New State, New Shop #6: Bench Assemblies and Flattening the Top

Like many other people putting together a shop and trying to work with wood, I wanted a good bench. I have a nice “machine” bench made of 3 laminated layers of 3/4” birch plywood 8’ long on two tiers of metal drawers with a nice 6” machinist vise, but I needed to build a woodworking bench, as it would be useful and building it would be an excellent experience (how little I knew!)

So after reading innumerable articles, I decided I would build the Holtzapffel bench as popularized by Chris Schwarz. I adapted my approach from the Roubo bench in his Workbenches book as it used easily obtained (here in Illinois) Southern Yellow Pine and finessed cutting those huge tenons by leveraging the composite nature of the laminated yellow pine strips.

So I got a mess of yellow pine, a jointer, planer and table saw (see early blogs) and away I went. It went pretty well though I ran into some fun problems for a newbie including the amazing reactive yellow pine and the fact that a 16 foot piece of yellow pine that looked clean would have an amazing number of knots. But I persevered and managed to cut enough 7 foot billets for the top. I then jointed them on two sides, planed and ripped the 20 strips needed for the top. That was a bit of a journey as I had never used a jointer or planer before and there are subtleties to their tuning and operation that, while I read extensively, I learned only through trial and error.

But I eventually got them milled, two or four at a time, glued up and clamped

And glued up the assemblies and the super assemblies. Eventually, I had an entire top.

I supposed if I was very skilled the sub-assemblies and assemblies would have been clean, flat and aligned so the resulting top would be flat and aligned. Such as not the case with me… :-) Instead, the various elements of the assemblies were all a LITTLE off- just a 1/32” or 1/16” here and there, but the net net was that the top wasn’t flat. I didn’t think it would work to try and build up the frame and attach the top with it not flat. So I needed to flatten the top – and the bottom for that matter, since that is where the frame would actually be attached.

Seemed like I had two choices, the traditional, organic method with winding sticks and a plane, or the router/sled method. While I sure want to learn how to use a plane properly, trying to flatten a 24×72” slab with an old plane and my meager (non-existent) skills sounded was more than a little daunting. I decided therefore to try the sled technique, as outlined by Marc Spagnuolo on the Wood Whisperer site.

So I took 2 eight-foot lengths of 2×6, flattened and jointed them, built the sled and flattened the top a la Spagnuolo.

It worked pretty well, but there were more than a couple of gotchas:

• As the the top was only a slab, setting and keeping the rails for the sled to ride on was tricky
• Jointing 2 lengths of yellow pine 8 feet long completely flat was a challenge
• Setting the router to JUST the right depth so it was deep enough but not so deep as to waste wood was not easy (read: took more than one try)
• Controlling the router back and forth (and back and forth and back and forth and …) was not only tedious but tiring and a bit of a challenge
• In particular, when the router is pulled back towards the operator it is not going in the “right” direction so it tries to climb and pulls toward the operator. This is hard to overcome, especially for those with no experience with this step

I got the job done, but it was a learning experience. If was to do it again, here are the parts I would change:

• Buy two 8 foot lengths of extruded aluminum and screw these to the 8 foot lengths of pine to p provide a perfectly flat, smooth surface for the sled to ride on. Attach them on the outside of the pine, just extending up enough to provide the riding rail while still keeping the pine as the sacrificial edge for the router to bite into
• Add a cleat at each end of the sled so that it doesn’t shift and is easy to mooch along the rails
• Accept that it is a very slow process and that the router needs to come all the way towards the operator BEFORE shifting laterally for the next pass. That way, the climbing part is more or less eliminated, but it means that shifting the sled is a little more awkward.

It’s one one the amusing parts of this journey is that after each step in building the shop, the bench, etc. I look back and say to myself “Aha, NOW I see how I should have done this!” Come to think of it, that’s how writing software (which I do for a living) tends to work as well… :-)

-- Ric, N. Illinois "Design thrice, measure twice, cut once... slap forehead, start over"