Very subtle eyebrow roof . . . suggest a way, please.

[GeneDavis]

I am going to be doing con docs for something that has been sketched using these sources as inspiriation.  I've watch these get framed, and all with straight rafters riding up over very gently arched wall plates.  Two layers of sheathing, thin t&g, get used atop.  Lotsa staples.

 

Curved roofs with these quite large radii are troublesome for me.  How would you do it?

[GeneDavis]

Thanks, Eric.  See my Suggestion, posted today.  These warped-roof versions of which I speak have no bell-shaped valleys around the eyebrow.  All the rafters have their top ends up either at a common "ridge" elevation, or are tied to hips.  Each rafter is pitched differently, unlike what we get using the Chief curved roof tools and doing the joins.

[robdyck]

Chief also won't let you warp a face or a face of a solid.

You may need to model this using terrain. I've done something similar before. You just need to know all the elevations and geometry before you start. 

The elevations along the hips are simple. So, plot out the elevations of the fascia using cad tools before you start modeling. Then draw a terrain the correct shape of the roof plane. Then add consistently spaced elevation points. You may need to figure out the slope percentage along each rafter so you can replicate a series of points at the correct slope percentage along each rafter.

It's fairly simple if you approach it that way.

[Alaskan_Son]

  

5 hours ago, GeneDavis said:

 

Using triangular roof sections just like you've shown in this picture is how I've done complex roofs like this in the past.  Its tedious.  You have to map out the fascia shape and then use that information to set your baseline angles and roof pitches.  Its doable but time consuming for certain. 

 

You can also cheat it with just a single roof plane (curved, pitched, and baseline tilted) but its not as accurate..

 

[VHampton]

That's a really creative solution. Probably the best yet. 

 

...another option could be to build the roof.

 

Then draw a molding poly-line in cross section with a slightly lower pitch - basically to create a rafter "molding". 

 

Then - in 2d view - a head on elevation allows one to draw the fascia (in the soft arch).

 

This gets converted to the "rafter molding" which will have shingles. It'll create the slight hump on the actual roof surface. 

 

A 2d molding needs to be created (and lowered by 1") so that it will serve as the fascia. 

 

It has potential to appear as the eyebrow effect per the image. Not too time consumptive at all. 

 

 

[GeneDavis]

That's an interesting way of doing the solid that is the hump.

 

Could the curve be the distribution path for the rafter tails?  All this guy's work has barenaked soffits with rafters exposed.

[VHampton]

Hi Gene... I would venture to say that the answer is yes.

 

When the initial fascia line is drawn in 2D elevation view (and curved to your liking)

...both the roof and the soffit will curve (when converted to a 3d molding) - just as per the real world images. 

 

The rafters tails on the underside of the roof "hump" can be made with 3d solids, and placed manually.  

 

My observation is this... There will be (3) moldings required. 

(1) for the initial fascia

(1) for the roof skin - a replica of the fascia - set one inch higher 

(1) for the soffit - another replica of the fascia - set one inch lower and probably one inch inward

 

The 3d solid rafters can be manipulated to follow the radius of the curve.  

 

Note: Radial moldings required some finesse so that they're a bit more fluid in appearance. The optimal way to make that curve look spot on, will be to make sure that the initial fascia has many facets - otherwise it'll look sloppy and choppy like in my quick example.

 

Hope this is a help.