The hexadecagon alternative: At the end of the last installment, the sphere was roughed out except at the ends where it was against the head and tail stocks. Before I move on, I wanted to circle back to something I said earlier about going from an octagon to a hexadecagon (16-sided). This variation takes off more of the waste wood using straight lines, to better show the proportions of the sphere and thereby relying less on achieving spherical by eye. I've always just gone to octagonal and then from there to spherical, but I decided since I mentioned it I should try the hexadecagon step. I did that last night; it turned out it was pretty quick and easy to do, and it might be something some folks might want to use, especially early on.
This step picks up after getting the blank to octagonal. The facets of a hexadecagon are 0.199 times the diameter. I set my calipers for that distance, then centered the calipers over the arris where the center section of the octagon meets the each of the 45 degree facets. I just did this by eye, then made pencil ticks on each side, and extended those around the blank (red dashed lines in diagram).
Capture97.JPG
Then you just cut away material to connect those lines with a straight facet, which is pretty quick now that most of the waste has been removed. Once the new facets are in, mark the midpoint of the new facets. Here's a picture after cutting the new facets and marking the midpoints.
Capture96.jpg
The final step is to round off the points to connect the lines with a smooth curve without cutting the lines away. I used a spindle gouge, but after doing that, I was thinking this might be a good time to use a negative rake scraper. There's very little wood to take off and a scraper is a pretty easy way to "rub off" extra wood without accidentally digging in and cutting a groove. Bottom line, I might have dismissed this approach prematurely. It's probably an easy way to help get the shape right if going straight from the octagon to spherical is challenging.
Rotating to the second axis: The next step is to turn the piece round in a second axis. The reason this works is another trick of geometry. Think of the pencil centerline from the last step as a circle you can hold, like a hula hoop. Imagine grabbing that hula hoop with each hand at opposite sides of the circle and holding it in front of you. Now, twirl that hula hoop so the top comes toward you and your hands stay in the same place. That path of that rotating circle describes a perfect sphere. So if we turn our “sphere-to-be” so its equator is spinning 90 degrees to the original axis, then cut away everything that lies outside our pencil line, we get a sphere. That’s the next step.
Replace the drive and live centers with compression chucks as shown in the next picture, with the equator (centerline) crosses through the centerline of the lathe under the compression chucks. It’s important to get the piece really centered in the cups; I find it helps to bring the tailstock up and squeeze the workpiece lightly, then rotate the workpiece back and forth against the cups to help if find center. It is common to end up with the area along the equator from the last step being a little flatter than it should be. If it is, the workpiece won’t completely settle in the compression chucks because it’s not contacting the cup all the way around. This is common; just do the best you can to get it centered. Even if it ends up off, the irregularity can be cured in successive rotations.
Capture98.jpg
There’s a little more subtlety to grain orientation might be immediately apparent and this becomes important as the process goes on. To keep track of things as you turn the sphere on different axes, it’s helpful to recognize the different grain, which I’ll call face, edge, and end grain. Face grain is what you see on the face of a flatsawn board. Edge grain would be on the edge of a flatsawn board (or the face of a quartersawn board). End grain is, well, end grain. In the picture below I show a sphere with the grains labeled. On a sphere, face grain is usually easy to see, it’s where the “circles” are. The edge and end grains can look similar as they’re both usually parallel grain lines. If there are visible rays in the wood (small but visible in this piece of mulberry), the edge grain is where they are cut in tangent and show most prominently. The end grain is usually darker looking; another clue for the end grain is that the end grain was on the turning axis of the lathe at the first step, so the equator from the original turning alternates between face and edge grain, so the pencil line from the first equator doesn’t go through end grain (this doesn't help you when sanding though). On this piece of mulberry the distinctions seem quite clear, but for some woods it can take a little looking to tell the difference.
Capture99.JPG
For the first rotation, it probably isn’t that important which orientation you choose. As a matter of habit, I make the first rotation of the sphere such that the face grain is facing the chucks (the edge grain will go there in the next rotation).
A couple words about the compression chucks. You can make these that fit in a chuck, you can mount them to a faceplate, or you can thread them. The faceplate pretty much requires that you dedicate one to the purpose, because it would be hard to get the wood block remounted in exactly the same place if you took it off. If you’re only turning one or two spheres, that’s not a big problem. I bought a 1” by 8 tpi tap and have liked those, but I haven’t been through the annual humidity cycle with them yet, so I don’t know if humidity-related changes in the threads will change my mind about that. For blocks that go in a scroll chuck, I mark then so the teeth of the chuck jaws always to in the same place, in hopes of them being as true as possible when I remove and remount them.
My compression chucks are not hollowed to fit the sphere – the sphere just registers on the ring of wood at the front. Because the workpiece is changing shape and diameter as you finish it, I don’t think it would be worth trying to fit the whole hollow to the sphere. The size of the “ring” that contacts the sphere depends on the size sphere you are turning. I have two sets I use most often. The larger has a contact surface that is about 1.25” in diameter and that works well for spheres from around 2.5” to 3.5” in diameter. The smaller ones have a contact surface about 7/8” in diameter and these work down to about 1.75” diameter. It’s a good idea to retrue your compression chucks every time, unless you *just* used them. An out of roundness in your compression chucks will keep you from getting a perfect sphere. Almost any time I’ve had a struggle getting to spherical, it’s because my chuck was out of round.
When I first started and didn’t have the threaded live center, I made a compression chuck with a hole in it that would fit over my conical live center (picture). Pay no attention to the step in the cone of the chuck, I cut that in there to hold something else unrelated to spheres.
Capture100.JPG
This worked, but not great. The problem with it is that it creates another hinge point in the chucking assembly, so another place for something to get off center. I think it could be improved by doing more shaping of the recess so that it is taper fit to the conical live center; I just drilled a hole and then eased the corners of the hole. I never went further with it because I got the threaded live center.
With the rotated sphere mounted and the lathe running, you will see your sphere, with a ghost image of the excess wood to be removed around it. The next series of cuts will be to remove the ghost image. Although this cut looks like a normal spindle cut, it is not. In normal spindle turning the end grain is facing the head- and tailstock, and so the wood fibers are running parallel to the ways. That’s why cutting from the larger diameter to a smaller diameter is cutting downhill with regard to the wood grain. When you rotate the sphere 90 degrees and put the face grain against the cups, the work is oriented more like a typical bowl. The wood fibers are running up and down or side to side as they rotate, perpendicular to the ways not parallel. As a result, using a spindle gouge and cutting from large diameter to small is actually cutting against the grain. The cut with the grain is from small diameter to large. For this reason, I switch at this step from a spindle gouge to a bowl gouge, and my shaping cuts are more like what you’d do on the outside of a bowl.
Capture101.JPG
I think I'm at my picture limit. . .I guess I'll break here and finish this step in another post.
Reply With Quote
