Friday, September 10, 2010

This One Rings a Bell (9)

Hello again. Post nine in this design build of a Japanese belltower. Nice to see you again.

I've been drawing 10~12 hours a day for the past 10 days or so, and much progress has been made. Trouble is, all that progress is not too apparent at this point as it has been of the 'two steps forward and one back' variety.

First the two steps forward: I managed to complete the drawing for the gable end verge board (hafu) and fascia, which includes the dreaded fukiji board. Here's how that came out:

Another view:

At this point you can see a gap between the peak of the gable boards and the top of the roof surface. The top edge of the roof will soon be transitioned down to the boards using what is termed in Japanese a minoko. A 'drooping verge' is one English translation for that term. I'll be talking about those parts in more detail soon enough, however the next matter I want to relate concerns that step back I mentioned.

After the roof form started to come together more, and I could get a better sense of how it was going to look, I came to see that the fukiji on the upper edge of the eave build up was a little smaller than I would have liked or is typical of other examples I have looked at. Since everything on the upper roof surface terminates on that board edge, there wasn't much point in forging ahead until I had that fukiji the way I wanted it.

However, readdressing the fukiji board issue brought me back around again to face the fact that I had not previously been able to figure out the one Japanese drawing I have which shows how to lay out an irregular hip atop a regular hip with fukiji. So I was back to that head scratching. And it was a really onerous nut to crack, let me tell you, and there was one night with me up at 3:00 a.m. trying to solve some of the difficulties, and another night with no sleep whatsoever. There were moments where I thought I had it all figured out only to discover 10 hours of drawing later, to my chagrin, that I had surmised wrong. That must be it...that wasn't it. And repeat.

In short, I spent a good long while submerged in Lake Fukiji, tromping along in the mud, feeling the weight of the water pressing down upon me, getting desperate for air. I guess it beats hanging out on a cliff, clinging by my fingertips, but not by much. It was late in the evening a couple of days back that I finally resurfaced from that lake. It was a very long step back through the mud on the bottom with those cement boots I had fashioned for myself, but I eventually made it to shore.

Through this saga, I learned much about that fukiji. Who would have thought there was so much to know about a little board which curves in and up like that? A board which gets a couple of page's mention in only one of my layout texts. So innocuous, yet such a challenge.

One of the issues I struggled with, though it took me a while to even recognize that it was a problem at all, was the manner in which I had bent the fukiji board inward on the drawing (and I had tried several methods). Here's how the board looked when I noticed the problem:

While the board may look fine in all respects, when you zoom in to look at the hip, which has a cutting panel fitted at its centerline in the drawing, you will notice a deviation in angle between the fukiji board and the geometrically correct angle (the dotted line on the panel) one should find represented on the hip:

The upper surface of the fukiji, while at the correct angle at the origin point in the middle of the eave length, has twisted slightly inward by the time it meets the plane of the hip centerline. The little twist inward results in the meeting point for the two fukiji, at their top outside corner, to be inward somewhat from where it should. And that deviation wreaks havoc when trying to get the upper roof surface to fit the fukiji. Things go point to point after all, even if the line of travel is curved.

So, I came to realize that although I thought I had the fukiji sorted out, in reality I had not quite got it right. So, I readdressed the drawing process for that part, and in the process bumped the height of the board up some, and projected it out further from the uragō piece below it. Here's a look at the corrected piece:

Zooming into the hip plane, the fukiji now terminates at the correct angle:

Here's a shot comparing the new fukiji (at left) with the old (at right):

The new fukiji conveys much more presence and mass, and protects the lower roof parts from the weather even better.

And from below, here's a picture which shows the difference in reveal between the new and old fukiji:

With the fukiji sorted out properly, I could revisit that ultra-complex drawing which I had spent four days on previously, and see if I could get it to work. This time it took me, um, five days!

Now, when dealing with an irregular roof, the problem is a more complex one, but even more so when the rafter is concave along its length and must also rise as each successive one is positioned along the eave edge moving towards the hip. But when you place this irregular construction problem atop a regular plan curved eave build up, with that aforementioned curved fukiji which itself has a face slope transitional between the eave build up below and the roof surface above, well, ya end up with some thorny issues. One problem which cropped up - again, something I hadn't noticed previously - was that of transferring the curve on the high pitch side common rafter over to the low pitch common side. I kept getting a slight kink in the line of the curve on the lower pitch side, just as it terminated on the fukiji:

I tried correcting this six ways from Sunday, but in each case the slight kink was there.

Finally, the eureka! moment came about 10:30 one night, as I realized how to solve that problem. It all boiled down to the point at which the center lines for the 45˚ hip below and the irregular hip above intersected. Based on the drawing I had to work with, this point appeared to be at the very tip of the irregular hip, however it turns out to be slightly away from that point. The drawing I have in the book, due to its small size, has no way of indicating such important details. This is in fact a common problem to any book on layout - the Mazerolle book is an exception in that it is of a much larger format with fold-out pages. It's really the only way with certain types of drawing to show the detail adequately on the printed page.

The problem now solved, the shape of that lower pitch curved rafter could now be established:

The hips at last have come out and here they are, a little over-long at this stage and not quite shaped right at their lower end, parked on the roof:

From above, the deviation between the upper hip and the regular lower hip is very clear:

This was a really insane drawing session, and unlike the nightmare of the Mazerolle sawhorse drawing from a few months back, in this case with a client waiting on me I can't simply put the drawing aside for a few months and let things percolate slowly, returning fresh later on to tackle it again. Not an option - I had to keep hammering away at it until something broke. Fortunately I still have my wits about me, and the dragon has been slayed for the time being. Let me know if it comes across as otherwise. At this point, my confidence is indeed a little on the tentative side, but I'm feeling good overall.

Finally I can start taking steps again and anticipate posting a little more regularly. Thanks for your patience and visit to the Carpentry Way. --> on to post 10


  1. Chris

    I can see the problem in your drawings, but I don't really understand the problem, so I'm miles from understanding the solution. I just hope the time spent will be richly rewarded in time saved later in the construction.

    On a more mundane note, how do the irregular hips transfer the weight of the upper roof to the regular hips?

    Keep on keeping on.


  2. Tom,

    sorry I have overlooked your comment for so long - my bad. The client was richly rewarded in a way when he left with the drawings to find someone else to build. I had charged but a small retainer for the drawing work in the understanding that I would do the building as well, and yes, I guess I got to learn a good lesson there.

    As for your question, the upper hips rest on purlins and upon the last foot or so of the lower hip itself. The purlins rest upon the lower hip, so the load from the upper hip to the lower is transferred through those intermediates and by direct contact. Most of the roof load is carried by the cantilever beams.



All comments are moderated, so if you're planning to spam this, know now that your clicking and pasting is in vain. I do read the comments before posting, so your mission is doomed from the outset. All this time and effort trying to put your inane spam onto blogs -- is this how you want to spend your time on earth?

Please do me the courtesy of appending your name to your comment, even if posting under the 'anonymous' option. No name = deleted.

Comments NOT accepted include:

-those containing links unrelated to blog content
-spam of any kind, or ham for that matter
-did I mention that attempted spam postings will be non-starters?