Thread: Testing a second hand surface plate

Attached are the tables from GGG-P-463C for flatness (two planes that sandwich the whole surface) and repeatability (deviation from local flatness). An 18" square plate has a diagonal of 25.4 inches.
flat.jpgrepeat.jpg

These numbers are in micro-inches. The numbers that Ray quoted are for flatness, and the repeatability (local flatness) requirement is tighter again. On the clip I linked to one of the statements made about the plate they were testing was that while it was flat to a particular grade, using a repeat-o-meter it failed the criteria and should be down graded.

Michael

Originally Posted by Michael G

Attached are the tables from GGG-P-463C for flatness (two planes that sandwich the whole surface) and repeatability (deviation from local flatness). An 18" square plate has a diagonal of 25.4 inches.
flat.jpgrepeat.jpg

These numbers are in micro-inches. The numbers that Ray quoted are for flatness, and the repeatability (local flatness) requirement is tighter again. On the clip I linked to one of the statements made about the plate they were testing was that while it was flat to a particular grade, using a repeat-o-meter it failed the criteria and should be down graded.

Michael

All round that gauging quality episode was good.

As far as specs go, If anyone here is a student or an apprentice, I pretty sure you have access to all the australian standards the ones of particular interest are:

AS 1004.1-1998:Surface plates for metrology - Cast iron
AS 1004.2-1998:Surface plates for metrology - Granite

These are respectively equivalent to ISO
ISO 8512-1:1990
ISO 8512-2:1990

I can't remember now who has the 00 and 000 grades in their standard. AS and ISO grades are 3,2,1,0.

Also there is a newer US standard for plates ASME B89.3.7 – 2013 which supersedes the GGG-P-463C spec, is has a couple of new appendices on uncertainty and traceability.

-Josh

I went back to day to have another look at that granite surface plate. I thought it was 3" thick and a B grade, but it turned out to be A grade and 4" thick. The next silly thing I did was bought it. Made in 1968, so around the same vintage as the rest of my gear
This of course means that I need to make up that repeat-o-meter to see how good it is. Once that is done I can check out the granite plate as well as the cast iron plate I have (although it will have to go as I don't need two - but I'd feel uncomfortable not knowing whether I was selling a plate on that was in reasonable condition)
One day I'll learn to leave the cards at home.

Michael

: ) It's a good size...Maybe they will average out,

If you do make a repeatOmeter, I have some rough idea's on how they are put together.

-J

All ideas are welcome Josh.
A friend emailed me through a page of the standard you mentioned with the testing requirements and they are pretty much the same. (I'm not sure why there is AS1004.1 and AS1004.2 as they say almost identical things). The test does not seem to have changed much at all.
I have a picture or two -
rahn1e.JPG
and it looks straightforward although a horribly clunky design. You can almost hear the discussion -
"The prototype works fine boss, we just need to neaten it up a bit and then we can release them onto the market"
"No. We're too busy at the moment. Maybe later if we have time"

The one thing I haven't worked out is the twin beam arrangement. I'm guessing that the knob pulls the two beams together and so moves the indicator ever so slightly to adjust the zero (total travel for a supramess is around 10 thou I think) with the bolt the plunger is on for coarse adjustment.

Michael

Hi Michael
A repeat-o-meter is something I would love to make as I also have a few surface plates of unknown quality. I have a supramess and 2 mikrokator ( 0.00001 and 0.000005 inch scales) comparators that I could use. Does anyone have a basic drawing of the mechanics of the repeat-o-meter or a link to same and we can see what can be fabricated. Ray might even do some quality assurance and measure its accuracy against the laser interferometer?
I will go hunting on the net and see what I can find.

They look pretty simple, but what do the 2 other handles do? (top an right end)
Presumably the actual contact area would have to be lapped super flat, as would the base?

Ew

Edit, Ahh, those handles....they are simple just handles aren't they?

But, at getting on to 12" long, isnt this going to be pretty useless for plates 18" and smaller?

Stuart

Yeah, the handles are handles. Contact on the bottom is via 4 pads; 3 fixed on the main section and one fixed on the pivoting section. They would not have to be lapped flat but need to be "reasonably flat".

Mark, what I was going to suggest is that someone makes one first and then it can either go touring or be drawn up so anyone else who wants one can make one up. Scaling as I can from the photos I think that one is far heavier than it needs to be (I estimate around 10kg) and you could get the same accuracy from something at least half the weight. I'm even wondering about using Al rather than steel. The mechanics are very simple. Looking at the picture, the main block with the handles sits on 3 pads. There is a pivoting section (using a piece of spring steel) with a 4th pad and the indicator reads off the top of that. If the 4th pad is in a hollow or on a rise relative to the other 3 then the indicator will show that. The diagonal struts are to guard against over travel and as I've speculated above, I think the knurled knob and split beam are to provide some vertical adjustment to help zero the indicator. At the moment I'm trying to find out the lateral spacing of the 3 pads. GGG-P-463c states the pad diameter as 3/8" diameter and at 5" longitudinal spacing; AS1004 states the diameter as around 29mm with 100mm longitudinal spacing. At the end of the day provided that your surface table hasn't got pits in it (that is, any hollows or rises are gradual) it probably doesn't matter which set up you use.
I'm hoping to have a play this weekend

Remember though this is detecting local irregularities. The best analogy that I can think of is a road surface. A surveyor can use instruments to give you the total rise and fall of a road over a certain distance, in surface table terms that is the flatness. This device, while it could do that with a bit of thought is more the thing that measures the depth of the pot holes that the surveyor won't pick up unless his staff is exactly in the right spot.
The standards differentiate between the two by imposing different limits on what is allowed.

Michael

PS for Stuart - yes, but for plates under 30" in the diagonal, it is slid over the surface in a random manner. Apparently smaller versions are made but you are not necessarily conforming to the standard then (although it would give a good idea)

Cool,

I'll draw it up to suit mahr.

-J

Not that I dont want to make a repeat-o-meter, but talking about Michael's problem. A repeat-o-meter isn't any real use until after you know the plate is flat(within spec)......... right?
A repeat-o-meter will pass this.(granted you'd need a small one )

any idea on where the flex point is? is it in the middle or on the bottom?

You could always just blue a known good object on it, like a straight edge...

Josh, the hinge is on the bottom (a strip of spring steel, which I just happen to have...)

Richard, I'll probably end up doing that but at the moment can't rely on that either. Classically to test flatness a straight edge is used with a dial indicator running on it (look up Planekator). These days of course you just send 60kg of granite on holidays to Victoria...

Stuart, you are right. A surface of constant curvature will pass a repeatometer test but if zeroed on a flat surface will work. However, another way would be to scrape something to the surface under consideration and then test both. The true (flat) surface will be the average reading. Once you have that well...

Michael