My goals in making and teaching chairmaking have more to do with making a great chair, having fun, and opening doors to accuracy and understanding. As I do this, I often try to isolate the variables at play and even if I abandon the idea, the thought process that I engage almost always deepens my understanding.
The first class in Melbourne was a continuous arm rocking chair, and once I realized what a different experience the troublesome timbers were giving the students, I set out to make a way to tame the rocker slot beast. In past classes, I've used the router to cut all four rocker slots in about 6 minutes and by the end of a class like this, students are bleary eyed and exhausted.
So I gave instructions on how to lay out the rockers and cut them by hand, and then we proceeded to use this set up to get 12 chairs to rock exactly the same, with nearly no balancing or fitting and no layout.
Here's the deal. Legs don't matter on a rocker. Yep, that's right, they are not the reference that counts. What counts is the relationship between the rockers and the seat. If the rockers are in line with each other, and have the correct position in relation to the seat, then the chair will rock beautifully. Of course, if the rockers are poorly positioned, the chair with tip to far one way or the other, so when setting rockers, the art of balancing them to each other and also to the seat at the same time is a challenge.
So, given my success in routing rockers in the past, I set out to make a jig that responded to the variables that count. By placing the chair upside down in the jig, I've already nailed one vital reference, the seat plane.
The next variable is the splay of the legs (and therefore rocker). The idea here is to split the difference between the front and rear leg splay and then to split it again from one side to the other. By coming to a single angle that the slots are cut and the rockers get beveled on the bottom, you can be assured that the curve remains consistent and balancing them will be possible. Imagine if the two rockers sit at different angles to the floor, they will actually be acting as different curves and become a nightmare to balance.
To knock out this variable, the jig that we used has a tilting board that is set to one angle for all of the legs. The only concern then, is that the legs register in the jig so that the slots are cut in the center of the leg at the full depth of the router cut, so there is no weaker side.
To achieve this, I cut a beveled edge on the bluish piece of ply that is attached below the surface where the router runs. When the chair is pushed up against the bevel edge of the plywood, the leg is in position to ensure that the deepest part of the cut is centered, regardless of the splay of the leg. It's all about relating the path of the router to the platform and then the leg to the path of the router.
The final variable is the height of the rocker slots as they relate to the seat. This is really where this process earns it's keep. By setting the heights of the two platforms in relation to the seat, the position of each rocker will be identical to all the others (the fact that they will perfectly align back to front is not even an issue with this thing!) The depth that each leg is cut might be different, but once again, it aint the legs that matter, it's the rockers relationship to the seat.
One clamp pulls the legs against the beveled board and the piece is ready to be routed.
Once we routed one rocker, we simply made an exact pattern of the two slots in space (knowing that they would all be the same) and made our rocker pattern to coincide.
This was a first go at this, and the results were amazing. All 12 chairs rocked the same and except for a couple of warped seats that threw it out a hair, there was no fussing or trimming. I don't know about you, but I've never found joy in shaving the end grain in rocker slots.
There is an even simpler way to rig this, and I will be constructing and showing it when I can, but I wanted to introduce this concept, if for no other reason than to get the point across that by isolating the variables that count, we can arrive at a new understanding of rocking chair construction that even a "purist" would find useful.
It's enough to make a Tasmanian smile!!