Friday, September 29, 2017

Drawing An Ellipse

I'm in the process of making a few Sellers-inspired step stools and I'm making the tops elliptical.  I once read about a method of laying out an ellipse in Dennis Laney's excellent blog, so I thought I'd try it out.  Dennis outlines a few different methods - this is just one of them.  This method uses a set of trammel points, and I had just made myself a set a few months ago.

I started by drawing the major and minor axes on poster paper.  My ellipse was to have major diameter 15" and minor diameter 9".
Axes drawn longer than size of ellipse
I set the two trammel points to a distance of 4 1/2" and 7 1/2" from the pencil - these are half the ellipse diameters.
Setting the trammel pencil and the two points
Then got a piece of scrap with one square corner and laid that in one quadrant of the drawn axes.  I used some plywood that was thicker than my trammel pins are long.  That way, the edge of the trammel from which the pin protrudes (the wooden part) can ride on the edge of the board and the points won't dig into the poster paper.
Using a squared-up scrap to guide the trammel points
Then, while holding the scrap board down firmly, I run the two trammel points along the two squared edges of the scrap board (both trammel points must remain in contact with the board at all times), while dragging the pencil trammel on the paper.
Starting the ellipse - pencil point is right-most
Sliding the trammel points along the board, keeping both in contact with an edge
Finishing the first quarter of the ellipse
Then it's a simple matter of moving the scrap piece to the different quadrants and repeating that process.
Working the next section of ellipse ...
... and the third quadrant ...
... and the fourth
And here is the result - a very pleasing shape.
Completed ellipse
I'll use this as a template for laying out the tops of the stools.  More on that later.

Friday, September 22, 2017

The Tale of Goldilocks and the Three Rabbets

"My rabbet moves further into the vertical wall with every plane pass," said daddy woodworker.
"My rabbet gets further away from the vertical wall with every pass, creating steps," said mommy woodworker.
"My rabbet is juuuust right!" said baby woodworker.

I've been doing some practice work with mouldings.  And when you use hand tools, that means cutting lots of rabbets.  When using a rabbet plane or a shoulder plane to cut rabbets, the iron should be just a little proud of the right side of the plane.  But how much is enough / too much?

Here is my shoulder plane.
The shoulder plane
The iron is proud of the right side by about 0.004".  This is not easy to measure, so I used a piece of 0.0025" thick paper to guestimate the amount.  The iron was more proud than one thickness of the paper, but just less than two thicknesses of the paper.
View from the plane's sole showing iron just proud of side
I used a wheel-type marking gauge to mark some poplar for a 1/8" wide x 1/4" deep rabbet and started the cut by setting the edge of the shoulder plane in the gauge line.
Starting the rabbet with plane edge riding in gauge line
Creates a little tiny shaving on which I can see the pencil mark I added to the gauge line
Right away it seems like it's cutting into the keeper wood of the vertical wall.  After two or three passes like that I was already into the vertical wall by about a millimeter!!!!  Holy crap, that went south fast!
Arrow points to the "new" vertical wall location
I continued with the plane held vertically and the plane cut fairly vertically - that is, I didn't keep moving the wall to the right.
Abandoned the first rabbet and gauged another line for 1/4" x 1/4" rabbet
For the next experiment, I nudged the iron to about a 0.002" proud of the side and started the rabbet as I did before by tilting the plane in the gauge line.  This time I did not move the gauge line very much at all, but it did move a little bit.
Stayed very close to the vertical line
Looked very good at the far end, too
I continued the rabbet to completion and I needed to clean up the vertical wall a little bit.
Vertical wall needed a little bit of cleaning (with shoulder plane on its side)
Well, this looked very promising, so I did another test with a board marked for a 3/16" x 3/16" rabbet.
This came out very good with little or no clean-up needed
The far end looked good, too
I have a Record rabbet plane and wanted to try using it for these experiments, sans fence and depth stop.
The Record #778
So I set its iron about 0.002" proud of the right side and gauged another 3/16" x 3/16" rabbet.  I angled the plane in the gauge line like I did with the shoulder plane, but got a very different result!
(Ignore the chipped corner)  The first pass got me way past the gauge line!
After a couple passes I was not even in the ball park.
I think the reason for this is that the Record rabbet plane has a significant chamfer on the edges of the sole, so it doesn't ride well in the gauge line.
Edge by my thumb nail is chamfered
So for the next test run I started the rabbet with the shoulder plane with its iron 0.002" proud.
Started rabbet with the shoulder plane ...
... and finished it with the Record to remove material more quickly
Nice square rabbet, but went past my vertical gauge line at the beginning ...
... and at the far end
My final experiment was not really an experiment, but a different technique altogether.  I used the shoulder plane to start the rabbet away from the gauge line, using my left hand fingers under the sole as a fence to keep the plane just a little away from the gauge line.
Plane held vertically, left thumb on top, left fingers under sole acting as a fence
This is easier than it sounds.  After a few passes, I was deep enough ...
About 1/32" deep
... to switch to the #778 for faster (and more comfortable) material removal.
After I got to the depth line, ...
... I cleaned up the vertical wall with a few passes of the shoulder plane
It's very easy to sneak up to the exact vertical gauge line that way ...
... and get a perfect rabbet
Matt Bickford espouses this last method.  It's easy to get it perfect after practicing once or twice.  On wider rabbets it can be tougher to start the plane, but you can still get a perfect rabbet.

The moral of the story is that if I want to cut rabbets, having the iron about 0.002" proud of the side is probably about the right amount.  But I can remove all doubt by rabbeting a little away from my vertical gauge line and cleaning up the vertical wall with a few strokes with the plane on it's side.

And Goldilocks said, "What nice rabbets you are!"  And they all lived happily ever after.

Friday, September 15, 2017

Some Practice With Moulding Profiles - Part 2

In the previous post I showed two profiles I cut into 1 1/2" x 1" x 15" long stock; a cove with fillets and an ovolo with fillets and .
Less than 90° cove on left, 90° ovolo on right
Buoyed with confidence, I tried some more complex profiles.  Having never made mouldings before, I struggled with the layout.  I thought there would be a more geometric way to do it, and maybe there is, but in the end I just kind of winged it with a grid and a circle template.
A 5/8" radius ovolo and 1/2" radius cove, separated by a fillet
You can see some diagonal lines from my failed attempt to find locations to use a compass to draw the ovolo and cove.

Anyway, I used this template to transfer lines to the end of my workpiece.
Found it helpful to write the number of the plane I'll use to make the convex and concave shapes.
The first rabbet is marked at the location of the fillet separating the two shapes.
First rabbet planed away (red X denotes an incorrect layout line for the ovolo - try to ignore it)
Second rabbet marked out - it will provide "runners" for the round plane to ride on
Setting a marking gauge to the vertical line.  I'll mark the ends and top of this rabbet,
then set the gauge with the horizontal line of this rabbet and mark the ends and right side.
Second rabbet cut
In the above picture, you may also notice that I removed 1/8" from the right side.  As this was laid out, the cove ended in a 1/8" flat.  I thought when I was laying this out that there would be a fillet of some kind there.  Rookie mistake.
Chamfering the ovolo area in preparation for the hollow plane
Ovolo cut ...
... and cove cut
Here's one important thing.  When the extent of an ovolo or a cove does not end in a fillet, I penciled in the extent of the profile on the surface or side of the workpiece.  You don't want marking gauge lines in these places as they would need to be removed and might affect the profile.
The completed profile (you can see remnants of my pencil lines denoting the extents of the cove)
The last profile that I practiced with is an ogee shape.  It's similar to the one above, but without the fillet between the ovolo and the cove, and with the addition of a top fillet.
Profile on the right, showing the order of the 4 rabbets
(there's a rabbet no. 1.5 because I forgot to write that one in at first)
Working on the first rabbet - it's 1 1/4" wide, so I got out the Record #778
Cleaning up the vertical wall of the second rabbet
Messed up here - cut the thin right-most rabbet before cutting the rabbet for the cove.
This was a problem because I hadn't gauged the rabbet for the cove yet and I removed a reference surface.
the resulting profile came out not perfect, but pretty good
It takes some practice to get the transition right between the ovolo and cove.  And with my H&R planes not being perfect, it's even more challenging.  But this was a really fun exercise.  Here are the four profiles I made.
Nice line-up
This was just a practice exercise - I don't have any current projects that require mouldings.  But when I do, I think I'll be able to make them by hand.

My challenge is that I don't really know how to design mouldings and don't have a feel for what might look good on a particular project.  Anybody know any good books (or other resources) that talk about this?

Until next time ...