Child's Sunday School Chair - Part 4: Stretchers and Undercarriage

The stretchers that join the legs were quick to make.  Three of the four were 11/16" diameter cylindrical with 3/4" long, 5/8" diameter straight tenons.  Square up a piece of wood to 11/16, octagonize, ease the facets with spokeshave and ... done.

The front stretcher only differed in that it bulged to about 1 1/16" in the middle.  After squaring a piece to 1 1/16", I used a template to draw the shape, then used drawknife and spokeshave to arrive at a square version of the stretcher.  Then octagonal, then roundish.

In process

A little sanding and, fini

To find the locations of the front and back stretchers, I measured down from the underside of the seat 5 1/2" and made a mark on the legs.  Then I stretched a rubber band from front left to front right leg at that 5 1/2" location.  I found a center point for boring each leg by measuring between straight portions of the rubber band.  The rubber band method also allows you to find the angle for boring.

Rubber band stretched from one front leg to the other
(dark wood board is there so you can see the yellow rubber band)

Getting the angle at which to bore the leg

The side stretchers were done similarly, except that the hole centers were 2" lower than those for the front and back stretchers.

The leg boring setup was as seen in the following picture: the leg was held in two cradles and clamped to the benchtop with a holdfast.  The sliding bevel was placed in front of the leg and a small round mirror was positioned about a foot behind the leg so I could see how I was doing relative to the angle of the sliding bevel.  Another mirror was placed to the left of the leg, with a combination square standing between the mirror and the leg to help judge vertical.  While boring with bit and brace, I kept track of both angles using the mirrors.  And I used a step stool to get me higher so I would be better able to bore from above.

Leg boring setup

There was one potential hiccup here.  The top end of the leg had already been tapered and that tapered end was placed in the left cradle.  The bottom of the leg was still full diameter, so the leg was not clamped horizontally.  Therefore, my sliding bevel, which was sitting on the bench top, indicated the wrong angle at which to bore.

I realized this later, after fitting the stretchers.  Turned out not to matter much.  It probably changed the boring angle by a couple of degrees, which was probably within my range of variability anyway.

Once the holes were bored, the legs were put into the seat and the lengths for all stretchers were determined using a couple of dowels.

Two sticks used to find the length of a stretcher

Then that length was transferred to the stretcher, which was cut to length.  Tenons were made on the ends of each stretcher using a drawknife.  Finally, a test fit of the undercarriage felt very satisfying.  It really tightened up the whole thing.

First test fit of the seat and undercarriage

Next time I'll get into the spindles and crest rail.

Child's Sunday School Chair - Part 3: Legs

For the legs, I started out with squared blanks of poplar, made them octagonal using a jack plane and then round(ish) with a spokeshave.  I tried turning the right rear leg, but the poplar was too soft to stay put on the pikes of my bungee lathe and I almost ruined the part.

Legs marked

At the upper end of the legs, I marked a line 4 1/4" from the top to mark the beginning of a taper, and 1 1/4" from the top to mark the tenon area.

Top of legs marked

Added a 1" diameter circle on top end as something to shoot for

In the picture below, you can see an original leg next to two new legs.  The original did not taper all the way to the end - it has a "shoulder" of sorts.  The original mortises were not tapered and the legs had straight cylindrical tenons.  My chair has through tapered mortises, into which I fit tapered tenons.  I experimented with two methods of forming the tenon - one like the original leg and one with a gradual taper from the 4 1/4" line.

Two methods of starting the tenon

Either way, I had to make the end about 1" diameter so that it would fit into the tapered tenon cutter that I made (and wrote about here) a few months ago.

Tenon cutter cutting a test piece

When cutting the real thing, I used two squares to make sure I was cutting straight
(tenon cutter is held horizontally in the vise)

The middle part of the legs contains undercarriage joinery and so it remained cylindrical, but the lower end of the legs tapered down to 1 1/8" at the feet.

Before fitting the stretchers to the legs, the legs had to be fitted to the seat.  So next up was to bore and taper the leg mortises.  To do this, the sightlines were laid out on the underside of the seat.

Layout for boring the leg mortises

I used a sliding bevel to guide the resultant angle, with a mirror placed on the bench to my left so I could monitor progress without stepping to the side.  A combination square was set along the sightline so I could monitor how well I was doing in that direction.

Boring at the proper resultant angle

A combo square with the edge of its blade right on the sightline ensures no leaning left-right

It's not in the above picture, but to avoid splintering the exit side of the hole, I clamped a sacrificial piece of wood to the seat.  After boring the holes, I tapered them with the reamer I made a few months ago (and wrote about here and here).

Reaming the hole to a 6° taper

The picture above shows the alignment of the top of the reamer with both the right edge of the square's blade and the sightline.  This helped make sure I don't change the angle of the mortise while tapering it.

All four legs fitted into their mortises

Next up: stretchers and fitting the undercarriage.

Child's Sunday School Chair - Part 2: Design Considerations

Last week's post described the little Sunday School chair that I picked up.  I used that chair as a model to make my own version.

The original chair

There were some things that I changed for various reasons.  First, I don't have steam-bending capability, so the arched rail was out.  I decided to keep the three interior spindles and add a thicker spindle on each side of the three interior ones.

I kept the middle three spindles in about the same locations as on the original chair.  For the outside two, I wanted to space them out roughly equally along the back curve of the seat, and away from the curve of the back by about 5/8".  Sketchup helped a lot with this.

Up top of the five spindles is a crest rail.

Partial Sketchup model (stretcher rails in undercarriage not shown)

I used the original chair to estimate the curvature - about a 25" radius.  It is 1" thick and 2 1/8" tall.  For the length of the crest rail, I used the original chair for guidance - the widest part of the arched back was 13 1/2", so I wanted the crest rail on my chair to be something similar.  Looking at some chair pictures, the crest rail is commonly wider than the seat.  I settled on 14 1/2" (which became a little less after shaping).

To determine how far back the crest rail should be, I looked to the original chair.  The top of the arch is set back about 1 1/2" from a line vertically perpendicular to the seat at its back edge.  I used the same for my chair.  For a crest rail about 13-15" higher than the seat, this gives a seat-back angle of about 10-11°.

For the undercarriage, I wanted to keep close to the original.  But I thought the original's leg mortises were a little close to the edge of the seat.

Underside of original seat

In the picture above of the original seat, the front leg mortises are to the right; the rear leg mortises are at far left (the other two holes were through mortises for the arched back).  The nearest edge of the front leg mortises was only about 1/2"-5/8" from the side of the seat and about 7/8" from the front edge.  The original chair did not have through leg mortises (they were about 7/8" deep in a 1-1/16" thick seat), and maybe that was a factor in the builder's choice of mortise location.  Another factor could have been that the original chair seat was made of cherry or some other hardwood.

A 7/8" deep leg mortise on the original seat

My chair is being made from poplar - seat, legs, stretchers, spindles, crest - all poplar.  Don't ask - it's what I have around the shop.  And with a weaker wood, I wanted a bit more meat around the mortises so that the seat wouldn't blow apart when driving the wedges or when someone sits on the chair.

So I moved the tops of the front legs toward the center of the chair about 1/2" and towards the back by about 1/4".  I kept the bottoms of the legs in the same position to give the chair the same stability as the original and then used my resultant and sightline calculator to figure out the new rake and splay angles.  The original front legs had 2° rake and 4° splay and the new chair has 3.4° and 6.7°.

Leg mortise locations on underside of the chair

I was also concerned about having the front and rear legs along the same line of the seat's wood grain.  So I made sure the rear legs were not directly behind the front legs.  However, I also needed to take into account the two back posts (outermost two spindles).  These were going to have their own mortise holes in the seat.  I don't show it here, but they ended up being almost in line with the front leg mortises.  C'est la vie.

I ended up moving the tops of the rear legs 1" forward and 3/8" towards center (might have been 1-3/16" and 1/2" - didn't document it properly).  The original rear legs had 14° rake and 6° splay and the new chair has 18.6° and 8.2°.

Next up: the chair build begins!

Child's Sunday School Chair - Part 1: The Original

Someone was giving away this chair, and I thought it could help me design and make my own version.  The chair had a note taped to the bottom of the seat indicating that it could be from 1930.

Child's Sunday school chair

Note on seat bottom

Front view (center spindle is broken)

Side view

Many of the joints were loose and it took only a little persuasion to get the chair apart.

The undercarriage

The legs are joined by a square (-ish) structure of stretchers, each about 11/16" diameter, with 5/8" tenons at the ends.  The front stretcher bulges in the middle to about 7/8" diameter.  The front and back stretchers are 5" to 5 1/2" down from the seat bottom and the side stretchers are about 2" lower than that.

The legs are 1 5/16" diameter in the joinery area, tapering at top to 3/4" tenons and at the bottom to about 1 1/16" feet.  I measured the rake and splay angles of the front and back legs using a bevel gauge and protractor.  The front legs have 2° rake and 4° splay, while back legs have 14° rake and 6° splay.  These were measured with the back feet propped up about 3/8" to make the seat level with the floor.

The back is formed from one bent-wood piece and three spindles.

The back

The arched piece is about 7/8" diameter, with 3/4" tenons at each end.  These had been fitted and wedged into through holes in the seat.

A large wedge in the arched rail tenon

At its maximum width, the arched back is 13 1/2" wide.  The top of the arch is about 15" above the seat.  Coming out of the seat, the arched piece angles back about 19°.  But you can see from the side view picture that it does not stay at that angle - the upper section curves further back after about 6"-8" of the straight lower section.

Arched back of chair, shows the backwards curve

The spindles fit into 1/2" mortises in the seat and 3/8" mortises in the arched piece.  The diameter of the spindles is about 1/2", and they bulge to about 11/16" about 4 1/2" up from the lower tenons.  The spindles angle back at 12-13° and the outer two also splayed to the sides about 3-4°.

The seat is about 1 1/16" thick, just under 13" wide and 12 1/4" front to back.  Grain direction is front to back.  The legs do not extend through the seat - the 3/4" diameter mortises on the underside are about 7/8" deep.  The underside of the seat is about 11" off the floor at the front and about 1/4" to 3/8" lower at the back.

The seat had a clean break front to back that had been "repaired" with two metal plates and screws.

Seat bottom

I drew the outline of the seat on paper and marked where the important features were.

Seat outline - both bottom (leg mortises) and top layouts shown (drawing not complete)

As far as materials are concerned, I'm guessing on most of it.  But I planed the surfaces of the seat where it had broken apart and it was a close-grained wood with a nice creamy light brown color.  Maybe cherry?

Seat pieces planed together in vise

I scraped some finish off a leg to see what it is made of.  Judging from the color and the ray flecks, it appears to be beech.

Leg material

The arched piece was of unknown wood.  As far as I know, oak and ash have been used historically for bending - maybe hickory, too.  I've never worked with ash or hickory, so I can't say for sure, but from what I've seen in pictures, this could be hickory.

Arched rail scraped to raw wood

The stretchers and spindles had a color similar to the legs, but I didn't see the typical ray flecks of beech, so I don't know what they were made of.

I'm making my own version of this chair.  Without wood bending capability I'll change the design to one with a crest rail.  I've been wanting to try some chair-making techniques that I've learned about (and have made some tools for).  I'll post about that next time.

Resultant and Sightline Angle Calculator

Finally I'm getting around to some chair making.  And to do that, I needed to work out resultant and sightline angles.  I won't bore you with definitions of these angles - you can learn about them on Schwarz's blog, or Curtis Buchanan's or Pete Galbert's sites (among others).

However I might bore you with some trigonometry.  The result of all this is that I made an Excel spreadsheet that will calculate resultant and sightline angles, given the rake and splay angles.  If anybody wants an Excel file with these calculations, send me a message using the "Contact Me" gadget of this blog and I'll be happy to send you a copy.

NOTE: in this discussion, I'm assuming the underside of the seat is level with the floor.

Here's a bit of the background mathematics:

Diagram showing how rake and splay relate to resultant and sightline angles

I'm going to treat a chair leg as a line with length, but no width or thickness.  In the diagram above, point E is on the underside of the seat where the leg originates and point A is on the floor directly below E (EA is a plumb line).  The horizontal and vertical axes are easy enough to represent in 2-dimensional space, but the third axis coming out of the page (toward the viewer) is along the segment AB.  Line segment EC can be the right front leg of a chair and we're looking at the chair from the front.  Hopefully it is clear that point C is where the leg touches the floor.

If I mark a line directly back to the horizontal axis from point C, that line intersects the horizontal axis at point D.  The angle formed by the plumb line and the segment ED is the splay angle, and it measures s°.

Splay of the front right leg is the angle off vertical when viewing from front

Similarly, if I mark a line from point C directly over to the third axis (along segment AB), it intersects that axis at point B.  The angle formed by the plumb line and segment EB is the rake angle, and it measures r°.

Rake of the front right leg is the angle off vertical when viewing from the side

You can also think of the rake and splay in inches.  In the case of the diagram above, the leg is splayed x inches to the right and raked y inches to the front.

The resultant angle, Θ, is a function of both rake and splay.  It is the angle the leg (segment EC) makes with the the plumb line EA and is the angle at which we need to bore the mortise hole in the seat.

The sightline angle, α, is also a function of rake and splay.  I find it hard to define in words, but I've read a couple different versions.  One goes like this.  It is the angle at which, as you rotate the chair and look at the leg, the leg looks perfectly vertical.

You can see the trig calculations at the upper right in the diagram.  These are easy enough to figure out (for this mathematically inclined dude), but writing them into an Excel spreadsheet can be challenging.  But that's what I did and here is the result.

Resultant and sightline angles as a function of rake and splay angles

To use the table, say you have a front leg with rake of 2° and splay of 4°.  Find the value for rake in the first column, go over to the column under the desired splay, and read off the resultant of 4.5° and sightline of 26.5°.

And in case I ever need to find resultant and sightline angles for rake and splay angles not represented in the table, the box at bottom (just left of center) allows you to put in any rake and splay.

On the right side box at the bottom, you can also find the rake and splay in inches (distance from where the plumb line hits the floor) by entering the height at the underside of the seat.  In this example, my small chair's seat bottom was 11" off the floor, resulting in a rake of 0.7" and splay of 1.3".

I know what you're saying.  "But Matt, what if I don't know the rake and splay angles?  Can I do something if I can measure the rake and splay in inches?"  Well yes, yes you can.  The following table shows rake and splay in inches, given the height of the bottom of the seat, as well as rake and splay angles in degrees.

Table of rake and splay in inches, given height of chair (H) and rake and splay angles

You've got to use this table in reverse.  Say you have a chair with height to bottom of the seat 11".  The bottom of the leg is 1" forward of where the plumb line hits the floor.  That's a 1" rake.  You also have the bottom of the leg out 2" laterally from where the plumb line hits the floor.  Splay is 2".  In the table, find 1" rake (black entries in the table) and you know you have a rake angle of 5°.  In that row of the table, find the red entry for 2" and that is between 10° and 11°.  Call it 10 1/2°.

I realize that I could make a table with rake and splay in inches along the top row and left column so you can look up the rake and splay in degrees, but I was losing energy.

You can do a lot with this.  I was copying a chair and I liked the spread of the legs at floor level.  But in the original I thought the mortise in the seat was a little too close to the edge.  So I played with the design, moving the top of the leg away from the edge of the seat while keeping the bottom of the leg unmoved.  Based on how much I moved the top of the leg, I was able to find the new rake and splay angles, and ultimately the new resultant and sightline angles.

It's fun stuff.  Again, if anybody would like a copy of this spreadsheet, send me a message using the "contact me" gadget on this blog.