of tenacity to kieep at this. I don't think many people would be able to stick to this for this long in view of all the problems with the orginal drawings and with Sketch-Up but that's why "you're the man"

  2. Thanks Charlie. I can be pretty determined sometimes. I have started and stopped my study of this French sawhorse several times over the past year. Sometimes I need to not look at it for a while, which seems to allow some digestion to occur. It's funny how things can move from complete incomprehensibility to a point months later where only a few details need be sorted out. Sometimes perspiration trumps inspiration I guess.


  3. To the 'Anonymous' poster: thanks for your comment, but I didn't allow it to be posted as you didn't put your name to it. I welcome all comments, positive, negative or whatever, but ask that a person append their name before I will allow it to go up. Sorry if I haven't made that clear on a more regular basis here on the blog.



All comments are moderated:

Anonymous comments will be deleted

Spam will be deleted

Comments containing links unrelated to blog content will be deleted