Shaker Bench II

My previous post was on a project I recently made off a plan of a Shaker bench.  The author lengthened the bench and added a backrest.  If you refer back to that post you will read how I battled with the three way assembly due the sequence given in the plans.  So I decided to redo this bench the way I thought was more logical and of course easier to do.

I decided to leave the backrest out for a number of reasons.  I didn’t have enough wood, I was remaking this bench just to try and find a better way to assemble the sub-assembly which didn’t involve the backrest, and the original Shaker design did not have a back rest-par for the course for these ascetic people I suppose.

In fact I only had one plank of Poplar and one plank of Kiaat  (Pterocarpus angolensis) available.  So I decided to use the bland Poplar for the underside (legs and seat supports),  and the rich-grained Kiaat for the seat.

I started by cross cutting the Poplar plank to length with my radial arm saw-4 pieces for the legs and the remaining long piece for the seat supports.

20131103_170501c 20131103_170520c

Above photos show single Poplar plank cross-cut into 4 short pieces (2 each per leg) and one long piece for the 2 seat supports.  The first photo shows the some offcuts sitting on the saw-bench underneath which are kept for testing the finish.

The pieces were then jointed and thicknessed.  My son developed an ingenious way to hold the dust extraction hose from getting in the way of the thicknesser outfeed table.  Basically an elastic tie attached to a tensioned cable at ceiling height:

20131103_195602c 20131103_195633c

Then the bench end pieces which make up the legs were glued together:


These end pieces then had an arch cut out of them to form two legs and they will then be lap-joined to the seat support.  Again see the previous post for more on this.

20131106_205738 20131106_213937

After the legs/side pieces were cut out, they were joined to the supporting seat supports. The lap joint was cut with a tenon saw and only a dry fit was done.


In the above photo you will notice that only the the seat support to your right has been joined.  You will also notice in both pictures how the seat support is not flush with the top of the leg/side piece, but is recessed by 12mm.  The part of the leg/side piece sitting proud will fit into a dado that will be cut into the underside of the seat, hence the three way fit that I referred to earlier.

Now we come to the part where I differed in my order from the previous bench.  Whereas the previous bench’s instructions called for the dado to be cut into the seat support first, I now took the whole dry fitted sub-assembly pictured above and laid it onto the underside of the seat.  I then marked out the dado directly and cut them.  This seemed liked the more logical (& esier way) to do things.  In reality it was much easier.

20140329_152807 20140327_185344 20140327_185322

I also added butterfly or dovetail keys on the seat for effect.  I made them from the same wood type so as to lessen the contrast. I placed four in a row along the centre line, two larger ones on the outside and two smaller on the inside. I used the Rockler jig to cut them out.

20140322_233447 IMG-20140325-WA0000 IMG-20140325-WA0004 IMG-20140325-WA0006 IMG-20140325-WA0010 IMG-20140325-WA0014

Kiaat tears out very badly so I quickly gave up on the handplane and continued with the #80 scraper. I mixed fresh orange shellac and applied 5 coats with a brush and finished with one coat of diluted poly.  The shellac decreased the contrast between the light Poplar & darker Kiaat.

twin bench c 1 twin bench c 2 twin bench c 3 twin bench c 4


Shaker Cherry Bench.

20130730 rough planks 20130804 planed & thicknessed 20130804 ripping to width

  Above photos show the initial sequence from left to right: raw Cherry planks – planing & thicknessing -ripping to correct width.

This project plan was taken from the December 2004 edition of Popular Woodworking by Robert W. Lang.  I chose to do this project because I liked the look of the bench, and not out of any particular need for this piece of furniture.  The plan made it look like a pretty straightforward piece to make, but in reality I found the methods employed and order of events difficult to lay out and assemble the piece as well as confusing to understand.


Some more initial shots, from left to right: the original planks from the lumber yard, planed & thicknessed and zoomed in view.

For example, the dados that run through the underside of the seat were cut first by the author, and I would have done them last-after making the sub-assembly (consisting of the end pieces that form the legs and the seat supports which run under the seat and end in an ogee).  See proceeding photos:

20131027_235733c 20131027_235829c

Above photos show the ogee-curved seat support resting under the seat (it’s actually fitted into a dado that is cut through the underside of the seat); and simultaneously lap-joined with the side pieces that make up the legs.

Then with the sub-assembly dry-assembled, I would have placed the whole thing onto the seat and traced the location of the dados directly.  The reason I would have done this  is because there is a three way alignment that has to happen here.  (i) The seat supports have to be centred so that they are equidistant from the ends of the seat.  (ii) The depth of saw-cut for the lap joint on the end-pieces had to take the depth of the seat dado into account as well.  (iii) The seat support lap joints have to be equidistant from the front & back edges of the seat.  (iv) Then the depth of the saw-cut in the end pieces had to be deeper (I chose 55 mm), in order the keep the seat support from becoming too thin and so compromising the strength of the bench.  (None of this was mentioned in the article-I would have at least pointed out that the shallower the saw cut for the lap joint on the end piece, the thinner (and therefore weaker) the seat support would have had to become due to a corresponding deeper lap joint saw-cut).    So starting with the dado, and then trying to line up all the pieces to this was very difficult and time consuming.  Another thing I would have done differently is to cut the ogees at the end, again after a dry-assembly of the whole seat.  This would have made the job of centring much easier in my opinion. I am remaking this bench without a backrest to prove this point.  So look up my next post labelled “Shaker Bench II”

DSCF0409 DSCF0410

The wood used for the seat was from a solid plank, so the one corner was not square and some Heartwood is showing white towards the back of the seat.  Because of the width, I had to joint it by hand as shown above.  For this reason the underside was not finished but left rough as is found in many antique hand crafted tools.

20130804 gluing up end pieces20130922_21020820130922_210219

Above photos show glue up.  The first photos is the end pieces that make up the legs.  Second photo shows the main assembly and the third photo a detail on the glue up of the seat support ogee.  I used bottled hide glue and left the piece in the clamps for 24 hours.

20130923_213351 20130923_224640 20130923_224705

Above photos show the bench assembled except for the back rest, which still has to come.  I chose the do this last so as to get it properly centred.

20130923_214006 20130923_21394920130925_22003520130925_200443

Above photos show arch cut-out on two end pieces to form two legs.

I had a template for the half oval cut-out for the end piece made from plastic (shown above).  I traced this onto the end piece, and then cut the wood out with a coping saw, rasp and spokeshave.  In the end, the  card scraper proved to be the best tool out of all to get the curve smooth, but I suppose a hand-stitched rasp would have made a bigger difference?


Above photos show back rest installation.

Finally I screwed and plugged the seat and back rest.  The best fit was achieved using  an Imperial drill and Metric dowel one size apart, in this case a 5/16″ drill with an 8 mm dowel.


Above photos show copper screws added as per plan and covered with dowels. Dowels were cut level with the wood with a flush trim saw. Four screws were used on the seat and two on the back rest.

Perhaps I like being spoon fed but I would have appreciated the plans mentioning the dowel diameter as well as their positioning on the bench.  On the other hand, if one read George Walker’s Design Matters article in the latest Popular Woodworking Magazine (issue #207), he makes a valid point where we should consider using plans up to a point, and then input our personal preferences, like what we need to store in the piece, or adapting the dimensions of the bench or chair to how tall or short we are.  If I were to apply that thinking here, I would have made the bench a half inch shorter, the seat 4 inches wider and I would have moved the back rest further back by sinking it into the back support with a full depth rabbet.  I would also have made the wood thicker than the specified 3/4″ for Cherry and relative to the length, as the back rest bowed slightly when leaned against.  Otherwise I would have used White Oak, 7/8″ thick.

And here are some photos of the finished project.

DSCF0417c DSCF0418c DSCF0420c