PDA

View Full Version : Visual Calipers



joe greiner
16th August 2009, 09:13 PM
This is my interpretation and adaptation of DJ DeLorie's parallax jig ( http://www.delorie.com/wood/tips/parallax-jig.html ). Although the physical phenomenon of parallax is the principle of operation, "visual calipers" seems a more appropriate name for its function. "Parallax" also implies some obscure high-tech magic, and the scary movie "The Parallax View" of 1974 hasn't helped at all.

In colloquial use, "visual calipers" generally means eyeball measurement, usually with precision and accuracy of about 10%. This tool is a lot better than that.

The general idea is to have two copies of the pattern, with one copy cut along the profile, and both of them offset from the axis of the lathe, as well as from each other. Then cut the workpiece until corresponding points are all in alignment from the turner's point of view. The advantage of this technique is that it enables "downhill" slicing of the wood fibers for a better finish. It also allows minute variations characteristic of manual cutting.

Most copying contrivances merely scrape to the required shape, and substantial sanding is needed. In fact, for some really mass production (such as baseball bats), the scraper may be the entire profile, pressed into service in a single motion. This works well enough when the medium is plaster of Paris, scraped while it hardens and expands; when the medium is wood, fine details can be lost to grain tear-out.

I made my version with an adjustable angle, and variable pattern separation, for maximum flexibility in further studies. It seems intuitively obvious, that greater pattern separation will provide more precise results. And of course, the near pattern must be far enough from the lathe axis to accommodate the rough blank. The far pattern could be farther yet, but depth of focus could be an issue.

I made this from materials on hand: surplus bed frame angle irons, a single stick of furring strip, and some orphan signs in extruded plastic; thin plywood or foamboard would also be suitable for the pattern supports. The threaded fastenings provide registration of the two patterns. The notch in the near pattern accommodates the tool rest support (aka "banjo"). The lock-down nuts are DIY weldments, tapped to suit. The angle irons can be inboard or outboard, depending on workspace constraints.

Very bright lighting provides better results. Further studies may explore the benefits, if any, of using contrasting colors in different parts of the patterns. Ideally, the tool will dynamically mimic the fit and gaps between the work in progress and a stationary template.

Cheers,
Joe

joe greiner
16th August 2009, 09:15 PM
These little vases are the first turnings using my Visual Calipers. They're replicas of small brass urns turned freehand by my uncle, Sgt. Richard Wylie, in the late 1940s. He died in 1950, and I've had to guess at a few historical details.

Sgt. Wylie worked in the office of Graves Registration after WW2. Thanks to his (and others') efforts, relatives of fallen servicemen and women can identify the locations of their graves in Europe. I suppose he dealt with the undertaking industry, and derived inspiration from crematory urns. Remarkably, the small urns measure 1" diameter x 1.6" high - very close to the "golden ratio" of 1.618. I don't know if Uncle Dick had any art training; I thought his profession was accounting. And he probably made them at an Army craft shop, or perhaps in the Pentagon's machine shop. I don't know if the Pentagon actually has a machine shop, but I'd be surprised if it doesn't.

At any rate, I measured the small urns carefully, and made a drawing in AutoCAD. I refined some of the measurements, and enlarged to a workable size. My first target prototype size needed Pine blanks that would have needed a lot of preliminary work. So I measured some smaller logs of Chinaberry, and reduced the scale of the drawing; printed two copies for the patterns, and proceeded to turn both copies over a period of about 8 hours maximum in two days. My reason for urgency was an impending meeting of our WT club at the shop of a retired Air Force General. The military connection was too good to resist.

Chinaberry, 2 5/8" x 4 1/2" (67mm x 115mm), EEE + paste wax, with paintball tubes for flowers.

Cheers,
Joe

hughie
17th August 2009, 12:51 AM
This reminds me of the old shadow profile machines we used to use in profiling existing shapes in order to duplicate them

Ad de Crom
17th August 2009, 01:27 AM
Joe, neat idea.
Isn't it the same as copying?
Sometimes I made a autocad drawing and scale it up, and than I made a mould of carton to make the same object on a larger scale.
Did that a time ago to make two wall lamps, a smaller and a bigger one for friends.
Ad

wheelinround
17th August 2009, 10:22 AM
Nice jig Joe now you just need that tool holder some cutters and away you go

joe greiner
17th August 2009, 08:45 PM
I guess so, hughie. Can you provide a reference or link to the shadow profile machine? I suppose an industrial application would have solved the problem of collimated light; the best I've figured out would be to use the Sun for substantially parallel light rays. With this tool, the radial lines help the turner to align his/her eyesight.

Righto, Ad. The CAD was just for convenience and precision. Freehand or mechanical drawing works just as well, with photocopies for the patterns. And some copiers can adjust scale too.

No dedicated cutters, Ray. Just parting tool, detail gouge, Oland tool, etc., depending on the location of the cut. It preserves the manual aspect of turning.

I recently read Richard Raffan's book on woodturning projects. He mentions turning about 2000 copies of chess pieces, then filtering them into groups of best matches for assembly of sets. I used the same procedure for my birdhouse earrings a couple years ago. ( http://www.woodworkforums.com/showthread.php?t=57866 ) In one of my previous lives, I occasionally used the title of Industrial Engineer, so I guess I knew the post-production trick. But it was nice to read his confirmation.

Cheers,
Joe

hughie
19th August 2009, 10:08 AM
Here you go Joe :2tsup:



http://www.akribis.co.uk/adverts/baty_R550_w.html

http://www.nikoninstruments.com/Information-Center/Optical-comparator-shadowgraph


http://www.alibaba.com/product-tp/103241806/Profile_Projector.html

joe greiner
19th August 2009, 09:31 PM
Thanks for the links, hughie. There's obviously a need for such devices.

Alas (or fortunately), I don't have enough real estate in my sliding-block puzzle of a shop. I reckon a tripod-mounted apparatus could use the Sun; 150,000,000 km distance should be parallel enough, especially for very infrequent use. Just what I need: another entry to the bucket list.:rolleyes:

Cheers,
Joe

Paul39
20th August 2009, 02:09 AM
Joe,

A slide or movie projector will work as a shadowgraph type light source. If you don't have one, check Goodwill and junk stores.

artme
20th August 2009, 03:06 AM
Hmmmmm. Methinks a good idea, but forsee some problems.

joe greiner
20th August 2009, 09:57 PM
G'day, Paul and artme.

For most accurate profiling, a conventional point light source, even from a slide projector, will suffer from "fanning" of the light rays, i.e. a conical distribution around the shape. Hence the high price of professional apparatus. The distortion could probably be corrected in software, or by brute force in CAD, with a couple reference points on the original. But simpler yet could be a tripod-mounted affair consisting of the original shape, a matte-finish screen, and a camera. Point the whole thing at the Sun for exposure, and Bob's your uncle. Enlarge the image as needed.

It doesn't work straight out of the box, Arthur. I first turn reference tenons at each end of the blank, using regular calipers. Then use those tenons to adjust the patterns into alignment. The tenons must also be long enough to reach the workspace, after allowance for the headstock and tailstock. Another hiccup could result from very long patterns due to flexibility; stiffeners should help there. And the cross sections applicable can't exploit the full capacity of the lathe, without substantial modifications. Please identify any other problems you see. I love to solve puzzles.

Cheers,
Joe

Paul39
21st August 2009, 03:46 AM
Joe,

Since you like to solve puzzles, here is something:

You do get fanning from the point source, which is how photo enlargers work. The fanning is the same in all directions.

Optical comparators are nothing but a point light source with the path folded in front and behind the object to be profiled. The expense comes with calibration and ability to measure.

See: http://en.wikipedia.org/wiki/Optical_comparator

Out of curiosity I looked up size of sun and distance from earth.

Distance sun to earth, 150 million Km divided by diameter of sun, 1.4 million Km equals a ratio of 107 to 1

So if you used your rig of camera photographing a ground glass screen with object between screen and light, the light 107 inches from the object, you could in theory have a light source of 1 inch to get the same sharpness as using the sun.

If one punched a 1/4 inch hole in a 2 X 2 inch piece of metal and placed it in the gate of a slide projector, and focused the hole at the plane of the object, I'll bet you would get a sharp accurate image of the object.

I used to put solid objects on a sheet of glass in a 4 X5 inch enlarger, focus on the edge of the object and make a sharp image to size.

If it was too big to fit in the enlarger, I would run the enlarger up to the top, put a sheet of paper on the easel, object on the paper, expose and develop.

These are called shadowgrams, see:

http://images.google.com/images?gbv=2&hl=en&sa=1&q=+shadowgrams&btnG=Search+images&aq=f&oq=&aqi=&start=0

If you use a scanner or copy machine, put a solid object on the glass, lay a sheet of black or white paper or cloth over it to contrast with the object, and scan or copy.

Paul39
21st August 2009, 03:54 AM
Joe,

I keep forgetting to comment.

Nice work on the urn copies.

joe greiner
21st August 2009, 09:57 PM
G'day, Paul.

I hadn't delved so deeply into the maths of the puzzle. Thanks for the links and your explanation. You might be exactly right about using a slide projector instead of the Sun. Maybe even a smaller pinhole in the slide.

Whatever the light source, I reckon some fanning occurs. The only truly collimated devices are probably CRT scanners. There was an article in Scientific American around 1975 about them, with additional discussion about streak printing a negative in several directions to reconstruct the image. I couldn't duplicate it in software. The CRT scanners capture spectra in many directions, and use a bajillion simultaneous equations to discover the underlying solids.

Imagine that: Hijacking my own thread!:rolleyes:

Thanks again.

Cheers,
Joe