Re: width of raised panel rails and styles
Alexander Grigoriev
>If you analyse the failure patterns in FWW test, it's not much a matter of glue surface.
For a butt joint it's a matter of low strength for pull from long grain surface.
For strongest joint, its failure mode would be wood shear. Lap and bridle joints had that.
In that test, mortise/tenon type joints failed because the tenon was too thin (1/3 of the stock). Floating tenon failed also because it was too narrow. Were it positioned closer to the outer side, it would experience less force and the joint would stand more.
The only joint where the glue failed was miter joint. Because it was end grain to end grain, it seemed to be starved of glue. I'm not sure if it was simply cut on a saw, or it was also shot with a plane after that. I would expect shot surface hold better, provided also with enough glue. Apply glue to both sides, let it stand a bit; only then put them together.
I would expect the strongest joint would be a miter with full-width tapered spline in a tapered groove (see picture). Groove bottom should be rounded. Spline thickness should go from 1/3 of stock on the bottom to 1/2 of stock on the top. Though it might still have fatigue issues with twisting force. For that reason, non-tapered spline might be better in the long run.
