Dial Indicators

[HopperParticipant @hopper]

Posted by Peter Krogh on 12/02/2016 00:56:16:

Deming tried to sell US industry on his philosophy but they wouldn't bite. So he took it to Japan where they were basically starting over. The rest is history. By the time the method was popularized in the US, commonly known as 'Just-In-Time' (which is only part of the system) it was the '90s.

Pete@been there done that wore out the tee shirt

I remember when the local Chrysler/Mitsubishi car factory introduced "Just in Time" production methods.

Instead of JIT, it was commonly referred to as NQIT – Not Quite in Time.

[MWParticipant @mw27036]

I actually am quite fond of mitutoyo but i would say that as i worked for a company that used nothing but mitutoyo measuring equipment and would say it is good enough for the job but consistent reading between inspectors are hard to come by, but you could take that as a general "rule" of measuring equipment (no pun intended)

Other than that, i do believe that it wont last as long as some of the equipment still being used today like baty, or moore and wrights famous slip gauges, i think they would have them "calibrated" every year or chuck'em, so its not meant to last basically, but were used nearly all the time.

Does beg the question sometimes of how accurate is enough, well i went to college where the verniers seemed to be the cheapest of the cheapo, but there were alot of old tools like baty and starret from their old premises.

Michael W

[Danny M2ZParticipant @dannym2z]

I was once told that the 'accuracy' of a measurement system should be a decade (10x) better than the smallest division.

It made sense at the time, but it was many years ago during training – so I have forgotten the logic behind this statement.

Also, there may be a big difference between the 'lifetime' of a measuring instrument used daily in a commercial environment and one used occasionally in a home workshop.

One 'problem' with purchasing Mitutoyo instruments is that they are so highly regarded that fakes have been manufactured. Read more about them here **LINK** Online purchases may be more risky than purchasing directly from a reputable supplier but still be careful as their purchasing department may not be techno-aware and just focus on the bottom line.

* Danny M *

[Neil WyattModerator @neilwyatt]

Measurement accuracy

closeness of agreement between a measured
quantity value and a true quantity value of a
measurand

Measurement precision

closeness of agreement between indications or
measured quantity values obtained by replicate
measurements on the same or similar objects
under specified conditions 2.15

Isn't that exactly what I said?

Neil

[Neil LickfoldParticipant @neillickfold44316]

It is interesting about brands and type of equipment. If it gets the work piece close enough for what you are doing than it is fine. It only becomes a problem when you are trying to indicate something and the indicator can not reliably put the part to that level of precision. You then either buy a new instrument with a finer reading amount, or make a design your parts in such a way that you are not reliant on the indicator for the precision. I did buy many years ago a Chinese finger type indicator with a 0.01mm reading. It would show after about 3 months that the work piece was indicatored to better than 0.01mm, but was actually out by 0.02mm. It went into the bin. The Tesa indicator I brought to replace it, is still very good and I check them about 3 times a year for repeatability and incremental reading error. An example of needing better reading indicators is when I modified the bearing races. It was a lot easier to indicate the inner race of the bearing to be reworked with a 0.002 mm indicator as to using the regular 0.01 indicator. Trying to gauge a 1/3 of a reading on the indicator is really for the very young eyes. With the 0.002 indicator, that becomes 1.5 divisions, much easier to see for sure.

Neil L

[Anna 1Participant @anna1]

Re Mitutoyo.

I still remember (over 50 years ago) the metalwork master saying " who ever heard of Mitutoyo" disparagingly referring to this foreign company, how wrong he proved to be.

I still daily use one of the first digimatic micrometers Mitutoyo made (bought about 30 years ago) Mitutoyo repaired it recently and it cost about £35 to repair, thats what I call service. I also still use a manual 6" Mitutoyo vernier that is about 45 years old and after all these years use still is more accurate to use than a new mid priced digital vernier that I bought recently and which is about to go in the bin .

regards

Anna

[MWParticipant @mw27036]

Yes, it only happens rarely but you find that sometimes your finest isn't fine enough for what you're doing, i found parallelism and concentricity and hole positions to be the worst ones, alot of what i do i'll spot through if i can, i dont suspect for a moment that i'm alone in that though.

Neil, i suppose the terms can be a little synonymous at times, i dont think many of the inspectors i knew would be able to tell me unless i said something like "repeatability" or "true to nominal".

Yeah i was aware of fakes of mitutoyo, will have to check against that graph though :P! it hasnt listed my product specifically but i guess the warning signs can be similar, if i remember rightly, i took steps to make sure i bought it from them directly through their listed retailers and the micrometer wasnt cheap, Funny you should mention it, i do remember telling one of the production engineers how much i bought it for and he said "ha! theyre listed at half the price in our catalogue", yet i was fairly certain i paid what it should cost, maybe theyre buying fakes after all

Michael W

[AjohnwParticipant @ajohnw51620]

Posted by Hopper on 12/02/2016 03:32:15:

Posted by Peter Krogh on 12/02/2016 00:56:16:

Deming tried to sell US industry on his philosophy but they wouldn't bite. So he took it to Japan where they were basically starting over. The rest is history. By the time the method was popularized in the US, commonly known as 'Just-In-Time' (which is only part of the system) it was the '90s.

Pete@been there done that wore out the tee shirt

I remember when the local Chrysler/Mitsubishi car factory introduced "Just in Time" production methods.

Instead of JIT, it was commonly referred to as NQIT – Not Quite in Time.

Just in time suited the Japanese well. Lack of space were many factories were so it paid to keep storage areas down. They had another cute trick too. If production faltered and stock wasn't being used as a result it still came in and they stored it all over the place even in corridors etc and reckoned it caused people to speed up and get rid of the backlog.

Personally I think certain western industries have used it for a very long time – process industries such as battery manufacture, probably chocolate and other similar things too.

John

–

[MWParticipant @mw27036]

I believe mitutoyo have their own metrology lab/institute based in the UK.

Michael W

[KWILParticipant @kwil]

Coventry for their Calibration Lab

[VicParticipant @vic]

Mitutoyo are a good company to deal with. I wanted a Micrometer barrel for a particular project that had to be reverse reading with a carbide ball and a barrel clamp. They didn't offer such a thing but after speaking to them they made a special for me by swapping some parts from what they had. Cost – same as a standard barrel.

[MWParticipant @mw27036]

sounds like a decent bunch of fellows. I like how they have time for an individual customer, it shows dedication.

Michael W

[SillyOldDufferModerator @sillyoldduffer]

Posted by Neil Wyatt on 12/02/2016 07:22:16:

Measurement accuracy

closeness of agreement between a measured
quantity value and a true quantity value of a
measurand

Measurement precision

closeness of agreement between indications or
measured quantity values obtained by replicate
measurements on the same or similar objects
under specified conditions 2.15

Isn't that exactly what I said?

Neil

Perhaps I do understand it after all then!

It just goes to show how careful you have to be about the written word. English may be a great language but it's all too easy to trip over shades of meaning especially when words have both lay and specialist meanings. Like Energy, Force, Work, and Power for example.

You'll have to forgive me. I come from a profession where "Entity" as in "Entity Relationship Diagram" can be defined as being a "thingy"…

Regards,

Dave

[Anna 1Participant @anna1]

Hi, Neil.

I have just read the NPL and your analogy to archery regarding precision /accuracy and have to accept your and NPL definition and understanding of these words..

It still bothers me though that a series of components that are innacurate and out of tolerance, albeit all with a similar error and are heading for the scrap bin, can apparantly, because they are of similar wrong size, be described as a precision component.

Best wishes

Anna

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Anna 1 on 12/02/2016 22:02:58:

… It still bothers me though that a series of components that are innacurate and out of tolerance, albeit all with a similar error and are heading for the scrap bin, can apparantly, because they are of similar wrong size, be described as a precision component.

.

Anna,

It may be more comfortable for you to think in terms of our knowing 'with precision' that the components are wrong, and by how much.

In case it's not obvious: This is the basis of statistical process control … the machine is run until it is no longer producing 'perfect things' and is then adjusted, to account for thermal expansion or whatever … which is a simple matter if we know that it is now producing consistently imperfect things.

MichaelG.

Edited By Michael Gilligan on 12/02/2016 22:44:14

[Anna 1Participant @anna1]

Hi, Michael.

Thanks for that way of looking at it.

Perhaps NPL should have chosen a new word altogether, I don't remember that many years ago precision had its current meaning. but I am probably wrong

Kind regards

Anna

[AjohnwParticipant @ajohnw51620]

As I am a Brit I get cheesed off with both pedantic use of the English language and certain august bodies choosing to alter the meanings of word to suite their own specific context – as do physicists with words like colour and flavour which is a little more excusable.

This is what the oxford has to say about it – it would have far more in the actual dictionary but it's a fact that the 2 words can be interchangeable in certain contexts. They are both meaningless without further information anyway.

I used to shoot precision 22 rifle. I might say that I can shoot 22 rifle with a high degree of accuracy.

What I have just said about my 22 rifle shooting doesn't really mean anything unless more information is given.

Precison 22 also needs more information. A bull in precision 22 rifle involves the equivalent of hitting a pin head at a certain distance.. Normal 22 or small bore as 22 is usually called uses a much larger bulls eye. The distance is the same.

High degree of accuracy. I might say my 10 shot average is 98.8, which it was providing I used competition rounds. But that doesn't mean much either unless i say this was shooting precision small bore. The competition round was worth one extra bull over the ordinary ones. Another factor – bit like using Mitutoyo rather than something less thoroughly statistically checked. I can only think of one instance when I failed to get 100 with the other type of target. Shooting against the army using their guns that they kindly provided with miss set sites in order to try and beat us so my first 2 shoots were out of the bull.

I'm not keen on Americanisms either. They should leave our language alone.

John

–

Edited By Ajohnw on 12/02/2016 23:50:01

[Bill PudneyParticipant @billpudney37759]

Nothing like a discussion over measuring things to bring the pedants out!!

cheers

Bill

p.s. When I was doing my time the instructor passed on a little gem, which went something like this….There are three basic types of measuring instruments, steel rule, vernier calipers (it was the 60s) and the micrometer. Whatever you make, however you check it as it's being made, use the next finer level of instrument to check it, when it's finished". This is something that is similar to Dannys post earlier.

This isn't intended to upset the users of bits of string with knots in.

[Neil WyattModerator @neilwyatt]

The OED may document English as she is spoke.

That doesn't help much when we are trying to better understand a technical issue and using words in the narrow senses associated with a particular topic.

Think of the 150 years of confusion because of how people interpreted Darwin's use of the word 'struggle' in the context of natural selection… (e.g. as a conflict between combatatnts rather than as individuals against the odds).

I thought I'd google 'NPL Accuracy precision' and see if anything helpful appeared, and lo!

http://www.npl.co.uk/upload/pdf/Accuracy%20and%20Precision.pdf

(I honestly haven't seen this before).

Neil

[pgk pgkParticipant @pgkpgk17461]

Posted by Neil Wyatt on 11/02/2016 17:10:42:

The way to get precision and accuracy in your mind's eye is to think of target shooting.

A tight bunch of hits away from the bull is precise, but not accurate.

A loose bunch of hits clustered around the bull are accurate but not precise.

A tight group on the bull is b0th precise and accurate.

You can work with any or the three – precision without accuracy means you need to identify the error and allow for it. Accuracy without precision can be partly offset by averaging several measurements.

Obviously if you have both, you are in clover.

Neil

We can add to these analogies. I used to be into target archery.

If one took an average bow and placed it in a stand with remote release and tuned it then its repeatability (in still air conditions etc) would likely stack all the arrows into the back of each other.

Give that bow to a master bowman and he'd shoot tight groups but would have to fiddle with the sighting to centre the groups to the target. He may be consistent but equally is consistently 'off'

Tell the master bowman that this was an average bow (as opposed to a top-named brand) and the group suddenly becomes less tight. He no longer expects tight groups so doesn't get them

In other words there's a difference in the accuracy/precision/repeatability of an instrument once there's human influence in it's use and a bigger difference once there's human doubt in it's 'precision'

[John HaineParticipant @johnhaine32865]

Neil, that's perfect, thanks for finding and sharing the link. Brilliant graphic.

[Neil WyattModerator @neilwyatt]

Perhaps what's most useful is thinking what these things mean in terms of ASSESSING and USING our own measuring instruments.

For practical use in the workshop, the most useful property is precision/repeatability. Why? because we usually make parts to fit each other rather a standard we can't measure.

If we use a device as a comparator then is absolute accuracy is irrelevant, all we need is for it to be repeatable and sensitive enough for our purposes.

Where we need accuracy is usually when we either (1) make parts to fit parts we cannot measure directly or (2) we have to measure mating parts with different tools.

Sometimes we can avoid (2) – for example, if making a shaft to fit a hole we can sometimes use a gauge inside the hole and make then measure both gauge and shaft using the same external micrometer. Of course the use of the gauge introduces another step and an another degree of inaccuracy.

The point is also well made that the skill/touch of the user or the mounting of the tool can have a significant impact on the readings obtained.

Other things can have an effect, I once compared two cheap digital calipers, allowing them to warm up and down over more than ten degrees C. One pair kept its reading rock solid, the other changed its reading up and down consistently with temperature (by much more than expansion of the scale could cause).

So a stab at a few practical suggestions, which others may choose to elaborate on/change:

1 – where possible use measurement tools as comparators.

2 – take repeat measurements ('blind' if possible) to get a fair idea of the repeatability/precision of a measuring tool AND the measuring process.

3 – if using two measuring tools, it's worth checking them against each other (e.g. use an inside micrometer between the anvils of an outside micrometer and see how well they agree).

4 – it's worth having a few 'standards' in the workshop. These can be the insulated bars supplied with larger micrometers, but a couple of gauge blocks or even bearings with accurately finished outer diameters. These can be used as a check on tool and technique before making critical measurements where absolute accuracy is important.

5 – Spend a while testing your equipment, ideally by measuring the same object or objects several times with each one. Your most precise and accurate devices should be come apparent rapidly – these may not be the ones you expect. If the ratty looking boot-sale mike turns out to be better than your £150 digital one, then keep it in a safe place!

6 – Be aware of the impact of temperature, especially on digital devices. Try and work at a consistent temperature (remember an industrial gauge room is kept at constant temperature and humidity).

Finally, cleanliness is very important. Just drawing a piece of paper between the anvils of a micrometer can make a big difference to accuracy and repeatability, especially if you are 'chasing tenths'. Keeping the scales of digital calipers clean and dry is vital (unless you have special inductive ones).

Neil

[AjohnwParticipant @ajohnw51620]

It seems NPL have a lot to do with the inadequacy's of some model engineering items – all things can be called high precision and people can finish up with something that doesn't meet up with their expectations.

It's just another case of some specialised field trying to change the meanings of words in their field – jargon in other words. Something that is wide spread and should be discouraged as usually there is no need for it.

Numbers have precision. A/D and D/A's have precision, in that case there precision can not be better than +/- 1bit at what ever physical resolution they work at. The other errors relate to what ever reference they use. That can be all sorts of things including any of them being held at a closely controlled temperature. The same statement could also be stated using the word accuracy. The power of the English language – context. They also have linearity errors. In all cases where these 2 words are used it's the error in obtaining the figures that is the important aspect.

A could example of context is the word tap. I can turn a tap on, use a 1/2 bsw tap, tap a tree, tap a table or even tap dance. Just like accuracy and precision the word is meaningless without further information and once the information is available the meaning is clear.

Precision small bore has a related message as well. The quality of the bullets. A score of 100 would be a fluke with standard rounds but would be possible with the normal targets. Some one might look at the precision targets and say that's not the same as hitting a pin head due to the size of the bull. Some one would then have to explain that outward scoring is used – touch a line and lower score is used. The other form is inward scoring. One target has 10 small ones on each sheet because a bullet might pass through the same hole or it might have missed the target completely. The other form has single large target. The bullet might tear the target as it goes through it. Precision uses a gauge, a magnifying glass with a bullet end sticking out of it. An approved person has to make the decision. Bit like using a mic really. Some people will get more precise actual real readings from them. The word accurate could also be used.

John

–

[Anna 1Participant @anna1]

I am just sad that engineering which all my life, to me, has always been synonymous with clear straightforward thinking, has been muddied by adding a new meaning to a word, rather than NPL simply choosing a new word to describe a particular situation. The Victorian engineers will be turning in their graves.

Kind regards

Anna

[Neil WyattModerator @neilwyatt]

You can't blame NPL.

This meaning was codified in international standards in 1994 with ISO 5725-1:1994.

Neil

For anyone with insomnia, this is chapter and verse on all the terms associated with metrology. I suggest taking paracetamol BFORE reading it:

/www.bipm.org/utils/common/documents/jcgm/JCGM_200_2008.pdf

BIPM is the Bureau International des Poids et Mesures

Edited By Neil Wyatt on 13/02/2016 11:08:44

[Author]

Posts