Dining Room Carver Chairs - Take 2

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Derek Cohen (Perth Oz)

Established Member
Joined
2 Mar 2005
Messages
3,277
Reaction score
1,515
Location
Perth, Australia
About a month ago I began a thread about designing and building two carver chairs to accompany the 6 bentwood chairs we have owned for the past 40 years. The chairs were purchased all those years ago along with a table, which is around 200 years old. We need a larger table, and time has come to replace it and add two more chairs.

8.jpg


The table planned will be a modern version of this, in Hard Maple and round Jarrah legs. The carver chairs need to match the table and blend with these chairs. Our taste is minimalist, Mid Century Danish, modern. An example of the lines I seek is this sofa table I built several years ago ...

28a.jpg


So I started thinking about the chairs I would build, and I took inspiration from this picture ...

est-living-sorrento-light-house-wolveridge-architects-3-750x540.jpg


... and began to evolve a design along similar lines. The 8 legs were started (just need rounding) and the seats carved, and then the tenons were cut. And that is where the problem arose. I had this idea for integrated tenons into through mortices in the legs. Very few chairs are built like this, and for a good reason - you cannot control for run out in the tenons, and run out make for weak tenons. And that is what I discovered. So I stopped the build, and decided to begin again from scratch.

I started looking more carefully at the chair I had come across and had treated rather casually. It has a name: DC 09, and it was designed by the Japanese- Scandinavian duo, Kyoko Inoda and Nils Sveje, in 2011. It is built by the Miyazaki factory in Japan.

Chair-seat9.webp


There is a challenge here - can I replicate it purely from photos? This is unlikely since one needs to examine an object in three dimensions to discover the subtleties of the design and construction. I have experience of this, having made an exact copy of Hans Wegner's "The Chair" or the "Round Chair" several years ago. What made this possible is that I own an original. One is mine and one is Wegner's ...

The-Chair.jpg


So the chairs I build will not be exact, but hopefully close. Actually, I am still on the fence about the arms and back and may modify this ... but will will see. We need to start with the seat. That is the key.

Help comes from two video I found ...





Some dimensions:

Chair-plan1.jpg


Chair-plan2.jpg


The doors above my bench make a place to pin details and photos ..

New1.jpg


The seat plan was scaled and drawn using images from the videos and photos.

Two half-templates were created - the first was a straight-sided outline of the seat. The reason for this is that I planned to use loose tenon joinery, which would enable the tenons to have straight grain for maximum strength, and the mortices would be made using a Domino. These would be 30mm long x 10mm wide, in other words would use custom-made loose tenons. The straight sides would make it easier to cut the mortices ahead of shaping the seats.

The second half-template was the actual outline of the seat, and this positions the tenons.

New3.jpg


Here the tenons are positioned ...

New5.jpg


A little jack-planning to flatten the underside of the seat blanks ..

New6.jpg


These are now sawn to shape ...

New7.jpg


New8.jpg


After this was done it became apparent just how the first shaping of the chairs differed from the DC09 design. Here is the first chair along with two legs. In the background can be seen the plan I had made for that build ...

New2.jpg


It looks quite good, with the angles and spacings appearing correct. In fact, they are quite different from the DC09. Below is the DC09 seat below the first seat. You can see the positioning of the legs ..

New9.jpg


Before I dominoed the mortices, the legs were mocked up to be sure of the angles. I discovered that they should be at 14 degrees and not 10 degrees, as they had been before. In the top right corner you can see a DC09 for comparison ...

New10.jpg


Guides for 14 degrees ...

New11.jpg


Marked and morticed ...

New13.jpg


New15.jpg


Tenons were made on the router table ...

New14.jpg


All is looking right so far ...

New16.jpg



The rule for chair backs is that thy need to allow for a lean of about 90-95 degrees. I have been thinking about this but decided to set aside this area for later. The priority is to get the seat shape correct - in three dimensions. - and its relationship with the legs (which were shaped in the first build - they are substantially correct, other than the upper ends being around 28mm against the - estimated - 40mm of the DC09. I will decide on whether to keep or replace these later). I have left the legs 1" longer in the lower half to adjust the tilt of the seat (the seat sits 18" off the ground and the ones I made previously are 19" in this respect).


I did stand the old chair along the table, and noted how low the arm rests were in reality - actually an ideal height for comfort. I do have some ideas how they may be "improved" aesthetically .... but this for later.

All advice gratefully received.

Regards from Perth

Derek
 
Last edited:
I was asked how I developed the drawings and plans, especially as there are numerous compound angles. It is a important question for those who want to use the ideas that are around us.

What I do is explore the videos and all photos, and then screen save relevant examples. From this I estimate or calculate size, approximate dimensions, and refine these over multiple examples. Here are some of the photos I used ....

















Regards from Perth

Derek
 
Derek did you give any consideration to making the seat to leg joints along the lines that Sam Maloof used with screws? As far as I know the joints have been proven to hold up long term.

Pete
 
Derek did you give any consideration to making the seat to leg joints along the lines that Sam Maloof used with screws? As far as I know the joints have been proven to hold up long term.

Pete

Pete, good idea. Early on I did consider using a Maloof joint, and spent a while with a number of examples. These are great and strong joints, but they fit in better with a certain type of chair. The joint is thicker and heavier-looking than I want, and they will not create the light aesthetic I am after.

Regards from Perth

Derek
 
A question about loose tenons

The legs will be joined to the seat with loose tenon mortice-and-tenon joinery. The tenons are Hard Maple, as with the rest of the build. I am in the process of choosing the stock for the 10mm wide x 30mm long tenons. The aim is to ensure that these are the best for a chair. Not any other purpose, but a chair.

The issue is grain direction: is it better to have the grain running horizontal or vertical?

Tenon1.jpg


Vertical grain should offer more rigidity, while horizontal grain more flexibility. While is preferred - your opinion?

The square stock is shaped on the router table ...

Tenon2.jpg


Tenon3.jpg


This is how it will look in the leg through mortice (minus the wedge and the rounded leg) ...

Tenon4.jpg


Regards from Perth

Derek
 
I've never made a chair so maybe I don't understand the question but why would you require flexibility in the M&T joint? In any case, I would imagine that the amount of flex would be very small but would it not encourage the joint to open up?
 
If you are putting wedges in do some test samples to see if one is more prone to splitting than the other. If no different than I prefer the look of the grain across the short dimension of the slot rather than along.

Pete
 
I've never made a chair so maybe I don't understand the question but why would you require flexibility in the M&T joint? In any case, I would imagine that the amount of flex would be very small but would it not encourage the joint to open up?

Chair joints are continually under stress. If there is no flex, or too little flex, the joint could break.

Regards from Perth

Derek
 
This is a long post with a good many photos. Treat them like a comic strip - flip by. It's just that there is so much more than one might realise that needs to get done when carving a seat. It is a 3D construction, and more complicated than joining square sections. This will become apparent as we progress.

These are the two seats. Both have been prepped with drilled holed to carve to depth. My plan is to work two seats alongside one another, completing a section on one and then duplicating it on the other, moving on, back-and-forth. The first stage is to power carve the rear of the seat using the Arbortech ...

Build1.jpg


Build-dots.jpg


Garage doors make for a handy white board to hang photos. These are the seats, and the angles provide the needed guidance.

Buil2.jpg


First shaping ..

Build3.jpg


It's rougher than it looks. A travisher begins the process of smoothing the curves ..

Buil4.jpg


Every now-and-then a scraper will refine the tear out ...

Build5.jpg


We begin the front section of the seat now, again using the Arbortech carver ...

Build6.jpg


And refine with a travisher ..

Build7.jpg


... always feeling the surface with a hand to detect any uinevenness ...

Build8.jpg


It is beginning to resemble the photos ...

Build9.jpg


At this stage it is time to cut out the seat surround, but before this can be done, the mortices (for the loose tenons) need to be preserved. This is not straight forward as they angle at 14 degrees. As a result, it is not possible to saw from end to end. The area around the tenon will require extra shaping.

My plan was to drill alongside the base, which would refine it and also create a curved root ...

Build10.jpg


Build11.jpg


Then as much as possible was sawn away with a jigsaw ...

Build12.jpg


Time to refine the front section and tenons. To do this, the underside of the seat needed to be made perfectly flat: this will provide a reference side to mark the boundaries at the front for carving to; also, taking down the bottom will define the bottom of the tenons ..

Build13.jpg


Build14.jpg


The shape of the underside is estimated using the templates for the top side ..

Build15.jpg


.. and refined with a rasp ..

Build16.jpg


Waste sawn away ...

Build17.jpg


.. and shaped ...

Build18.jpg


Again and again ...

Build19.jpg


Build20.jpg


Finally the front of the seat can be shaped as it flows from the end of the tenons ...

Build21.jpg


With raking light, the tear out and bumps are more easily seen, and now sanded down with 80 grit (more will be done much later on) ...

Build22.jpg


Build23.jpg


And that's it for today. Underside shaping to come.

Regards from Perth

Derek
 
Looking great! I've never tried to reproduce something as complicated as this but the process of looking at a piece, understanding it's construction, the stresses it is managing, and then replicating form and function is something I really enjoy.
 
Time to finish the seats ... sort of ... well, very close.

This is where we start - two seats completed on the upper side ...

Build33.jpg


Looking at the seat blanks from the side, it is apparent that they are 2" thick ...

Build34.jpg


... and they need to end up looking slim, like this ...

Build40.jpg


Clearly, there is a great deal of waste to remove from the sides. That is, not to thin the thickness, but to taper the sides to create the impression of thinness.

Here is another view, which better illustrates this ...

Build45.jpg


The MFT makes a good bench for carving ...

Build35.jpg


Lines are roughly pencilled on the seat and then the Arbortech is used to carve away as much waste as possible ..

Build36.jpg


I relied on a Auriou 10-grain rasp to do most of the shaping ...

Build37.jpg


... until I was loaned the largest rasp I have ever seen. Large enough to cause most here rasp-envy. It says Nicholson on it, but it appeared larger than a Nicholson #50, and far, far coarser than the 11-grain it is advertised as having. Any ideas?

Build38.jpg


I moved between rasps and spokeshaves to shape and smooth the curves ..

Build39.jpg


To be frank, I worked without a specific plan other than to create fair curves - the curves almost decided what waste to remove, while I monitored the photos I had for reassurance.

In the end, with everything sanded to 80-grit, this was the result. From the underside ...

Build41.jpg


The upper side, from the rear end ...

Build42.jpg


And side ...

Build43.jpg


And one more photo, taken at an angle similar to the "slim line" view above ...

Build44.jpg


There is still work remaining in the seats: every time I look at the chair photos new details become apparent. So I shall get on with the legs, and return to fine-tune the seats later.

Regards from Perth

Derek
 
New Legs

Every time I sit down to write up the progress made I feel like apologising for how little there is to show for the efforts made. I have made this observation before - it is tough to copy from photos when there are no measurements to follow, but it is especially so when the piece is curved and changes shape from different angles. I keep finding new detail to add in, and it seems like a never ending carousel.

One example, the front underside of the seat is concave and not convex. I added this, but need to do more ...

L1.jpg


The seats need quite a bit more work, mainly refining details. For example, the sides need to be tapered more. For later.

For now I return to the legs.

I did make 8 legs when building Mark 1 of the chairs, but now I am about to re-make them completely. Why? Because the first set of legs were designed for a chair which was inspired by the DC 09 Chair, but now that I am attempting to get close to this design, the legs also need to be in keeping.

Step one was to create a template, and then mark out 8 legs. These were bandsawed close to the lines, and then cleaned up with spokeshaves ...

L2.jpg


The second time around is always easier, and marking out the mortises was much more efficient by measuring the front of the legs ...

L3.jpg


... and then dropping the verticals ...

L4.jpg


This is made possible by ensuring all sides were kept square when the leg blanks were made.

As before, the Domino was used to mortice the through tenons. The legs are clamped to a mortising fixture.

L5.jpg


The mortises are 30mm long and 10mm wide, and through the 30mm thick legs. Since the Domino 500 can only rout to a depth of 28mm, it was done by working half way from each side.

The mortise needed to be marked very accurately, and to do this I used a marking jig I developed and wrote about recently ...

L7.jpg


The article is here: https://www.inthewoodshop.com/Powered Tools and Machinery/DominoDW500AsAMortiser.html

It was possible to be as accurate as this for both sides and all the legs ...

L6.jpg


This ended with a pile of legs blanks plus mortises ..

L8.jpg


Early on in the build of Mark 1, I attempted to use a half-round bit on a router table to round the legs. This was a disaster as the Rock Maple blew up, leaving me with wet trousers. Looking back on this sad experience, I realised that I had literally bitten off more than I should have chewed. So I was determined to try again, but this time rout in stages, little-by-little ...

L9.jpg


The leg at the rear is one I attempted purely with spokeshaves. It was a miserable experience - Rock Maple is well named.

Here are 8 semi-finished legs. The corners have been rounded, but much of the leg is still square-ish owing to the tapered profile - the legs start at 35mm at the top, are 30mm by the mortise, and end at 22mm at the feet. Consequently, there is still a whole lot of shaping still to do.

L10.jpg


Look closely and you can see the flats on the sides.

There followed a lot of spokeshaving.

A progress shot of 4 completed legs and 4 incompleted legs ...

L13.jpg


"Completed" really means "done for now". There is constant refining. Slowly the square becomes less so, and then round, but with imperfections, and then eventually there are just fine tracks ...

L11.jpg


I must admit that it is so tempting to leave it like this, where fingers can caress the tool marks ...

L12.jpg


This would look better on a different style of chair, and the DC 09 is better suited to a sanded finish.

Here are two of the legs to gauge progress. Note the photo on the wall in the background for comparison (also recognise that the arm section of the legs has been left long at this stage)...

L14.jpg


L15.jpg


Until next time.

Regards from Perth

Derek
 
Back
Top