A comprehensive educational community devoted to trim carpentry, finish carpentry and architectural millwork. Hosted by nationally recognized author and finish carpentry specialist Gary M. Katz.
  Radius Trim: What's Right and What's Wrong  
Kurt's Question: How do you use the Crafstman Style door/window head on a radius opening?
Here's an example of the style in a segmented arch. Obviously something about it isn't quite right.
Classical heads are difficult above a full-circle or circle-top arch, but not impossible because there's an easy point of 'square' termination. In the photo left, the head terminates 'flat' onto the columns. In the photo right, keystones are misused as rossettes. A keystone, like a voussoir, owes it's definition to its use: supported by two equal and flanking forces.
Here's a closer look. I've never seen a horizontal keystone. That's another reason these ornaments defy the definition.
Here is an example that's somewhat like Brent's solution to the radius head problem when using the Craftsman Style. I think this solution works very well. It has to be the best one, in my opinion, because classical rules and proportions are maintained.
In fact, I kind of like the simple yet elegant way this carpenter has interpreted classically inspired designs. These openings were designed and built by Paul Allen, one of the contributors to the JLC Finish Carp. Forum (Paul also supplied the pics for the two photos above and the one below). I like the use of the keystone as a tablet, and the way the crown breaks forward around it. I think the segmented arch enhances this opening, rather than diminishes it, and only because the trim is correct.
Whereas, in this exaggerated use of keystones, it's even easier to see that a segmented radius arch can pose real problems. Here the keystones are vertical, but because they're not positioned between flanking and equal forces, they make no more sense to the eye then the previous ones which were horizontal.
The problem doesn't occur with Palladian arches, ones that have horizontal flanking head jambs, because the casing can be mitered around the arch-to-horizontal-head transition.
But on segmented radius heads, installing rosette poses a real problem. On a normal horizontal door head, the rossette is positioned with the bottom edge slightly above (1/4 in. reveal) the inside of the head jamb. Whereas, on a circle top window, the rossete must be posistioned with the top edge at the spring line of the radius (similar to the columns above and the circle-top window below). But on a segmented arch, there's no good choice. In the example above, the rossettes are places above the spring line, to match nearby flat-topped doors, which necessitated custom elongated rossettes.
In this example, the rosset is placed with the top beneath the spring line (as it would be on a circle top head), and a transition miter is made in the casing, another 'detail' that wouldn't occur in classical architecture.
It would seem to me that the best solution to a segmented arch, is to incorporate it beneath the horizontal door head (Brent's and Paul's solution), or use a three or four-centered arch or ellipse, as in the photo left, and the ones below.And THAT'S why carpenter's need to learn easier ways of using and building ellipital openings and trim.
  Supported by corporations who care about education in the construction industry.  
  Mastering FInish Carpentry DVDs