One of the funny things about drawing in 3D is that while it can often help with visualization and problem solving, sometimes you really need to be able to set things up in 2D first. Case in point is the curved eave layout on this pentagonal gazebo roof I am doing as a design exercise. With hip roofs, the eave projection and rafter spacing are directly tied, and in Japanese carpentry practice one of the crucial aspects to getting a hip roof right revolves around getting the common rafter spacing sorted just perfectly. And with a eave that curves up, and a hip rafter curving up, the length of the eave changes from what it would be if the eave edge were straight.
It was in trying to move my drawing along a bit by adding in the soffit paneling that I discovered a problem with my 3d work. I had drawn the curved eave as a 3D construction in the first go-round, and the defects that arose because of this took a while to manifest. It turned out that I had so make a slight adjustment to the eave projection to correct the problem. So, it was back to the drawing board, and this time starting in 2D:
The projection traces, which are numerous, have been stripped away from the drawing to show it a bit more clearly.
With the 2D plans of the pieces figured out, I was able to construct 3D pieces from there, now including the soffit paneling and lattice (komai). Here’s how that looks at this stage:
A view from underneath – at this point the material coloring is temporary and I will likely be making some changes before long:
There will be another soffit/ceiling, fitted between the wall plate and the outrigger plate, and I haven’t started to draw that yet.
On top, I have roughed out the hanegi, the bowed cantilevers which support the eave:
So, it’s moving along. At this point the hanegi are perhaps a little too thick, but their shape is about right:
Next up are the fulcrum beams for the hanegi, plus other framing details which associate, and then I will start on the lantern support framing and work out the actual roof profile. At this point I am thinking the roof surface will be moderately campaniform (bell-shaped, ogee-shaped).
Thanks for coming by the Carpentry Way.