This One Rings a Bell (8)

Oh yes, I was writing a blog – now it all comes back to me. Where’d all those days go? No, in case you are wondering, I did not win an all-expenses paid holiday to some island in the Caribbean. I have however moved house, joy of joys, and that cut three days or so off my drawing time. It also gave me some wicked asthma, which is how my body has decided to react to the Ragweed (Goldenrod) currently in maximum pollen release mode. I’m recovering well enough now that I’m staying (for a little while) down in Connecticut,  where such noxious pests are not so widespread. There are other pests, like the drivers around these parts, but I’ll save that diatribe for another occasion.

So, on the days where I haven’t been preoccupied with other matters, which is most of the past 10 days, I have been drawing full time. Progress has been won at the end, but an epic struggle was endured with things looking dire at certain junctures. I’ll spare you most of the details.

I’ve been working exclusively on the upper roof, focusing on getting the common and hip rafters established before I begin working on the other structural members. One of the advantages of the Japanese double roof system is that it allows one greater latitude in where the supporting elements can be placed. All that should become clear soon enough in the days ahead, but today’s post looks at those pesky common rafters and those irksome hip rafters.

These were the toughest rafters I’ve ever laid out, more onerous than the fanning rafters in the exposed eave. The intense difficulty is due to the fact that the upper roof’s shape is not only irregular-plan but also of double curvature – and the whole works is perched atop a curvilinear base in the fukiji and other perimeter fascia build up layers.

Let’s see, trying to draw something curved in two directions and placed on top of a curved plan… Result? Near insanity, let me tell you (not that I have far to go to reach that condition!).

It was a tough nut to crack, and the process was hindered considerably by the fact that SketchUp is rather poor at dealing with curvilinear objects in general. Several parts were re-drawn multiple times until I could obtain a satisfactory result.

It all began, early one morning, with the seemingly innocuous task of establishing the curvature of the high pitched (7.5/10) common rafter:

Then the other parts, namely the low pitch (5.5/10) common and the hip rafter profile, are developed from that rafter’s profile. Sounds easy enough, what could go wrong?

Initially I followed a drawing shown in one of my layout books- with the aid of a magnifying glass – and here’s how that drawing looked when it was complete:

That little sketch took me 4 full days to complete. If I pull the traces out of the development, you can see more clearly the principal players:

The high pitch common is on top, the low pitch common to the right, and the two faces of the curved irregular hip are splayed out on each side of the irregular hip’s plan, in green.

Another view:

Clicking on the above images will make them more legible, if you’re feeling brave enough to look what may appear as little more than a mass of tangled wiring.

Here’s a zoom into the tip of the irregular hip:

The brightly colored bits are cross-sectional views of the eave build-up components, both at the face of the hip and at the centerline:

Did I mention that I spent 4 days on that friggin’ drawing? I was about as nervous at the end of it all when it came time to ‘raise’ the 3-d parts off the 2D developments as I would be after having cut a stick of wood when a bit uncertain in the layout.

And you know what? The layout didn’t quite work! I am absolutely sure I followed the book very closely, and yet things just didn’t quite come out right at the hip – nor did the high and low pitch curves correspond properly. There was a bit of teeth gnashing at that point.When there isn’t much hair left to pull, you grind your teeth – that’s how it was explained to me.

So, back to the drawing board. After several mock ups and re-draws I was finally able to iron out all the wrinkles and produce a high pitch side and low pitch side that had perfect correspondence with each others curvature where they met at the irregular hip. Here’s a view of the underside of the mock up:

The light blue piece in the middle represents the plane of the centerline of the irregular hip rafter. A click on that image will make it all come clear I hope.

With the rafters sorted out, I could then commence building up the roof underlayers atop the rafters, which will comprise 2 layers of 3/8″ ply and some peel and stick and ice & water shield, yada, yada:

The hip rafter is not developed yet. Well, it was, incorrectly, at least three times so far, but that was then and this is now. Round 11 and I’m a little punch drunk. It should come out fine this time, knock on wood.

I placed the roof parts up in the main sketch to see how things were shaping up. Wanna see? Here’s a look into the gable end with the shallower pitch:

Now, the gable end rafters will not be running up into the middle of the roof and meeting each other, as it appears in the picture – I’ve left them long for the time being as I have yet to sort out the details in the gable opening.

Here’s a view looking at the face of the steeply pitched side:

A corner perspective:

One of the consequences of having the eave curve begin right in the middle of the eave edge is that the ridge and purlins are curved as well. It’s one option among many, and the one I went with this time.

Lots to do yet, but the worst appears to be over. Ha-ha, where have I heard that before!

I can complete the hips tomorrow and sort out the gable detailing including minoko. Once the overall form is looking right, then I can locate the fanning cantilevers and the rest of the supporting cast, and as none of that work is double curvilinear, it might almost go quickly. Well, I’ll hold off on the optimistic time projections for the time being.

Thanks for coming by the Carpentry Way today. Comments most welcome. –> on to post 9

13 thoughts on “This One Rings a Bell (8)

  1. Chris,

    its just gorgeous! You are a brave man to walk into a situation with a curved irregular hip. Most would run from it. But… Is plywood a good choice for a doubly curved roof? I would be afraid of getting a 'bubble' in the center of a sheet and not being able to nail it down properly. I had a singly curved roof long ago and talked the architect out of ply and into boards which worked splendidly. Anyway, its gorgeous.


  2. Tom,

    sorry for the late reply. I think that your idea is a good one and that with a concave roof surface, getting the sheets to lay flat might well be a hassle. It wasn't a problem with a convex roof surface, but I can see why it would be with a concave one. So, I'll look to use boards there.

    Dale, glad you are enjoying the thread so far and thanks for the compliments on how it's looking at this point. Changes are afoot however…


  3. Hi there, i just came across this site because im searchinglike this, anyway what program are you using? thanks…
    i like the way you did it.. its a work of art!

  4. Hi Chris,

    You have peeked my curiosity here on a matter: irregular hip gables.

    Do you know how prevalent the Japanese use irregular hip gables over the regular hip gables on larger structures such as Sanmon? For example, some really large gates, such as Chion-in's Sanmon, seems to have an irregular hip gabled roof (I might be wrong, but it darn looks like it).

    Lastly, what is the Japanese term for the irregular hip gabled roof? I know that the regulars are called irimoya but can't seem to find anything related to the irregulars. 🙂


  5. Michael,

    glad to have piqued your curiosity.

    The irregular hip version is, on larger and more expensive structures, rather more the norm than the type with regular hips.

    An irimoya roof is called an irimoya roof regardless of whether the hips are regular or irregular. The term for irregular hipped roof construction however is fure-zumi yane.


  6. Thank you Chris. Does it matter to the Japanese carpenter, the angles used on the irregular hip? Assuming that a traditional regular hip gable has a ridge angled on 45 degrees?


  7. Michael,

    are you referring to the plan angle of the hip?

    The irregular hip angle is not determined simply by whim, but in interrelation to the proportions of the gable to the roof, whether the gable is convex or concave down the roof slope, etc. so, yes, the angle of the irregular hip, both in plan and in elevation is a very important design consideration.


  8. Yes, all of those curves and roof members should come into play to determine that angle. However, I am baffled even more by the “slope” of the roof so to speak on each side of the roof. By quick intuition, I suppose that the slope on the front side (usually the long side on a rectilinear building) will have a much steeper slope than the short side caused by the irregular hip? And with that, how do a tiled roof cope with this in relation with tile laying? Wouldn't the pendant tile on the long and short side have differing angles? This is all so interesting as irregular roofs aren't really dominant around my area.

    Thanks once again, Michael

  9. Michael,

    what you end up with when the hip is placed at an irregular angle is that the roof surface in front of the gable is at a slacker pitch than the main roof surface. I can't speak so well to the tile issue, a this is outside my area of expertise.


  10. Chris,

    I have taken your recommendations and have bought my own copy of Mr. Togashi's gate pattern text. I liked it so much, I got his temple layout book too!

    One question that I can't seem to fit the puzzle together to is the 弛み (tarumi). In the layout text, he appends a fraction (such as 3/10 etc) to the concave curve (assuming we are working with a concave profile roof). However, explicit steps into how to “use” that number to generate a concave curve is still shrouded in mystery. I believe the method illustrated in the first picture of this post describes the method I'm trying to seek. Any light shed to this issue will be much appreciated, thanks.

    Keep up the good work,

  11. Sag in the body of the roof is usually a percentage of the length over which the roof is curved. A given length might be 'L', which the sag will be 4/100 of 'L' typically. The deepest point of the sag might be in the middle of the length, or it might be at some other point, depending upon how the length might be divided up into parts. There are various methods for determining the tarumi, however those books you obtained are pattern books so they don't show such techniques. Those books are about the proportions between parts and the slopes of the roofs, arrangements of structural members for aesthetics, etc. You're expected to already have a grasp on such things by the time you come to look at pattern books. The first picture of this post does give you some clues however….


Anything to add?