Tuesday, March 30, 2010

Following Mazerolle: Lucarne Et Nolet Éventail (II)

Today has been full of ups and downs and ups. This morning I woke up to an e-mail from one of the purchasers of the Japanese Carpentry drawing essays. This gentleman had completed the hopper project, and the kōkogen slope data table, and was ready for the exam. Good work!! Now I have to prepare an exam of course, which will take a few hours, but I was psyched to see someone work their way through the 130-odd pages successfully. The pace has been set, the glove thrown down as it were.

I then set to work on the fan-tail dormer in the Mazerolle charpente text and was running into various difficulties. After a couple of hours and mounting frustration, I took a break and went down the street to buy a little 12/4 pattern grade mahogany (genuine), and run it through the jointer and planer. This is some of the material for the upcoming build of a Japanese free-standing screen, or tsuitate.

I was telling the guys down the road that I had just about reached the end of my rope here with the Mazerolle book and was ready to pack it in and put my attention to something else. Returning home, I had a bit to eat and sat down in front of the drawing again for one last look. Yep, it has some problems alright, just like all the rest of them.

I tried a little reverse engineering, and this time had a breakthrough where I was able to determine what exactly was screwed up in the original sketch. Freaky! In the previous post on this topic, I mentioned how the portion of the drawing which developed a section view of the ridgepole was erroneous - now I have figured out that the view produced is actually of the ridge in a plumb section. That view is not much use, mind you, for placing the piece directly on the drawing and connecting points, but it did clear up the location of some of other lines which had developed from it.

Those lines were also in error due to a portion of the drawing showing some other lines which were simply not where they should have been. Here is the problematic section of the drawing in the book, just to illustrate the nature of the difficulty:

Notice the line running along from the apparent section of the ridgepole (way over on the left) to point P' all to the way to the right? Then notice the line below it marked with an 'N', which means niveau, a fold line. Those lines are not parallel, but they should be. The line running from k to P' should be parallel to the niveau.

Here's a view a little closer in:

Notice the line from the lower surface of the ridge, marked "Dessous du faitage paralléle...(to line kP') - it says that the lower surface line runs parallel to the other line going from k to P'. Well, parallel it should be, but the line kP' is not going the right way anyhow, so neither is the lower line. Notice in the drawing the points marked 1, 2, 3, and 4 - these are used to transfer points to another view in the plan and since line kP' is not aligned correctly, and neither is the lower surface line, points 2 and 3 end up being in the wrong place -- not by much but enough to screw things up if you marked the wood directly by transfer on the other part of the plan. With a little reverse drawing from the completed 3D parts on top of the 2D plan, I was able to determine the correct arrangement. I still don't get everything on the drawing, but had enough to produce developed views of all the parts in the end.

Anyway, I survived it. Here's the completed fan-tail dormer, or Lucarne Éventail:

Standing on the floor looking out:

If you stood inside the dormer and could see the exposed wood (not how it would be done I guess in most cases), this is how that would look:

If you could fly over top, here's the view:

It might be hard to tell, but the noulet, or little valleys are pronouncedly trapezoidal in section. They are that shape as it makes for simple bevel cuts on the rafters that attach to them (which in turn makes for a stronger abutment and allows conveniently for a mortise and tenon connection if desired).

Finally, here's how the developed drawing looked:

You can see how I brought the hip rafter and the noulet pieces down to the floor to check and work on correcting the drawing. Here's a closer view showing some of the connections for the lower end of the hip rafter:

The tan colored shadow is the footprint of the rafter on the plate. That hip portion, BTW, was another part of the original drawing that was in error, as it showed the hip rafter foot without the curious little asymmetric birdsmouth (like a high heel in a way) on the bottom. The illustration in the book showed it with a full-width birdsmouth, which is impossible given the position of the hip rafter on the plate.

And last, here's a look at the noulet layout zone:

One valley piece is on the plan in the 'actual length' position, another is kicked up at the actual orientation and slope, and the third, over to the right side, is a development from the first, rotated over to another side. If you look carefully at the kicked up valley piece, you will see it is not square or rectangular in section, as noted earlier: the lower surface is much slimmer than the upper surface.

That's all for today folks, and thanks for dropping by today. If you feeling a little overwhelmed by all the drawing, don't worry - me too! I'll be back to cuttin' wood up soon enough.


  1. A thought and a question.
    Thought - it seems odd that some one of the authors calibre should have drawn with mistakes - so I wonder if the errors came in when the engraved plates for printing were made by another.

    I have only used SketchUp a little, preferring another package that I have suffered the learning curve on. In one of your posts you mention 'lifting' the 3D image from the plan - is this a Sketch~Up thing or a phrase of your own to cover the development of the drawing into 3D.

  2. Guy,

    I also find it quite odd that there are so many mistakes in the text. I can accept that some errors could have come about via the engraver of the plates, but this only explains a certain number of the errors I have found. It does not explain all of them away.

    There are several types of errors in these drawings I have found so far:

    1) places where a projection line connects to the wrong place, usually close by where it should have been. These I can see as engraver's mistakes easily.

    2) places where a given developed part is a mirror image of what it should be, or has a component within it rotated incorrectly. If the engraver's plates had to be made as mirror-images of the original for the purpose of printing, I could see how such problems could arise.

    3) constructions which are in fact impossible. Examples are some of the through-tenons in the sawhorse piece, and the molding for the Lucarne Biaise which climbs the gable. These I feel cannot be engraver's mistakes.

    4) examples where the geometric method shown does not produce the parts accurately. An example of this is found in the above posting, with the section to determine the ridgepole cross-section, and in the previous two dormer examples where the method for finding the noulet is simply not working. I find it hard to believe that these mitakes could have happened by an engraver's efforts.

    As for your question about SketchUp:

    SketchUp is primarily a 3D drawing approach. I use it a little unconventionally in that I draw things in within it in first in 2D plan and (sometimes) elevation views, and then develop the 3D parts and structures from there. In a sense, I am using the drawing as if I am doing full-scale lofting - the marks on the floor are transferred to parts placed directly over them.

    In some cases where I have found the 2D development that Mazerolle shows does not make sense, i often have enough information to be able to construct the 3D part anyhow. Then what I do it copy the 3D part out of the drawing and place it back down on the floor, in the correct location and orientation, so that I might be able to reverse back from there to uncover the 2D drawing. That's what worked for me in the above post.

    Without the 3D, given the mistakes and dead ends that result from the examples in Mazerolle's book, I would have quit long ago. The 3D helps me fill in the missing information and confirm the working of the developed drawings. That said, I am taking a break from the Mazerolle book for the next little while.



All comments are moderated:

Anonymous comments will be deleted

Spam will be deleted

Comments containing links unrelated to blog content will be deleted