Broken back, may also be an indicator of how the builder feels at times.

Descending ridge makes it all come together, I agree. How about splayed, descending ridges.

The pictures help with the definitions. The curb has a fascia. Now I have a new detail to look for. Tim

Chris, Thanks for all of this info and pictures. As far as the irimoya. Does the main slope (if the adjacent is the gable end) dictate. For instance would you use an 8/12 on the main, locate the bottom of the gable according to the 5ths or 7ths. and then locate your hips in plan thereby giving you the pitch of the roof below the gable?

I also looked for some proportioning guidelines (western classical). found this.

"If vaulted the heights of the vaults in rooms that are square is a third part more than the breadth of the room. But in those which are longer than they are broad, it will be necessary from the length and the breadth to seek for the height, that they may bear a proportion to each other. This height will be found adding the breadth to the length, and dividing the whole into two equal parts, because one of those halves will be the height of the vault"

Roofs

"Therefore the breadth of the place to be roofed, must be divided into nine parts, and two given to the height of the ridge".

Vitruvius

According to this: 24' span (breadth)/9 = 2' 8"
2'8" (2) = 5'4"
5'4 rise, 12' run = pitch of 5 11/32 or 23.69

never heard of that one!

Cheers, mo

Well, the Mansard seems to stem from a type of roof popularized in an urban context. I have no idea about French barns however, and whether they employ a Mansard roof.

The pictures of the thatched Dutch barn you note - I wouldn't call that a Mansard roof at all. That said, I came across an British roofing website yesterday that did call a roof with a slacker lower pitch (than upper) as 'Mansard':

http://images.google.com/imgres?imgurl=http://www.k-roofing.co.uk/images/mansard_roof_slate_council.jpg&imgrefurl=http://www.k-roofing.co.uk/contractor_examples/slate_roof_refurbishment_council.htm&h=667&w=1000&sz=202&hl=en&start=117&usg=__sbW_l_ZOp13NaBi3z6dAKTwEYyw=&tbnid=Vr0-Uz5ZA35UyM:&tbnh=99&tbnw=149&prev=/images%3Fq%3Dmansard%26start%3D105%26gbv%3D2%26ndsp%3D21%26hl%3Den%26sa%3DN


I think that site makes a mis-interpretation of the definition that a Mansard has two roof pitches. This type of roof is really a variant of the gable roof. In the case of the barn, slackening the pitch at the lower sides of the roof allows the space in the aisle below to be more readily used. Often, this shape of roof develops by adding a slacker pitched shed-roof extension to an existing gable roof, though I can't say if that was the course of development in the case of the two roofs discussed above. In the case of Chinese roofs, the affixing of a slacker pitched flanking roof to the core of the building, to give a more pleasant walking space below was common. The tiles for the roof were bedded in thick clay, and it was simpler to feather the clay out along the fold of the roof, which resulted in a roof shape that was concave down the fall line.

As for convexity/concavity in Mansard roofs, that is fairly common too, especially with the steeper lower pitched roof, which is often made concave as a means of gaining an eave. There's the convex example linked to in your earlier post (http://rktect.blogspot.com/2006/02/howards-blog-of-fame-mansard-roof.html), and if one makes the lower pitch convex and the upper concave, and smooth the fold transition right out, you end up with a roof that looks ogee-shaped, or a 'Hershey Kiss' if you like. The French term for this shape of roof is "pavillion imperial" The French have some really cool whimsical roof forms!

Hi Mo,

without getting into the details for how the Japanese develop the irimoya roof, which is rather complicated, I can say that the way you went about it makes good sense, and can add a couple of points to consider.

The purpose of making the roof pitch at the gable end slacker than the pitch of the main roof is to allow the gable opening, and barge board, to grow a little larger. This is done to counteract the optical illusion of the barge board looking scanty/puny in case where gable end roof pitch is the same as the main roof. The difference in pitches between main and adjacent roofs should not be radically different however - the Japanese rule of thumb is to make the gable end pitch 0.5/10 slacker than the main, which in a base-12 system is 0.6/12 slacker. So, if your main pitch is to be 8/12, then you might want to make the adjacent 7.4/12. In fact, one way to go about the layout, in plan view, would be to locate the irregular hip line and project it up the roof, then determine in elevation the gable proportion and intersect that with the hip line. I hope that makes sense.

Another trick involves the gable end and the barge board - if it is plumb, then when seen from the ground it will actually appear, to the discerning viewer, to slope backwards slightly. The way to counteract this is to incline the barge board forward slightly (not by pushing the ridge out only to create a prow, but by treating the point at which the barge board comes down to the roof surface as a 'hinge' and tilting the whole barge board assembly forward). The amount it should lean forward is only slight - about 1~1.5" or so. Done properly, the barge board needs to be made slightly rhomboid in cross section so as to keep its upper surface flush to the roof material.

Thanks Chris --

I can't argue with your logic, and to be honest I have my own pet project and obsolete words I insist on using.

The blank stares and raised eyebrows are sometimes fun when you use one in conversation, so too are the conversations that follow, though as often eyes glaze over, such minutia boring most to tears.

Maybe Marina the pop culture philologist will spark a greater interest in such things, though I'm guessing probably not.

Hi, I don't think this is too far off topic. Ive been sitting in front of the computer, paper, and pencil awhile to figure this one out. A variation of the irimoya. This view is axonometric so the view is a little wonky.

Im going to finish this design and I need some help. I can post the progress if you all like.

Here in the picture. Not an easy task to frame. Any suggestions for the framing connections for this.

I was thinking posts at the walls ofcourse. Posts under the hip and gable connections somewhere. Midspan plates from there to the other hip to support lighter common rafters. And a ridge. girts in between the interior posts to hold the adjacent commons.



What connection possibilities can you all think of where the hip terminates into the main slope? Thanks.

now that i think about it a bent for the gable and a hip going from the corner to the principal rafter. must be another way.

I've done that in stick framing with hips to the ridge, which continues on through to the dutch gable. Then jacks from the extended ridge to the hip. I like your bent idea better.

I was looking at a booklet from Alpine Truss co the other night. They also call that a Boston Hip. They refer to a jerkinhead as a Tudor Hip.

In a Japanese roof, the rafters are all carried atop the purlins (and you can count the wall plate and ridge as 'purlins' in that regard). The hip rafter is sunk into the plates so that it's top surfaces (the hip rafter is always backed) are in plane with the top of the common rafters. The joint used to trench the hip into the purlins is a bit complicated, but essentially is a form of partially-lapped cog joint.

The hip rafter in an Irimoya should not continue up beyond the lower foot of the gable. If you go all the way up to the ridge, the hip section is wasted, since it only supports roof sheathing in one plane, and causes the interruption of the commons into jacks and cripples -a bunch of extra work for no gain.

I think a good solution, considering that you are talking about using a bent to support the gable end, would be to use continuous plates/purlins atop the principal rafters of the bents, and park the hips and commons atop the purlins. The 'brick bond' approach to framing yields a more resilient structure than using tie girts between bents, IMO.

An alternative would be to use a truss roof system instead of bents. The Japanese also frame trusses in traditional construction, generally of the King Piece type, and there are examples of trussed Irimoya roofs.