Stressing over numbers

[not done it yetParticipant @notdoneityet]

Also I noticed that at first he said his material was 20mm but when he measured it was 19.99mm But even Stefan called it 20mm. That was a little confusing lol.

Wasn’t that the starting diameter? If so, that was of little significance – it could have been 19.50-20.49 and still have been initially measured as 20mm and 19.99mm, measured with a caliper, is as near as necessary so rounded to 20mm. It is only that last mm, or so, that is really important – where he reverts to a micrometer for accurate measurements.

[JasonBModerator @jasonb]

You don't actually say how you are setting the DRO on the lathe, if that is wrong than all measurement based on it's readings will be wrong. For OD the method would be to skim the part and measure with whatever you are confident with then enter that value into the DRO, final cut should be to a similar depth as the cut you set things with to keep tool deflection the same.

I only have traditional micrometer and one and one digital, the traditional does not get used much and the digi has not been out in years yet I manage to make things that work so sounds more like tequnique, I'm a fan of making one part to fit another not making two two a size. Whats that saying "all the gear …………..!

[Anonymous]

Posted by JasonB on 26/05/2020 16:24:59:

Whats that saying "all the gear …………..!

Remind me, I seem to have forgotten, if I ever knew in the first place.

Andrew

[Michael Edwards 1Participant @michaeledwards1]

Posted by JasonB on 26/05/2020 16:24:59:

You don't actually say how you are setting the DRO on the lathe,

Jason thats normally what I do. I take a cut then enter that as the new base reading. I normally end up setting this a few times over the machining process.

[Chris GunnParticipant @chrisgunn36534]

Michael, a lot of folk have mentioned tool spring, it must be a thin boring bar you are using, so susceptible to springing. When I get near to the size I want, at the penultimate cut, I go through and then reverse the feed and let the boring bar cut on the way back too, then measure, then set the final cut. Then I go through again, and withdraw the tool with the lathe stopped, and without changing the setting, and measure again. it should be on the money, maybe on the tight side. if it is, I advance the tool through the work with it stationary, and then start it up and let it feed back out again, and it will probably take a touch out again, and then you should be pretty close, all other things being equal. Patience is a virtue, and I take it slow, especially if it is an expensive casting.

Chris Gunn

[Phil H1Participant @philh196021]

Michael,

Forgive me if you have already tried this – I might have missed part of the thread. Are you experiencing the same variations with one of your traditional micrometers because manual types don't give anywhere near the kind of variation provided you apply the correct 'feel'.

I have experienced this kind of variation when using an expensive digital vernier (it was demoted to the scrap bin).

Phil H

[SakuraParticipant @sakura]

There was reason for apprenticeships, it was a learning time. Experience takes a long time to accumulate, I've been working in engineering for 55 years, every day is still a school day. I still get things wrong, sometimes things I shouldn't get wrong but that's engineering. A man once said to me " if you can get it right every time, stop doing this and go on the Stock Market, you'll make billions".

[JasonBModerator @jasonb]

Posted by Andrew Johnston on 26/05/2020 17:20:46:

Posted by JasonB on 26/05/2020 16:24:59:

Whats that saying "all the gear …………..!

Remind me, I seem to have forgotten, if I ever knew in the first place.

Andrew

………………No Idea

It is something that I seem to keep seeing on the forum, many beginners seem to be feel that they need to be working to umpteen decimal places, the easy availability of relatively cheap measuring equipment that is now available may well be one reason for this. They seem to feel that having all these measuring tools, DRO etc will mean they can make better parts yet don't really know how to get the best out of them. For example Barrie wrote a very good post the other week about using telescopic gauges and the lengths he went to for consistant results yet I'm sure many users don't take that time and will for example stick the tele gauge into a freshly bored hole while the work is still warm for cutting, etc.

The OP's comment about imperial being worse as there are more decimal places is also interesting as the acutal amount is basically the same eg 0.01mm and 0.0005" (0.0004 rounded up) which are what a lot will read down to so why more errors when using imperial than metric?

Many don't know what finished size is really needed for their part, yes there are time when you do need to hit the exact size with very little deviation but that is far outweighed but the times there is more tolerance or even when a part is just cosmetic

The recent threads on the Stuart 10V are another example where hundreds must have been made by filing the bottom of the bearing housing and hand fitting the bearing using just a steel rule and firm leg calliper. Yet measured fits to two decimal places (metric) are being looked for.

I posted a video of the current engine yesterday, could not tell you what many of the finished sizes are to several decimal places, the majority had the hole done first, a few small ones by reaming and larger by boring some of which were then honed or lapped these were cut to nominal size and just measured with digi callipers. The mating parts were again measured with digi callipers and when I got close to size they were turned until I got the fit I felt was needed, so no measurement taken.

So what I'm saying is get a feel for your machines, cutting tools, etc and the fit of the parts you need rather then just making two parts to numbers that may or may not have been measured correctly and the allowances to those numbers for fit may also not have been calculated correctly. If you don't use the numbers you can't stress over them

[Michael Edwards 1Participant @michaeledwards1]

Thanks JasonB, good words of wisdom there, and I can see where you are coming from. I think a change of mind is actually needed. If the parts fit then its a 'Go' / 'No Go' situation.

I did spend some time machining last night I was working on the Cylinder covers for the Stuart S50, not a critical measurement but I wanted to see if what I seen on Stefan G's video I could put into practise.

I was aiming for 28.575mm (1 1/8) and I got 28.60mm I was so pleased with that.

(part moved slightly in picture when I took the photo so part not on centre after taking photo, measured it 10 times for consistancy just to see if i could replicate the feel of the micrometer).

 

 

 

 

 

Edited By Michael Edwards 1 on 27/05/2020 08:42:44

[SakuraParticipant @sakura]

Exactly right Jason. People seem to forget how much excellent model engineering was carried out with a very average lathe, calipers and files.

[Martin ConnellyParticipant @martinconnelly55370]

I think part of the general problem for beginners is that they do not understand or have not been involved with the concepts of tolerances, fits and the need for parts to be interchangeable or not as the case may be. It has probably been discussed before or appeared in the magazine at some point. As has been stated earlier make the bore smooth and parallel then fit the piston to the bore. If the bore is sized at 20mm it is 20mm plus or minus X where X is whatever you want to achieve or can achieve. If you are not making a part to be sent off to mate with another part made by someone else somewhere else then X can be quite large and not be a problem. A bore of 20.1mm does not automatically make the cylinder scrap.

Martin C

[Howard LewisParticipant @howardlewis46836]

Consolation, I hope.

When as an Apprentice, I scrapped a large job., which nearly reduced me to tears, I was told "The man who never made a mistake, never made anything"

For years beside my desk was a card saying "Experience is what allows you to recognise the mistake the next time you make it"

Very little experience is wasted, even when things go wrong, if you can find what went wrong, you will have learned, hopefully, how to avoid that pitfall in the future..

Although I use a digital calliper a lot of the time, (pressing the jaws onto the job by hand ) for greater precision, I prefer a mechanical mic. With a digital device the limit is the resolution, which is likely to be in the fourth decimal place. But at which end? Just about to drop to the next lower reading or to the next upper?j .

With a mechanical mic, the thickness of the line is less than the graduation, so if you have a mic with a vernier barrel, you can estimate, by eye to better than the last decimal point. As long as you have the same feel each time!

HTH

Howard

[Tim HammondParticipant @timhammond72264]

Posted by AdrianR on 26/05/2020 12:18:03:

I watched a video by Stefan Gotteswinter recently about how to turn precision diameters, it is worth watching **LINK**

Adrian

I tried this technique yesterday on my cheap 'n cheerful mini lathe. Didn't work for me. Ho hum.

[not done it yetParticipant @notdoneityet]

Michael,

As Jason sort of indicates, consistency in reading is important but that reading will always have an error of plus or minus one in the last digit – at the very best.

I have two (cheapish) dro’s that indicate to 0.01mm but, if one looks at the specs, they are certainly not guaranteed to better than perhaps 0.03mm in terms of absolute accuracy. One of them occasionally drifts in the last digit, the other is rock steady, but neither are necessarily better or worse than the other for most purposes. That is why I rely on a micrometer for my best consistent readings of final size. I always use the same measuring instrument for taking readings, unless the two agree when measuring a standard item. Mostly, I don’t particularly care about absolute accuracy – as long as the parts fit. Shrink-fitting smaller parts always needs careful attention.🙂

It is the same with surface finish – the only one ever seen, afterwards, is that last finishing cut – all the rest have gone. If taking 4mm off an item one could, just as well remove 3 1/2mm with an angle grinder and just finish the job on a mill, taking 0.1mm cuts. That is why there are ‘roughing’ and ‘finishing’ cutters available to us.

[CirclipParticipant @circlip]

There are so many who will give up the hobby if ever liquid crystals develop permanent leaks. Probably have to revert to callipers and fractions and re learn the art of FITTING. Unless of course one is supplying the Swiss mechanical watch industry or NASA.

Regards Ian.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Michael Edwards 1 on 27/05/2020 08:26:55:

Thanks JasonB, good words of wisdom there, and I can see where you are coming from. I think a change of mind is actually needed. If the parts fit then its a 'Go' / 'No Go' situation.

…

High accuracy and tolerances are needed to meet the demands of interchangeable manufacturing which is a different world from my workshop! Interchangeability has major advantages in manufacturing; spares can be exchanged without fuss, and – once set up – it's no longer necessary to employ large numbers of skilled Fitters.

Making parts on the interchangeable system is difficult. It originally required specialist tool-makers working in tool-rooms equipped with high-end facilities making accurate gauges, jigs, and fixtures for use on the shop floor. The jigs, fixtures and machine settings provide most of the accuracy, not the production worker.

Before the interchangeable system, a different approach was used. Typically, work was done to rather lower standards of accuracy, and parts then had to be 'fitted' together, in effect tool-room gauges are replaced by the job itself. With care, items finished this way will be every bit as good as their interchangeable equivalent. The down-side is it's uneconomic in mass production, and spare parts can't swapped without more fitting. However, 'fitting' remains the best way of making prototypes, small runs and repairs. And it can be done without high-end measuring equipment; ungraduated spring calipers can be used as comparators. For many purposes it's only necessary to establish 'Too Big', 'Too Small' and 'About Right'. In this context a micrometer becomes the slave not the master.

Accurate measuring is difficult, and I suspect many claiming accuracy in home workshops are deluding themselves! It needs training, practice, controlled conditions, and careful technique. Just buying the most expensive Mitutoyo available doesn't guarantee accuracy because the operator needs to know how to use it properly.

Keeping a sense of scale is difficult too. Here's a magnified 20mm line:

Inside my CAD package, the line is represented as floating point coordinates with 15 to 17 digit precision. The 20mm line isn't exactly 20 units long, but it outperforms a mechanical micrometer by at least 13 digits.

The blue line is 20.01mm, and the red line is 19.99 mm long. The two circles are 0.4mm diameter. Even at this magnification, about 8X on my screen, there's no obvious difference between the red, blue and white lines.

Necessary to zoom in on the 0.4mm circles to see differences.

The point is making parts to fit together at 0.01mm accuracy is really hard and usually unnecessary. For example, making a shaft to fit inside a 20mm bearing, the shaft only has to be small enough to fit freely in the hole, but not extravagantly loose. In practice, fits can be quite sloppy – for most purposes working to about a thou (0.0254mm) is more than good enough. There's even advantage in loose fits, chaps who make clocks often say high-accuracy working results in over-tight mechanisms – clocks and model steam engines both work better when made a bit loose.

As always the validity of any forum advice depends on what you're making. I mostly make utilitarian stuff where good finish doesn't matter much and I only do high-precision when absolutely necessary. I don't make models that have to work and look good. I'm more likely to measure and compare with a DRO, Digital Caliper, spring calipers and other parts than a micrometer. My micrometer is only wonderful when it's really needed!

Dave

Edited By SillyOldDuffer on 27/05/2020 10:55:54

[Clive FosterParticipant @clivefoster55965]

Objectively the combination of inexpensive DRO systems, affordable inserts and at least reasonably accurate dials on affordable new and most decent used machines are game changers for Model Engineering. But the way we look at things and standard newbie teachings haven't caught up.

Back in the day Model Engineering was very much a craft. You made things to fit with caliper and rule based measurements. In post war years maybe a vernier or micrometer would have been affordable. If really posh perhaps a Unique "dial gauge" too. Which was about the state of play in the mid 1970's when I caught the machine bug. But I got lucky with a shop clearance sale for a dial gauge.

Hard saving needed if you wanted a machine with reliable dials but used Myfords, Boxfords, SouthBends et al could be found. Even more hard saving to get a new machine. I recall my then boss mimbling hard over a new Boxford ME 10 or a new Portass, cheaper but no dials, to replace his Zyto. He got the ME10, I suspect the quotation from Portass was one of the last they ever issued.

Over the years affordability, availability and capability of machines and tools have steadily increased and experienced folk work the tool-room way relying on accurate machines and accurate measuring equipment. I guess I'm now rather better equipped than the local section workshop was when I worked at RARDE / DERA / DRA / Qinetiq. Maybe not toolroom but high precision expected. 45 years in I still haven't the experience to use what I have to its full abilities.

When you have a good DRO set-up and reliable dials the approach is different. Much closer to the old production line operator working to stops on the machine set by the tool setter. Except you are both tool setter and operator. The basic modern way is to take a cleanup cut, set DRO to zero, measure, work out how much there is to come off, reset DRO so it will read zero when the job is to size and have at it with a check measurement before the last cut on the first part. It should come out right. If not adjust the final zero position and all the rest will be right. If they are not then more practice needed. Learn to love the many memory registers on the DRO.

I don't have a DRO on either lathes so I set the dials to be zero at finish size. Same thing, different equipment. If I did have a DRO I'd still set the dials too as I find it much easier to control cuts by handle angle rotation rather than watching flicking numbers. My view is dogtal is great for getting start and end points, not so good running in between.

Rant mode on. Why can't I get a DRO that makes it easy to connect a switch integrated into a stop so the machine turns off auto feed or threading when the DRO hits zero!

Clive

[Michael Edwards 1Participant @michaeledwards1]

Gents thank you very much for sharing your knowledge and thoughts. I feel alot more educated and happier now with my machining.

Ive been told many a time that "I dont just scratch my a**e hole, I tear a new one" and I suppose its true. Because I never know when to stop. I'm very rarely satisfied with anything I do I always pick faults in it.

Thank you all again

Regards and stay safe

Mike

[Anonymous]

Posted by JasonB on 27/05/2020 07:35:08:

………………No Idea

Ok, thanks. Seems a bit rude to me, even if I agree with the underlying point. But what do I know!

Andrew

[Author]

Posts