Dining chair - one down... 5 to go.

The first ‘sitting’ in the chair was quite promising, so I kept going.

Fitting the piece for the backrest was first. There was a bit of trial and error with the backrest, but it went fairly smoothy. I opted to use 3 Dominos in each end to connect to the back legs. Cutting the Dominos in the backrest was easy - the really tricky part was how to index the locations on the back legs. With some careful layout lines, it worked - at least on this first chair.

The next step was the arm rests. These were also challenging from an indexing/locating standpoint. I used a single Domino oriented vertically to connect the arm rest to the back leg. The front of the arm rest is connected with a 1/2"D, 2" long dowel. The photo above shows the 2" thick over-sized arm blanks in the correct position with the Dominos and dowels in place. It was time to start thinking about removing some of the bulk and doing some more shaping.

I really wrestled with just how much shaping to do before the glue up. On the one hand, parts were way easier to shape before they were glued - but if I inadvertently removed too much material in any spot - it could spell disaster. I was also nervous about the fact that this was a very unfamiliar process, building a very unfamiliar form. I opted to play it safe and only wasted out what I was absolutely certain about.

The back rest struck me as a pretty safe place to do some shaping. There was a lot of material to remove from the blank, and the joint between the leg and the backrest would be really tricky to access once it was glued. Below are 3 shots of the rough shaping.






You will also notice the heavy chamfer to the back leg - just behind the back rest. This was another area that would be really difficult to get to once it was glued up, so I did it before hand.




I also refined the shape of the arm rests. I left the top flat to allow for clamping. I did shape the underside of the arms before glue-up. I used the RAS for this and took advantage of the radius to scoop it out.




I also further refined the front edge of the seat to make it look a little thinner. The above photo shows the thinned out seat edge.

At this point, I was pretty comfortable with how much I had refined the shapes and did not want to push my luck. Time to glue it up.




The complexity of this chair became apparent as I planned for the glue-up - and the advantage of screws became crystal clear. I had to glue the arm rest at both ends and then install both legs onto the seat at the same time. And glue one end of the back rest... and have the other two legs in position in order to index everything and give myself proper clamping surfaces.

The above photo shows the first glue-up. You can see the vertical Domino in the back leg. I had to make another pair of clamping pads for the top of the back legs to allow for even clamping pressure to glue the back rest. The first glue up worked well, as did the second and final one. Phew!





Riley was the first person to try the chair all glued up.

Now for the stage I was most looking forward to - the final shaping.



The first day of shaping was pure bliss. I was armed with a pair of spokeshaves and a small army of (left handed) Auriou rasps. The shaves were marvelous shaping tools and made quick, tear-out free work of the curly walnut. For those areas the shaves could not get to - I used rasps.










Jill took a few action shots.



I kept the mock-up close at hand - you can see it in the background.





You can get see the cross section of the legs by looking at the ends. You can also see the scooped out underside of the arm rest.




This felt a little weird - but it was the best way to shorten the arms by 3/8".

There was one aspect to this chair that I did not fully resolve in the mock-up and I hoped I would be able to figure it out by the time I needed to. It had to do with the three way joint where the arm rest meets the back leg - right where the back rest joins the back leg. The ‘armpit’.




In the above photo - you can see how it looked. It was not a smooth transition - but I was not totally clear with how to resolve it. I had hoped that during the clean-up stage, it would just come to me. No such luck.

So I called my good friend Pat for some help. Pat had been along for the chair-making ride since the beginning. There is a very long list of things that I admire about Pat, but in this particular case, I was after his ability to see. In his typical polite way - he found this weakest aspect to the chair and called me on it. It was the reminder I needed (and was looking for) not to settle. We spent at least an hour on the phone talking through possible solutions and the overall design of the chair. I hung up the phone with a few ideas, a renewed sense of hope and gratitude for his help.

I realized that the problem with this area was the three converging elements were all competing for equal amounts of attention. One of the intersections needed to be visually dominant in order for it to work. As soon as I realized this - the solution was quite clear. The connection between the back leg and the arm rest had to be dominant - the back rest to back leg connection was secondary.


I did a quick test to the mock-up to see how it would look. Other than the fact that the Domino broke through :) - it looked pretty good. Just to be certain, I did another test in Photoshop. Way easier to move a few pixels around than to add wood!

(above - the photoshop altered test)

(the finished transition)

The photo shop image did it for me and I carefully proceeded to make the change to the chair. I was sweatin it though - worrying about the Domino breaking through. Thankfully it didn’t, and I was quite happy with how it worked out in the end.

With all the shaping done it was time to move to the sanding stage. I was warned by several people that this was by far the worst part of making one of these types of chairs... the countless hours of sanding. I had finished the shaping stage with fairly fine files, and was hopeful that I could start with a fairly fine grit of paper. I started with 180 grit which seemed to be ok. The 180 took many hours but each grit after than went increasingly quicker. It reminded me of using waterstones - the 1000 grit stone does the bulk of the work.

I mixed my own Maloof type finish - 1/3 tung oil, 1/3 linseed oil and 1/3 varnish (P&L 38 to be specific). 8 coats later - the chair was done. Click on any of the images for a larger view.
































We have been enjoying the chair for the last 3 weeks now and I am pleasantly shocked that there is only one change that I will make to the remaining 5 chairs. I will remove a bit more material from behind the pommel - the ridge in the middle of the seat. It transitions nicely now, but if you slouch in the chair to read the paper for example, you can feel a bit too much of the transition into the pommel. Other than that though - it is quite comfortable. It is also heavier than I would like - but I am not sure how to resolve this. Part of it may be the fact that this was unusually heavy walnut - but I suspect I could have started with thinner stock for the seats.

There are quite a few people to thank for contributing to this project. The 20 or so family, friends and neighbours who generously lent their posteriors to make sure the chair was comfortable.

And two special thank-yous. To Pat for all his thoughtful insight and for not letting me settle. I have always appreciated the fact that we can be on opposite sides of the continent and yet have very meaningful design discussions. And to Derrick - who generously fielded all my emails and phone calls and provided insight into the process.


And now for the other 5...




But before I get to these - I have quite a few planes to get to. I have been busy in the shop working on several planes - including this one...

Dining room chairs - milling, cutting joints & the seat

My Dad came over and helped me with 2 very important tasks. The first was to put the log back together so I could mark the boards to get the best layout for parts. I was particularly interested in getting the seats sorted out. There was one massive board that was wide enough to get three single board seats. By putting the log back together, we were able to get two more seats that were slip-matched.



These two, 18" wide book-matched planks were set aside for the future table.




These were perfect for the front and back legs and the arms.



The above shot made me somewhat sick. I was half prepared for lightning to strike me down for cross cutting this amazing wide board. I stood there for at least 20 minutes pacing around building the courage to do it.



I got over it pretty quickly as soon as I saw the seat blanks.



Figuring out the layout for six chairs was way more work than I anticipated. There were some defects to work around and it was a struggle to wrap my head around all the waste. Thankfully - all the stock was at least 8/4 with some being 9/4 and even a few 10/4 pieces. Dad helped me mill everything flat on the jointer and then through the thickness planer. I needed a full 2" thick for the arms - the rest of the stock needed to be 7/4. Once we had it down to 2", I re-sawed several of the boards to take a single piece of 1/8" veneer. I could not stand the thought of all that curly walnut in the cyclone. And who knows - those veneers may end up finding their way into a future sideboard.

(a few pieces of the 1/8" veneer)

Here is the pile of chair parts. The front legs are on the saw horses, the seats are on the bottom of the pile, then the back legs and the arm rests on top.

As I was milling the parts for all 6 chairs, I realized that it would make a lot of sense to work on all 6 chairs up to a certain point and then take one of them and finish it off. I decided to cut all the joints for the front and back legs and prepare all the seat parts.

The leg joints.

The front and back legs are attached to the seat using the joint that Sam Maloof invented. It struck me as an elegant and sturdy way to attach the leg if one wanted to shape the joint so it would blend together. The front leg was the easiest - it is actually a 90 degree joint - so that is where I started. I did get a trial run of these joints on the prototype, but poplar is pretty squishy compared to walnut so a full set of new jigs was required for the walnut chairs.

This joint is fairly simple and straightforward, but it is not easy. It does require an extremely high level of accuracy when cutting the initial dados - about .001" of wiggle room. Set-up is critical - everything must be 90 degrees. These joints reminded me of plane making from the standpoint that if the first cut is not done accurately - you will be fighting with it at every point thereafter. Success (or failure) is based on the previous step being perfect and any deviation or error will send you back to the beginning. It was at this point that I was grateful for my plane making experience.




Here is a shot of the completed front leg to seat joint. One other early decision was to use dowels and Dominos instead of screws and plugs to hold everything together. I used a 1/2" diameter, 2" long dowel to connect the front legs to the seat.





There is also a 1/2"D, 2" long dowel connecting the top of the front leg to the arm rest. The top of the front leg is at a 1.5 degree angle which required a jig to ensure the dowel was square to that face.

Locating all the holes was a little nerve wracking. You can see the pencil layout lines on the top of the front leg. These are the rough bandsaw cut lines - there is further refinement after these cuts and I was holding my breath that none of the dowels would break through.

The back leg joint is much more complicated. The back legs are tilted in 3 degrees (closer at the top than at at the bottom) and rotated in by 5 degrees.

(this is the only shot I have showing the 3 degree angle to the seat)

The above photo shows the back leg fit to the outside piece of the seat.



An inside view of the joint.

I used two, 1/2"D dowels to connect the back leg to the seat. I made one a little shorter for 2 reasons. I figured it would make alignment during glue-up a little easier if they were staggered, and I was also concerned about the furthest dowel breaking through the back leg during the shaping.






I cut the dowels in the seat component first. When I located the holes in the leg, I offset them by .006" - two pieces of paper. I wanted the dowels to act like a bit of a drawbore to pull the joint in tight along the edge parallel to the dowels. The above photo shows the joint going together for a test fit.

(you can see the 5 degree rotation in this shot)

The seat.

I have watched many videos of Sam Maloof shaping chair parts on the bandsaw. It is an amazing display of control and confidence. I was by no means going to attempt to use a bandsaw to that extent... but I was comfortable doing some of the shaping.



The above photo shows the contouring to the seat done on the bandsaw. I went really slowly - this was about an hour and a half to do. On the second board from the right, you can see an almost disaster - there is a dark shadow line that indicated a cut that almost went too far. What took so much of the time was figuring out how to keep track of the layout lines on the opposite edge - the one you could not see.

After the seat was glued up, it was time to shape it. A good friend of mine strongly suggested that I invest in the Festool RAS 115. It is technically a sander, but really, it is a grinder. And a marvelous grinder at that! Armed with 24 grit paper, I was able to shape the seat at an alarming rate. Not only did the paper hold up surprisingly well... the grinder itself was really easy to control. I worked through the grits from 24, to 26, to 50 and then to 80. Once I got to 80 grit, I switched to the Festool ETS 150. This sander was great to smooth out any of the ridges left from the RAS 115. I started with 50 grit, then 80 and worked my way to 220.

Below are 2 shots of the contoured seat. I took these in low light (single light source) which was the best way to be able to see the contour and not be distracted by all that pesky curly grain:)






I was pretty happy with how the contouring came out and decided to keep going. I figured I could make refinements further along if the shape was not quite right.




Shaping the front edge of the chair was a lot of fun. Spokeshaves were the perfect tool for this task. There were a few spots I could not get into easily but the Auriou rasps solved that problem in short order.

With the front edge done, I was finally going to be able to put the legs on and see how it felt. One quick problem though... the 3 degree tilt and 5 degree rotation to the back legs made it impossible to clamp them in place. I made up 2 jigs one for each leg.




The jig consists of two angled pieces of walnut - one at 3 degrees and the other at 5 degrees. These were oriented (and glued together) in such a way that they would be parallel to each other and allow for clamping and the eventual glue-up. I screwed two pieces of 1/4" plywood to the top and bottom so they would be indexed off the edge of the seat. I glued a piece of 220 grit sandpaper to the walnut - I did not want anything to rotate out.



Finally - I could take the first test ‘sit’.