>I've had to turn several platters and bunch of mirrors lately and I find getting a flat gently curve on the bottom of the platters or the outer softly curved face of the mirrors is really difficult. You think you have it perfect and then you put the finish on and you can see very shallow ripple somewhere, usually the middle. I even stop the lathe and sand with the grain which you would think would get rid of any circular uneveness. I'm going to try a different lighting arrangement and look more closely. Granted most people would barely see it unless the light is just right and probably wouldn't care either, it's just a perfectionist sort of thing.
>I always have this issue. I use an engineers square to check the flatness. I make sure the bottom is flat before I get to the 220 and finer grits.
I use belt sanders (Porter Cable has a small belt sander). 5" ROS, 3 1/4" ROS, and cork sanding blocks to achieve a flat surface in the body of a platter.
The chuck appears to impact the "ring" position. I have gone to the #6 Oneway jaws which can push the ring out to 8" where I find it easier to control the "ring" effect. I notice that Vicmark has a set of jaws which can be used in compression at 10"+. For me that would put the "ring" of a 16" platter at the slope of the body for the platter, which would eliminate the "ring" effect. My best success comes from using 10" glue blocks for mounting. But a glue block is extra effort so I rarely use a glue block.
All this to say that my experience of the "ring" effect is related to mounting method more than cutting technique or sanding (though improved cutting and sanding technique certainly reduces the effect). I have also noticed that thicker surfaces, such as a table top, shows no "ring" effect.
>Interesting. I don't believe the chuck has anything to do with the problem I'm having. It's really more a problem of the cutting and shear scraping causing a very slightly unflat surface. I will try a straight edge but since my pieces all have a gentle curve I don't know if it will show. These are very slight depressions that only show up when I get light reflecting off the glossy finish. I'll try to take a photo tonight to show what I'm talking about.
>I have this trouble integrating long flat curves, too. I get around it by using a large, very sharp, gently curved scraper held flat on the rest for the final light passes. Any remaining unevenness can generally be felt with the fingertips. I take care of that with my Milwaukee offset drill with a 3" medium-durometer sanding head, making sure to start with a grit that is coarse enough to do any required leveling. If you start too fine, it just gives you smoother ridges. :-)
>The problem is that a wave only has to be something like 0.003" before we can see it, and 0.005" before we can feel it, and the glossier the surface, the easier it is to see.
A trick is to cheat on a platter or flat plate, by making the center slightly higher than the outside edges of the surface. Yes, flat is best, but a slight dome in the center won't accent the small irregularities in the surface like a flat or slightly dished surface.
I always use the tool rest as a reference and set it somewhere between flat and no more than 1/16" farther away in the center on a very large platter. That the tool rest has to be straight tool rest is a given. NHot all of them are. After that, there is a pecking order of cuts to make a flat surface. Some get there the 1st try, while others take awhile, and some will never get there.
The prefered tool is the scraper that Ellis described. The direction of the cut is important. It is for me anyway. With the tool angled INTO the wood, it is easier to get a flat cut by pushing the scraper into the wave from the outside. Make sure the scraper is always moved with a whole-body motion, using the tool rest as a reference, and NOT with an arm motion.
If you were to make what I call the "natural scraper cut" by pulling the tool across the wood starting from the center, the tool will follow the wave and only move it, rather than remove it.
A common problem is the little dimple in the center. The solution is to hold the scraper slightly BELOW center, and don't stop or hesitate when you get to the center. Keep moving the tool and don't remove it from the wood until you have passed under the center. You will have a little nubbin to remove, but there will not be a dimple.
Doing the same thing will always help when you start from the center as well. Start the tool on the other side of and slightly below center. That way it will start cutting as you cross under the center of the piece and there is no dimple because there was no hesitation from starting the cut in the center.
Another good tool for making flat surfaces is a sharp skew chisel with a curved cutting edge. Hold it flat on the tool rest and roll the curved edge across the wave. You should be able to see the wave as it is being cut. This tool gets dull fast. No more than 30 seconds and you will start seeing tear-out and it needs sharpening, and the sharper the better.
You can make a tool-rest follower that looks like a little T-Square. The head will follow the tool rest, and the tool is held along the blade of the "T". Making a groove or a step in the blade will reference the tool. This can be used with either a gouge or a scraper.
As a last resort, I have a 6" x 80-grit disc sander that does an excellent job at removing waves.
Get one of the longer ones, 6-8" long. Back light the surface, bend the stick to it and look for where light shines through. Great for checking longer gradual curves, inside or out.
>John-I found this in one of those shop drawers where things go to get lost. Don't remember buying or ever using it for its intended purpose. I have found it useful for detecting those circular ripples that you describe. I just push it straight into the bottom of a bowl/platter/whatever across the center and then look at it with a solid background behind it. It will show very small defects.
>Bill I have 3 of those plus a plastic one. :) They won't show the shallow ripples I'm talking about. These can't be felt when you sand them. My last passes with the sandpaper are with the lathe off and sanding with the grain with each grit. These are really subtle ripples that don't show up until the finishing stage.
I'm going to do another mirror tonight. I did really well on my Oak mirror last night with just the slightest ripple toward the outside,none in the middle. We'll see how it goes. I may try Russ's idea of the skew sideways although some of my woods tear out easily and I prefer to shear scrape them. I'll let everyone know how it goes.
>This is hard to do. I measured it with a dial indicator and the from the dip to the rise in the center is only .010". I can't believe how noticeable that small amount is. It's very hard to photograph because you have to move your eye around to see it. If you hold still it's barely noticeable.
>John, I too have had the ring effect that rick mentioned and I too believe that it is due to relaxation of the stress caused by the chuck. I recently turned a 20-inch flat-bottomed walnut tray that was dead flat until I dismounted it from the chuck and trued up the base. After finishing, it has a slight ring when viewed by reflected light. I use a rectangular cabinet scraper as a straight edge for checking flatness and surprisingly, for consistency of a curved concave surface. If you place the straight edge on the curve, it provides a reference against which your eye can detect minute aberrations from a smooth curve. It also works well for convex surfaces, allowing you to find flat spots and variations of curvature that are not apparent otherwise.
>I was having a similar problem and tried shear scraping with all my tools with no luck. Finally purchased one of the carbide cutters from Easy Rougher folks which has a 4 inch radius (rather than one of the "straight" ones)--mounted on 1/2 inch square stock so that it would be presented to wood at 45 degree angle and used it to shear scrape the piece. Worked very well for me.
>Somehow I thought we were discussing a concave curve not a convex one. For this I would try a card scraper with the lathe running. Then I would sand it with a ROS with fresh sharp paper with the lathe off or running very slow. (100 rpm or less)
Another approach might be to make a caul for the sand paper that has the profile in it and use that to sand.