Thread: Metrology for Amateurs

0.025" = 0.635mm

Originally Posted by Bryan

I'm interested in calibration techniques available to amateurs with basic equipment.

Great idea Bryan, a summary of techniques that are self checking and don't require external standards could be included.

I'm thinking of basic tests for flat, square, straight.. checking for square with reversal etc..

For dimensional accuracy/calibration I think a good set of gage blocks and a clean workspace is the best starting point..

Regards
Ray

Hi Fellas,
On a Russian site there are a couple of videos on metrology, the soundtrack is in Russian but the videos are pretty well self explanatory.
Виктор Леонтьев - YouTube
Regards,
Martin

RC has covered "25=0.64mm", but I should add I believe there is an error in 25=24.98, I think it should read 0.025"=.0248"*
So its two 1/10th of a thou out not two 1/100ths.
25 thou is pretty much full scale on that DTI, I didnt check my numbers against the paper work that came with it. I'll check that later.

They were checked against a master height gauge graduated in 0.00001"(not to be confused with calibrated). But I figured for this it would be "close enough". The master height gauge was zeroed at a handy height. The DTIs were mounted on a vernier height gauge then zeroed on one of the steps. The Master was then wound down and back up to zero to check DTI was repeating zero. Then wound "the first number" of thou up and the DTI was read.

Stuart

*to get to the second decimal of thous on the DTI I would have had to "guess". Though I will repeat the tests later to see is I come up with the same numbers. This shouldnt be taken as an effort to check them to the nearest 0.00001". I'll try harder next time.

On the basics:
Accuracy and precision - Wikipedia, the free encyclopedia
Caliper - Wikipedia, the free encyclopedia
Understanding Errors In Hand-Held Measuring Instruments : Modern Machine Shop
http://www.ignou.ac.in/upload/Unit-5-62.pdf

I like this one:
http://www.mitutoyo.co.jp/eng/pdf/E4329_QuickGuide.pdf

Ok I tested them a little different to last time.
I set the DTI's pretty close to 0 then used the master height gauge to set 0. Read the master, use the master to move the Dti a certain distance. read the master and work out the length it moved.

0.64mm on DTI, 0.02494" on master. 0.00025"error cheap one
.025" on DTI, .02516" on master. 0.00016" error 0.0005" compac
.0020" on DTI, .02020" on master. 0.00020" error 0.0001" compac
.008" on DTI, .00811" on master. 0.00011" error girod
0.64mm on DTI, .02545" on master. 0.00025" error swivel mit

This is the inspection report of the 0.0001" compac. I have no idea what its trying to tell me.

What level of accuracy are you looking for Bryan? For absolute measurements I'd think gauge blocks are the way to go.
If you feel the need to check a mic you'd really need(?) at least an optical flat for checking the anvils for flatness and parallel.

I have one of these kits which has gauge blocks for checking the mic at different "parts of a turn"

If you want to get really excited there are kits of optical flats that check the anvils for parallel at different "parts of a turn" also.

But its not exactly "basic equipment". Some may say a little OTT.

Stuart

Thanks Stuart, for taking the time to do that. I didn't have a particular scale in mind. I just wanted some help to think about exactly the sort of testing you've done - to quantify and compare errors in different instruments. Just for fun really. It seems the concept is pretty simple. You use the most sensitive/accurate thing you have to check the others against. Your master height gauge at .00001" is way finer than I can achieve. That's a quarter of a micron I think, which is a bit mind bending. I now have a Millimess indicator, which at one micron is the finest resolution available to me. Haven't had a chance to play much yet but I will be interested to see how some of my other stuff compares with it.

And the other part of the answer is to try to understand the different sorts of measurement error. Thanks Chris for those links - which I haven't got through yet, but a quick look at the Mitutoyo one showed very simple rigs using micrometer heads as the reference. Out of curiosity I did some ebay searching and 2um seems to be about the finest mic head commonly available. Looks like The Mahr will be my reference for now. Maybe I'll get some gauge blocks one day.

Ray, the geometric checking wasn't what I had in mind. That's a great topic but a different one and deserves its own space I think. Here I was interested in how a mug like me could try to verify his instruments. I guess I was inspired partly by Stuart's rational response to some of the hot air we've had lately, and partly by a couple of acquisitions. Thanks everyone.

0.254micron
Dont get to hung up on the little numbers, remember its just in a shed. It was repeatable and sould be "pretty close" but........

Reading to the nearest micron should give you more than enough trouble.
I have checked some of my gear against some of my other gear and if they agree I call it good. Whats the chance they would both be out the same amount? (and in almost all cases it really wouldnt matter to me if the size wasnt correct in the absolute sense anyway)

Stuart

Originally Posted by Stustoys

0.64mm on DTI, 0.02494" on master. 0.00025"error cheap one
.025" on DTI, .02516" on master. 0.00016" error 0.0005" compac
.0020" on DTI, .02020" on master. 0.00020" error 0.0001" compac
.008" on DTI, .00811" on master. 0.00011" error girod
0.64mm on DTI, .02545" on master. 0.00025" error swivel mit

Stuart, where you haven't specified a resolution can we assume it's .001"? On that basis the bouquet goes to the Girod with an error around a tenth of its resolution. The rest (including the cheapie) are mostly around a quarter, which seems reasonable. The surprise though is the .0001" Compac, with an error twice its resolution. Any thoughts on that?

Bryan, industrially calibration is arranged around a hierarchy of equipment.
Buried deep in the bowels of the metrology department and never seen by mere mortals is (usually) a set of slips*. They are brought out once or twice a year and used to check the next level down - whether that be the working slip blocks or a height master or what ever. Typically a DTI in a stand is used to do this - it can be checked by measuring the slips and seeing that the graduations are accurate. The 'master' slips are then compared to the lesser device and any error noted. The working set of slips can then be used to check micrometers, calipers, DIs, DTIs and down the chain. Of course the further you go, the greater the cumulative error.
The summary of this long winded answer is that you probably need to get a set of slip blocks, call those your master and then check every thing to those. They appear on Grays and Ebay every so often and for general shed use are probably as accurate as you need - they are generally well looked after because of what they are.

Michael
*Why slips? No moving parts, compact and easy to store, less worries about temperature etc (compared to laser calibration or other more complex devices)

Originally Posted by Michael G

Bryan, industrially calibration is arranged around a hierarchy of equipment.
Buried deep in the bowels of the metrology department and never seen by mere mortals is (usually) a set of slips*. They are brought out once or twice a year and used to check the next level down - whether that be the working slip blocks or a height master or what ever. Typically a DTI in a stand is used to do this - it can be checked by measuring the slips and seeing that the graduations are accurate. The 'master' slips are then compared to the lesser device and any error noted. The working set of slips can then be used to check micrometers, calipers, DIs, DTIs and down the chain. Of course the further you go, the greater the cumulative error.
The summary of this long winded answer is that you probably need to get a set of slip blocks, call those your master and then check every thing to those. They appear on Grays and Ebay every so often and for general shed use are probably as accurate as you need - they are generally well looked after because of what they are.

Michael
*Why slips? No moving parts, compact and easy to store, less worries about temperature etc (compared to laser calibration or other more complex devices)

That just brought back some memories from long ago.

Buried deep in the bowels of the metrology department and never seen by mere mortals is (usually) a set of slips*.

Very true and all done in an airconditioned room , except that in my case the airconditioned room had all round windows so mere mortals could see in and find out all the secrets .

That is precisely how it was done. Sometimes they even let us mere mortals touch the lowest of low slip blocks lol.

The

Typically a DTI in a stand is used to do this

is a purpose made comparator with fixed base, column and a very precise indicator on top. The base is very precicely lapped and generally has a slotted surface (from 30yrs ago) so dust etc can be pushed off the surface during the seating process. For those who have not experienced this.

Thanks for the reminder Michael.

Dean

Hi Bryan,
Sorry.
Girod is 0.0001"
Thoughts on the Compac error? (Well one springs to mind but Im not game to say it )
But really, while both are 0.0001" indicators, the Compac is being checked over 2.5 times the distance of the Girod. So the Compacs error per inch is less than the Girod. I'll do a bit of a map of them both later.

Hi Michael,
As far as checking off a slip block with a DTI. Thats how my master is zeroed, but my master goes to 0.00001" my DTI's only 0.0001". So you're left "guessing" oops I mean extrapolating. Is this the done thing in Metrology?(I'd think you could pick up an error of maybe 1/100th a gradution. Not saying you would be able to give an accurate number, just that you could tell it was "not zero").



Was just looking up the specs on my master.
Deviation of central length =/- .00005"
Variation in length .00004"
Block feeding error =/- .00005"

Stuart

Originally Posted by Stustoys

As far as checking off a slip block with a DTI. That's how my master is zeroed, but my master goes to 0.00001" my DTI's only 0.0001". So you're left "guessing" oops I mean extrapolating. Is this the done thing in Metrology?(I'd think you could pick up an error of maybe 1/100th a graduation. Not saying you would be able to give an accurate number, just that you could tell it was "not zero").

Was just looking up the specs on my master.
Deviation of central length =/- .00005"
Variation in length .00004"
Block feeding error =/- .00005"

Stuart

Interpolation is allowed - usually 1/2 a division, so in this case you could call a measurement to the nearest 0.00005", but you are left with the uncertainty of the indicator which you would probably state as +/- 0.00005" anyway.

Metrology is about uncertainty - while the deviation may be +/- 0.00005" and the feed error +/- 0.00005", all up when checking something against a height master I would be saying that it measures 1.2345", +/- 0.0001". The resolution of the DTI doesn't actually matter as much as you are using it as a comparator as you will zero it against the surface of interest and then zero it on the height master to get a number. What is important is that the height master when compared to a slip is within accuracy limits.

Michael

Well here's a rough & ready test, FWIW. Reference indicator mounted on the rear of the cross slide, bearing on a bar in the chuck. Subject indicator in toolpost, bearing on opposite side of bar. That big thing lives there so it came along for the ride.

I tried 4 indicators - 3 cheapies and a Mitutoyo. Over .1mm (full range on the Mahr), 2 of the cheapies were out by a micron. The other two were more like 1.5. So that's the long travel one that lives on the cross slide and gets a rough, dirty life. And the Mitu that I bought new in 1985 and have hardly used. Interesting. Either way it's plenty close enough for the work I do.