Once the glue had cured I flattened and shaped the seat blanks. Flattening did not take much time as I’d made sure that the glue joints were square to the face of the seat, and so some gentle traversing passes with the jointer plane cleaned up the glue seam and flattened the top.
The underside of the seats needed a little more work to bring the seats into a consistent thickness, and for this I used a scrub plane from Lie-Nielsen. This is a new addition to my tool chest, and the first time I’m used it on a project. Although more compact than a jack plane, the scrub has a much more pronounced (3″) radius to the iron which means that it can take a much deeper bite; perfect for rapidly removing material when processing stock. Taking traversing cuts with the scrub rapidly levelled the underside of the seats, and left a scalloped texture. For show faces I would remove this texture, but for the underside of chairseats or drawers it is a wonderful surprise for exploratory fingers, not to mention proof of the handwork process. So I left the seats as they were. My initial impressions of the scrub is that as a rapid stock removal tool it functions perfectly, and as a very simple tool (no depth adjustment mechanism or lateral adjuster) it may also be a great tool for children or beginners. My original reason for buying it was to assist in preparing the timber for the oak Roubo bench (coming later this year!) and I expect it will do well at that task.
The seats topped out at 1″ thick, which for a chair aimed at 3-5 year olds should be plenty thick enough. Chris Williams tells me that John Brown’s childsized chair seats were in the region of 1 1/8″ thick, while the historic chair I’m following on this build had a hefty 1 5/8″ thick seat which was beveled down to 1 1/8″ at the edges. I didn’t have material in stock thick enough for a 1 5/8″ thick seat, although I may try that approach on my next child-sized chair. But as I say, the 1″ thick oak I’m using for this pair of chairs will be fine.
The seat is trapezoidal in shape – 17 1/2″ wide at the front and 15″ wide at the back giving a 1 1/4″ taper on each side, and 12″ deep. I jointed the front edge and then gauged the back edge off that. When orientating the seat, I placed the glue seam as deep into the seat as possible so that it lands away from the three leg mortises. With the front and back edges jointed and down to dimension, I then laid out the sides. There wasn’t much meat to remove from the sides, and on a workpiece of this side sawing off thin slivers at the saw bench is awkward work, so I cut down my layout lines on the bandsaw to remove the excess, before cleaning up the edges with a low angle block plane. The face vise on my Sjoberg bench has all the holding power of cottage cheese, and planing such a large amount of end grain with any meaningful pressure encouraged the seat to rotate in the vise. Setting a saw bench underneather the workbench to support the seat held everything secure while I worked the edges.
The final element to the seat shape is the rear edge, which has a gentle convex curve. From my collection of photos of this chair, I judged the centre point of the curve to be 1/2″ higher than the corners of the seat, so that’s the curve I went with. A drawing bow tensioned to give the right curvature made easy work of laying out the curve, which I traced with a marking knife before cutting on the bandsaw (although a coping saw would have done the job just as well). To fair the curve I used a combination of spokeshave, 9 grain cabinet maker’s rasp, and a 13 grain modeller’s rasp.
The seats are now ready for the mortises to be drilled. After that I will be able to round the front corners of the seat (whch are currently looking rather sharp). Laying out the mortises requires having definite corners at the front of the seat, which is why I didn’t round them at this stage of the build.