Angle plate accuracy standards?

Peter

Could you not buy a 'standard item and machine it to the necessary tolerance on your own machine? --

Chris Edwards (in deepest Dorset) "There *must* be an easier way!"

Buy two more and scrape all three for mutual alignment. That's the only way you'll get 0.015mm or better errors on a cheap 100mm angle plate and prove it.

It's time consuming, but satisfying. DAMHIKT

Mark Rand RTFM

I could (and probably will) machine the angle plate I bought to be accurate to around 0.02 mm (20 microns). However I don't think I could get it accurate to much better than that - eg the 0.003mm (3 microns) in

100 mm that DIN875(00) implies - with my present machines and without an accurate angle plate for reference. **The thing here is, an angle plate is supposed to be a reference of some sort **

- whether for machining, or for marking-out/checking - and if the reference is less accurate then the machine then ... that's bad.

A cheap DIN 875(1) / BS939(B) 100mm engineers square, around £3-10, will be accurate to 15 microns or so - 1 part in 6,600.

The angle plate I bought, and it's by no means the cheapest available, is actually only accurate to 1 part in 150, or less.

-- Peter Fairbrother

Okay - but can I ask you how you do that?

Thanks,

- Peter Fairbro

There is a simpler fixture you can make to measure the angular error if you have a good surface plate. You need a block or plate about the height of the angle block and thick enough to stand on edge. Machine two opposite sides as parallel as you can measure. Drill and tap into one end near both of the parallel surfaces. Run screws with locknuts part way into the holes. Stand this fixture up on the surface plate with the screws pointing toward the upright web of the angle plate, and adjust both screws to touch it. Flip the fixture over and slide it up to the angle plate. If the angle plate isn't perfectly vertical one of the screws will clear by twice the error.

A 1-2-3 block with tapped holes might work. Stand it up on end, insert the screws into a 1x3 edge.

You could make a fancier version with a dowel pin next to one parallel edge and a sensitive dial indicator next to the other.

jsw

You need a decent surface plate to start with. Scrape one side of each angle plate to be 'flat' to the surface plate.

Letter the angle plates A, B, C

Scrape and match A-B, then B-C, then C-A and then back to A-B etc and keep going till they're (all) good enough. Like the man says, it's not quick, but it is satisfying and with perseverence, guaranteed to get the result.

If you haven't got a good surface plate.... get 3 and do the same with them first to get them as flat as you want/need.

Richard

Thanks.

Can one scrape granite?

-- Peter Fairbrother

I've never done it, ut I don;t see why not. You would probably need a carbide scraper, or rather a scraper with a carbide tip.

This is an atiquated text, but considered by many to be the benchmark for the technique in general. Personally I find it 'readable', but many who are used to more modern bullet point and comic-strip presentations find it too 'wordy'.

Richard

Thanks! I dunno about wordy, but at 500 plus pages it is rather long ... actually I'm a bit gobsmacked!

I have a good surface plate and a scraper, so I may give scraping a try

- someday - but wow, it surely does look like a long business.

Do people still use it much, in real industry?

--Peter Fairbrother

Page 46: "Should a granite plate become damaged it must be returned to the manufacturer who has proper facilities for refinishing it".

jsw

>

Peter, there is lots of discussion about scraping on the Practical Machinist and HSM forums (fora?), some of it quite 'forthright', but I think it's fair to say that scraping is much less widely used now than it used to be, but it's not dead.

For your use in this context, only the first 75 pages or so of the book is relevant, the rest of the book relates to rebuilding machine tools using the scraping techniques described in the first chapters.

Richard

Out of interest how accurate would the surface be if it were machined with a properly set up surface grinder?

Alan

In a commercial environment I'd expect within a few tenths (~0.006mm) without too much bother. In arm waving terms, it starts to get a lot harder as you get bigger.

Richard

This is directly relevant, in partucular some of the embedded links are useful:

Richard