Results 1 to 1 of 1

Thread: Turning spheres part 7: Turning on third axis, finishing cuts, troubleshooting

  1. #1
    Join Date
    May 2005
    Location
    Northern MN
    Posts
    389

    Turning spheres part 7: Turning on third axis, finishing cuts, troubleshooting

    Before I get to turning on the third axis, I took a couple pictures that better show the tool orientations I use.

    This first pic shows cutting away the initial ghost, which is large because of all the material left on the ends where the piece was held by the steb centers. As I said before, I use an unsupported pull cut using just the tip. Since the bevel is not engaged, you're freehanding the cut. But I do it this way because using just the tip of the gouge I can remove wood with only a small amount of cutting edge, and the wood comes off pretty easily because it's mostly a peeling cut parallel to the grain. My cut is not in a smooth arc, in part because I don't have the bevel to guide me, but it doesn't matter because I can clearly see where my cut is in the ghost, and I can also see that I'm staying plenty clear of the sphere within.

    Capture106.JPG

    Once most of the ghost is removed and there more wood and less air involved, then I switch to the steep angle shearing cut, shown in the next picture. I showed a picture of this before, but the angle of the tool was too steep. This tool orientation does a nice job of slicing across the end grain.

    Capture107.JPG

    So then, picking up after completion of turning on the second axis. . .

    Rotating to the third axis: At this point, things are probably looking pretty spherical, but there will likely be a lump near the areas of the sphere that were held in the compression chuck in the last step. This lump is to be expected, and is removed in this next step. Remount the sphere in the compression chucks with the intersection of the first and second equators centered in the cups as shown in the picture.

    Capture108.JPG

    Do the twisting and adjusting again to get the workpiece settled in properly centered. When you turn the lathe on this time, you should see only a very limited ghost image, and it should be confined primarily to the middle quarter or third of the sphere, along what will become the third equator in this step (or the 90th meridian for the earth analogy). If you see a lot of wobble when you remount this time, especially away from the middle, best to check the seating of the workpiece. Depending on how things went in the previous two steps, having a fair amount of wobble at this point can still happen, but if there’s a lot of wobble in this or any subsequent rotation, my first suspicion is always that the workpiece isn’t seated correctly (or that I was lazy when I started and didn’t true the compression chucks before using them and they aren’t true).

    Repeat the turning process as for step two. In this orientation, the fibers of the workpiece are again perpendicular to the ways in a “bowl like” orientation, so the downhill cut is from small diameter to large, as in the last step. Accordingly, I use the same gouge (although touching up the sharpness is a good idea, you’re getting close to finish cuts here) and same cutting orientation. Most of the out of roundness should be toward the largest diameter, so be careful when cutting close to the chucks that you don’t cut too deep. Usually the piece isn't too lumpy at this point so I can go straight to the shear cut with bevel support. Check your progress against the two equators (the first one, the single line, and the second, the double line). It is not unusual that you may cut one or more away bringing the sphere to round, most likely the first equator. Don’t worry about this unless you are cutting dramatically more on one part of the sphere than another; that would indicate something is not aligned well, check the seating of the sphere in the chuck. Also check to see progress of the cut to the equator along the full path to see if you need to cut more in one area or another; try to bring it up evenly across the whole workpiece. Mark the third equator with three lines. Note in the picture there are a couple small uncut areas, one an overdeep cut from turning on the second axis, and a small flat spot near the first equator (single line). In this example I was roughing out a sphere from green wood that was to be dried and re-turned later so I didn't bother to take this out, as I would if it were a dry sphere to be finished. More on this issue later.

    Capture109.JPG


    Are we there yet?: Maybe. When the sphere is done depends on several things, including what your tolerances are, and how close to true you were when you started. If all cuts were close to perfect, and if the workpiece was perfectly centered at each rotation, then in theory it is now a sphere. However, if when turning the second axis (first turning in compression chucks) you had to cut away much of the equator on one or both sides, then the final sphere will be a little smaller than the diameter you had at the start of the layout process, and you will need to rotate the sphere again to bring the whole sphere down to it’s new diameter. The simplest test is to rotate the sphere in the compression chucks again, this time so the second and third equators cross under the cups and the first equator is in the middle between (this is the same orientation as when the workpiece was roughed out). Taking care to see that the sphere is seated properly, rotate it and see if it is running true. For me, often there is still a little irregularity and so I true it up again in this orientation, keeping in mind that it’s now back to a spindle orientation (endgrain facing the compression chucks), so the downhill cut is back to being from large diameter to small.

    You can continue rotating the piece through the three axes and truing it up until you’re happy with the result. Nothing is perfect, there is only, as my wife says, “perfect enough.” Even if you make a truly perfect sphere at this point, it will come out of round slightly with seasonal changes in moisture content. Sanding will not remove many, if any, deviations from spherical, but it can make them less noticeable when the sphere is handled. But if there’s a small ridge, sanding will blunt it and probably make it not detectable by the hand.

    Good finishing cuts are better than sanding. You don’t want to have to sand spheres aggressively. If you have to take off a significant layer of wood to get below some tearout, you either have to take the entire surface of the sphere down by the same amount, or oversand the trouble spots and risk making a noticeably out of round spot. In addition, particularly on softer woods, the face grain is generally softer than the end grain, with the side grain generally in between. If you do a lot of sanding to remove tool marks it’s pretty easy to create a flat spot on the end grain. It’s a far better thing to leave a good surface from your tool. Your tool will cut in a circle even across varying density, and will leave a truer surface to sand.

    For me, the best surface I can produce on the lathe is usually with a “shear-scrape” cut with a bowl gouge. As shown in the picture, this involves turning the gouge over with the flute facing the work, then dragging the far wing on the work at a steep angle. This was a piece of green hickory that did not like to cut cleanly across the end grain. I think it will cut better dry.

    Capture110.JPG

    While this is very effective at leaving a high quality surface, you don’t want to use it until you’re nearly done and just removing a tiny bit of wood. In my experience, it’s a bad idea to use it when the work is not round. The reason is that you have the knife edge of the gouge pointed straight into the wood. While that’s part of what makes such a nice cut, if there’s any “bounce” in the tool, the bounce will drive that knife edge into the workpiece and cut a groove. Cutting a groove with the shearing cut with bevel support (described previously) is less likely I think, because if the tool is bouncing a little, some of that is taken up by impact on the bevel, and the tool is less likely to cut a groove when it impacts the workpiece. So best to wait to use the shear scrape until the piece is round or very nearly so. And, like most any other lathe cut, it will work best if you sharpen immediately before. Others may wish to use a negative rake scraper. Just speaking for myself, I seem to get a better finish from the shear scrape with the bowl gouge, but those able to put a nice burr on a NRS may get as good or better results.

    Troubleshooting: I’ve already mentioned several times the importance of making sure the workpiece is centered in the compression chucks, as being off center is probably the most common thing to interfere with coming to spherical and therefore require many iterations to complete the sphere. Something I haven’t mentioned yet is if you get a deeper-than-it-should-have-been cut somewhere. I have a couple pictures below from a sphere where at the octagon step, one of the facets that should have been dead straight ended up cut a little concave, and as a result the middle of that flat was too low. When the piece was rounded after rotating to the second axis, when I was approaching the equator I still had an uncut area near what was the middle of one of the 45 degree facets. As I got to the equator, I still had a small “trough” that was uncut between the “pole” and the equator. When this happens, you have three choices: 1) live with it (not necessary to); 2) sand it out later (not recommended for reasons discussed above); or cut it out, and in doing so slightly reduce the final diameter of the sphere. I generally choose that last; the reduction in diameter is usually really, really small, and a nicely turned though slightly smaller sphere is usually more desirable than a slightly larger but blemished sphere. But everyone will make that decision for themselves. If you go ahead and cut it out as I did in this example, you will need to do at least two more rotations after this one to bring the whole sphere down to its new diameter (but again, this is a very slight reduction). The right panel in the pic shows where I've cut out the defect on the right side, but not fully on the left yet.

    Capture111.jpg

    Especially after 2 rotations, residual wobble should be very slight, and probably only detectable by touch, rather than by eye. If you are seeing a visible wobble after 2 rotations or more, it is very likely the workpiece isn’t seated properly or that your compression chucks need to be trued up.

    Next time, sanding, finishing, twice turned spheres. . .
    Last edited by Dave Mount; 06-02-2021 at 10:00 AM. Reason: typos

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •