|Project by A.W. "Pappy" Ford||posted 2252 days ago||1041 views||2 times favorited||6 comments|
This was the first project I’ve done since high school woodshop class. Copied here from my non-LJ blog as I chronicled the project. I apologize for the poor image quality:
Part 1 Blog entry – Sept 21, 2007
So, here I am. I have a 10″ table saw, a decent router, sander, and a few hand tools. I suppose if I’m going to call myself a woodworker I should start on some kind of a project. If necessity is the mother of invention then it would only make sense my project should fulfill some form of need. After only a few days spent setting up the workshop I realized I need some place to sit or my feetsies are going to be screaming an awful lot from standing on the concrete. I need a stool.
Sounds simple, right? Four legs and a seat. Uh huh. Thankfully I didn’t ask my online friends about this prior to getting the project started – after the fact most of them have said this is probably far too complicated a project for a newbie. “Build a box,” they say. “But I don’t need a box,” I say. Since I’d already purchased some decent oak for this project and did, after all, need a stool, I decided to forge ahead regardless of the difficulty.
Here is the basic drawing (somewhat to scale, 10 pixels = 1 inch) that I started out with. Click the thumbnail for the full-sized image. So what makes the stool so complex for a starter project? To start with, the angle of the legs. Here in the original drawing the legs are 10 degrees from square with the seat. Easy enough to cut using the miter gauge on the table saw, but becomes tricky where the joints are concerned. Based on sound advice I did change the degree of the angle back down to 5 degrees prior to starting.
The first problem the angled legs creates is that the legs cannot be secured at right angles to the seat. To do so would cause adjacent sides to be disproportionate – basically, one side would be wider at the base than the other, leading to an unstable stool. One should never have unstable stool… :)
To solve the problem I turned the legs 45 degrees so they angle out proportionately from the center of the seat to the corners. Problem solved, but (oops) this threw another monkey wrench into the design. Now that the legs are turned 45 degrees the crossbars running from leg to adjacent leg join the legs right at the 90 degree corner – not an easy joint to cut out at all, and one which I had no idea how to accomplish with my limited set of tools. Looks the the crossbars need re-thunk…
Think, thank, thunk, and the design changed again. Instead of taking the crossbars between adjacent legs I moved them to opposite legs, with a dado joint cut in each crossbar to allow them to fit together in the middle, creating an “x”, as shown in the photo at right. In addition to solving my problem, it halved the number of joints to be cut into the legs and increased stability, reduced wobble. ( By the way, I use the generic term “joints” here because I honestly don’t know what kind of joint this is called. I believe it is a mortise and tenon joint. If I got it right send me a cookie or something. If not, somebody please correct me.)
I routed out the mortises(?) in each leg, but because I don’t have any chisels yet I had to make them a little longer than the height of the crossbar due to the curvature of the router bit. Certainly not perfect, but correctable with some glue, sawdust, and a little wood filler. I wasn’t too concerned about this since these joints are turned towards the inside of the stool and barely visible.
Next I cut out the plywood seat support and attached it to the top of the legs with 1 1/2″ wood screws. I set it up without glue to verify that everything was level, and that all the 5 degree joints sat properly and I was pleasantly surprised to find everything set up quite nicely to within 1/32″ – not too shabby if I do say so.
At this point let me back track a little and explain why I should NOT have used the plywood seat support in the design. The legs, crossbars and seat are all make of oak, a nice hardwood. It will shrink and swell throughout the year due to temperature and moisture, but the plywood will not. Over time this will cause issues with the seat, making joints loose because the parts won’t all move in tandem with each other. What I should have done instead, was directly attach the legs to the seat using a mortise and tenon joint. I didn’t for two reasons: first, the joint would need to go all the way through to the top of the seat at a 5 degree angle in order to be stable, with the top edge of the leg leveled off and sanded down even with the seat – something I did not know how to accomplish. Second, if I had the knowledge I’m still very short on tools, and again have no chisels to do it right. Faking it with the crossbars in the legs was okay, but the seat top would have to be perfect or the whole thing could fall apart or at least not be safe to sit on, nor pretty to look at.
Part 2 blog entry – Sept 23, 2007
When we last left off with the workshop stool project I’d completed the legs, crossbars, and seat support. That left the seat and finishing to be done before calling the project complete.
The oak I bought for the seat was too thin, only 3/4 inch. So the first order of business was to cut two matching 12 inch by 12 inch pieces and glue them up to achieve a thicker seat. The very first advice I got from every woodworking community online was that you can never have enough clamps. Apparently, I didn’t listen to this close enough, because I thought 2 clamps (all I have at the moment) would be enough to handle the glue job. They weren’t.
After taking care to choose which side I wanted visible, I doled out a liberal amount of Titebond wood glue, aligned the two pieces, and clamped them down on opposite corners. To start with, this left too much play and the top piece kept floating on the excess glue, and two clamps did not provide enough even pressure across the entire surface to force out the excess glue. Now, in a panic as it was too late to stop, I pulled the bench vise off and used it to clamp down a third corner, which did help enough for me to be able to keep the two pieces lined up right and squeeze out much of the excess. Then I covered the piece with a shop rag to keep it from getting marred and placed 20 pounds of cinder block on the 4th corner.
Sloppy, sloppy, sloppy, and unfortunately it showed in the results. The three clamped corners formed a nice tight seal, while the fourth left about a 1/16 inch gap, with the dried glue forming a sort of wedge between the pieces. I’d have included a picture, but my crap-o-matic $10 digital camera won’t capture that kind of detail no matter how well I light it or how close I get…
It then bears repeating: you can never have enough clamps. In hindsight I would have placed clamps on the corners, in the middle, plus a clamp over each of the edges to guarantee proper alignment.
“Red Four, this is Red Leader. You need more clamps.”
“Cap’n, I don’t think the dilithium crystals in the warp core matrix can handle the stress: we just don’t have enough clamps.”
Oh well, live and learn I guess. Next I used Minwax Wood Filler to fill the holes inside the gap where air pockets had been. Using a 1/2 inch round-over bit, the next task was to route all the edges of the seat to give it a nice smooth top that wouldn’t poke me in the arse. A full course of sanding, from 80 grit, 100, 150, and then 220 and the seat piece is finished.
Attaching the seat was an easy affair. A little alignment and 5 woodscrews later and construction is done.
E gads! It was right about this point that I realized I should have done all the stain and finish work prior to buttoning it all up. I was so concerned that I wouldn’t get all the angles lined up right I didn’t even think about the order of putting it all together. Fortunately, the stool design is open enough I can reach everything with a brush without too much trouble.
The finishing pattern was as follows: single coat of Minwax Golden Oak stain, one very thin coat Minwax Polyurethane High Gloss, no sanding between this and a second coat of polyurethane (see note below), very light sand with 380 grit, and a final coat of polyurethane. Although the Minwax guide recommends sanding polyurethane between each coat, I’ve read that with polyurethane it is very easy to sand through the first layer and into the stain beneath. If this happens, the only way to fix it is to sand off all the stain and finish and start over. Instead, it’s recommended to apply a thinner than normal first coat and to sand only after the second and successive coats.
So, while enjoying my morning coffee this morning and waiting for the second coat to dry, I came across an interesting tip I wish I’d known about a few hours earlier. Wood magazine has a special issue out this month called “215 Fixes for Common Woodworking Mistakes“, and on pages 110-111 they have an excellent article on grain fillers. I never knew such a thing existed. With open grained woods, such as oak, ash, mahogany, and walnut, it is impossible to get that glass-smooth surface with just stain and topcoat. Run your fingers across one of these woods and you’ll feel the pits created by open pores and channels in the wood fiber. Grain filler, a mixture of solvent with silica, is designed to level up these pits and give you a surface worth bragging about.
Although I came across this information too late to use it this time around, you better believe I’ll be trying it out the next time I work with oak.
Well the only thing left to do is nail the steel furniture feet onto the legs – close enough that for the purposes of the blog it’s all complete. Hindsight is a great teacher and I definitely learned more of what to do and what not to do. Thank you to everyone that gave their advice, sent me emails telling me I was doing it all wrong , and of course the authors of the million or so articles I’ve read in the last few weeks.
Stay tuned, cause Pappy’s gonna be back soon with another project and I’ll need lots of correction and ass-kicking by wiser and more experienced folks than I.
Count on it. :)
-- --==[ Pappy ]==--