Roof Face w/ cross slope

[Benson Shaw]

Some recent posts about roof face with cross slope made me curious. It is a geometry problem.

 

Here is one solution in a simple rectangular context.

Process starts at a control edge with known High/Low points - edge of adjoining shed roof, top of a sloped wall, etc, plus the proposed high point.

 

Things needed for parameters of the Roof Face command, all determined via vwx geometry:

1.  Locate the roof face axis = a level line across the imagined new roof face at height of the control edge. (h=4 in example). Analysis tool helps this.

2. Locate the the line perpendicular to axis from control edge low point, and . . .

3. Find the angle along that line from control edge low point up to the axis height.

 

Video note -

A 3d poly was created to represent plane of the new roof and find the axis. 3d poly is a perimeter. Analysis tool needs surfaces.

3d Poly was converted to Mesh. NURBS surface would also work. Analysis tool happy.

 

Wish/Enhancement - Roof Face command adds a 3d 3 point mode, sort of a hinge:

           Select the roof face footprint. Click the control edge low. Click the control edge high. Click the roof face high point. The command runs.

The current modes, including 2nd Click Height, all employ axis parallel to ground plane (or maybe working plane - too confusing!).

 

-B

[line-weight]

I think there are some unnecessary steps after you get to this point:

 

 

Because you can specify the roof pitch by "rise over run" - you don't need to measure the pitch angle.

 

At this stage you already know the length of that line (the run), and you know what the rise needs to be along its length (4.000 in the example).

 

 

 

[line-weight]

Also...regarding the "taller fascia" on the steeper sloped portion - this actually highlights a real-world construction detail question, because the fascia is not really taller. Measured perpendicular to its bottom edge it's the same height because it = the roof face thickness. The apparent discontinuity in height where the two fascia portions meet is because of the way the junction between the two roof faces is modelled (the join is projected vertically up from where the two planes meet).

 

When you use the connect/combine tool to join roof faces you're given the option of how this join is made - vertically or mitred:

 

 

This requires a bit of careful thought if you are modelling something in any detail (aside from the fact that the connect/combine tool is rather temperamental) because doing vertical joins is kind of easier to keep track of but doesn't reflect the reality of what's going to happen in most real world multi-layer roof buildups. Essentially you need to decide what's going to be the "reference plane" for the roof (I'd say you should have in mind what makes sense for the person building it - so it might be the top-of-joist plane) and then let the geometry work out from there. In many cases the actual "valley" in the roof surface will not be vertically above the join between the roof reference planes. And it can all get rather tricky if you want to set things out so that the valley ends up in a particular place...

Edited January 19 by line-weight

[Benson Shaw]

@line-weight Thanks for comments! I was just experimenting with the geometry and found one way.  Rise/run vs angle is about same effort, but extruding the perpendicular DOES provide both rise and run from OIP, which can be convenient.  "Facia", sorry, I should have said abutting face. That was my point, anyway.  I didn't take it as far as the construction detail.  As you say, the mitre join is way to go.  Roof joinery has such interesting challenges, design as well as build.

 

-B