Workshop working tolerances

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Dave C on 03/09/2018 09:42:58:
.

Any recommendations for a good gear cutter anyone?

.

This doesn't qualify as a recommendation, Dave, but it may be worth giving Leek Gears a call: **LINK**

Take a Look

Family name = Leek

Location = Leek

Always looks 'proper' when I drive past, but I've not yet had the need to use them.

MichaelG.

[MWParticipant @mw27036]

Generally speaking it's not actually that bad to control sizes, especially if you already have what you want to fit it to, to hand.

For example, you're machining a bar to fit into a cylinder of sorts.

You don't really even have to be concerned with numbers too much at that point. Just creep up on the nominal size that you know the other is and keep checking it by hand against the matching part after each and every light cut.

You can then go from either a force fit to a push fit to a sliding fit, if it feels a bit tight to begin with, just let it wear into it for a short while.

This is why you didn't have to be a walking calculator to successfully manage sizes years ago.

It obviously changes and gets tricky when you want to make lots of bits for parts that you don't have to hand, that repeatability is what industry continually wrestles with.

There would be people who really don't like that idea but in actual fact this is the way craftsmen used to work, by feel and intuition. 

Michael W

Edited By Michael-w on 03/09/2018 12:15:53

[colin hawesParticipant @colinhawes85982]

Tolerances are for making spare parts interchangeable or special fits such as force fits where a critical difference between parts is necessary so actual precise dimensions are unlikely to be essential on models. Colin

[Neil WyattModerator @neilwyatt]

Posted by Michael-w on 03/09/2018 12:08:57:

There would be people who really don't like that idea but in actual fact this is the way craftsmen used to work, by feel and intuition.

It's how Rolls Royce used to build Merlins until Packard rejected their tolerances as too broad.

Contrary to the urban legend, RR were actually grateful for Packard's input.

Neil

[DanielParticipant @daniel]

Builders work to the nearest CM.

Furniture makers to the nearest MM.

Boat builders to the nearest boat.

[Dave CParticipant @davec87625]

Just as an update and also possibly help to others.

I eventually found a company nearby that was prepared to take on the cutting of the 2 gears for me.

They were ready for collection within 2 days and finished to a very high standard. The service and assistance given was superb as was the price.

The company is a medium sized business and is happy to carry out small / one off jobs.

I'm not sure if the advertising will be approved by the moderators but the company is Northern Broaching Services in Radcliffe, Manchester. Contact details are available on the web.

Hopefully this may be of some help to others in the future for gear cutting or more complex machining.

I have no connection with the company other than that of a very satisfied customer.

[BufferParticipant @buffer]

Heres a sort of relevant question. What is the best way to turn work to your chosen diameter without sneaking up on it taking smaller and smaller cuts. Which is what i seem to end up doing? And it takes an age. And dont say get a DRO!
Thanks Rich

[not done it yetParticipant @notdoneityet]

Take bigger cuts, but measure how much is removed for x divisions on the machine scale. Each cut of the same size will remove the same amount, so at least the last two cuts need to be about the same number of divisions on the dial, taking into account whether one or two divisions more, or less, for that last cut.

If you need it any closer than 0.01/0.02mm, cut slightly oversize and finish with abrasive.

With practice, and confidence, it works.

Even with a dro, one cannot expect a 0.02mm (or less) cut to be accurate immediately after a 0.2mm cut, or more, if your tolerance is really tight.

Edited By not done it yet on 25/09/2018 19:43:55

[Clive FosterParticipant @clivefoster55965]

Richard

Best way to work to chosen size is sharp tools, trust the feed dials and take a decent finishing cut. Anything under 5 thou / 0.1 mm is too little. Very much in the lap of the gods as to exactly what the cut will be. Especially on a small and, inevitably by absolute standards, floppy machine with a tool of unspecified wear.

Accurate sizing requires a stable cut which can only be got with significant load on the tool. To quote Dennis Turk lathe rebuilder extraordinaire "Everything is made of springs.". If your finish cut is too much smaller than your main working cut changes in load can alter the relationship between dial and actual feed. I habitually cut at 50 to 100 thou for main metal shifting and consider 20 thou a decent finish cut. Thats on a Smart & Brown 1024, a massive beast by home shop standards and very stiff. Even so I'd not put money on a 1 thou cut taking 2 thou off to within a couple of tenths. Usually doesn't mean always. On a 20 thou finish cut if I'm more than a few tenths out either something is wrong or I've been careless.

Clive.

PS :- "not done it yet" types faster!

Edited By Clive Foster on 25/09/2018 20:04:23

[BufferParticipant @buffer]

Thanks, being self taught it never occured to me that a fine cut after a bigger cut can cause a problem. I always thought you got the accuracy from trying to take a tiny one at the end but that never seemed to work that well.
Thanks again
Rich

[Neil WyattModerator @neilwyatt]

Posted by Richard brown 1 on 26/09/2018 13:29:17:
Thanks, being self taught it never occured to me that a fine cut after a bigger cut can cause a problem. I always thought you got the accuracy from trying to take a tiny one at the end but that never seemed to work that well.
Thanks again
Rich

If you want to finish with a small cut, run the cutter past again on the same setting after the last big cut. This 'works out the spring in the tool'.

I prefer to take a large cut but not so much as to compromise finish.

HSS and ground finish carbide (GT) are good for fine finishing cuts. Moulded carbide (MT) gives the best finish for cuts of at least about 2/3 the tip radius.

Neil

[Jeff DaymanParticipant @jeffdayman43397]

Thousands of years ago when I was getting my first formal machine tool training I was always taught to plan the cuts in the lathe so the last cut was at least 20 thou, right to finish size. The instructor was always yelling "don't take skin cuts!" meaning cuts of a few thou. He was an ex British Army armourer, and could make anything from forged chain to a pocket watch, he was extremely skilled.

Mind you this was on large Colchester and Hardinge lathes, with tremendous rigidity. On smaller less rigid lathes you might need to reduce cut depth slightly, or reduce feed per rev. I have always followed the advice of "no skin cuts" and got good results most of the time. Some steels are such that it is very hard to get a good finish though, and in that case a grind or hone or lapping op may be required to reach desired finish at size. In free machining steels, take a big final cut right to size.

For a beginner I'd suggest learning with HSS tools, then much later, experiment with carbide tools, particularly on small lathes.

Just my $0.02 worth, I'm sure many dissenting opinions will follow as usual. Don't be afraid to try various tools and steels, draw your own conclusions, and practice practice practice.

[Tony Pratt 1Participant @tonypratt1]

Best advice I can give is rough with one tool & finish with another nicely sharpened tool, you won’t get a decent size & finish just by using one tool for both tasks, and this applies to milling also.

I also don’t agree with large finish cuts using HSS, carbide is a different story, but hey what do I know?

Tony

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[Neil WyattModerator @neilwyatt]

For a long time I advocated the HSS first approach.

I now use inserted carbide for almost everything, and only use HSS when I haven't got a carbide tool the right size, such as screwcutting.

HSS is less stressful for beginners as you don't need to work it hard to get good results.

The downside of this is that when you change to carbide you either get poor results or chip inserts due to pussyfooting around with them. They do like high speeds and generous cuts. Some people won't persist long enough to make the transition and go back to HSS.

My experience of uncoated CCGT inserts on aluminium, brass, stainless and steel has changed my mind. These can cope well with HSS-style cuts but also allow you to turn up the heat. I recently bored out apart in 6082 using 2mm cuts for roughing and finishing to size with less than 0.01mm cuts with the same insert. That's a 200:1 cut depth ratio. The light cut swarf came off like a spider's web.

I would now advise beginners to consider starting with the CCGT-type inserts, either from the start or to transition from HSS.

Once you are confident with carbide, CCMT inserts are great for ripping up steel and inserted carbide parting tools are great too if you have power feed or a steady hand and a steady nerve to keep up the feedrate.

Neil

[Jeff DaymanParticipant @jeffdayman43397]

Posted by Barrie Lever 1 on 26/09/2018 18:49:18:

Posted by Jeff Dayman on 26/09/2018 14:16:15:

For a beginner I'd suggest learning with HSS tools, then much later, experiment with carbide tools, particularly on small lathes.

Jeff

What is your reason for HSS tools in small lathes? I always get best results with moulded carbide and I do know how to sharpen HSS.

Just asking.

Best Regards

Barrie

Hi Barrie, I am basing this comment on my experience teaching a number of people over many years the basics of lathe work. Many have bought carbide tooling and struggled mightily with it mainly because their small or miniature lathes do not have the power or rigidity to use carbide tooling, or the setup is wrong, or they have bad Chinese carbide tools from a local discount tool firm (horrible stuff). Several of these guys were at their wits' end – could not make a single good cut. When they changed to HSS tooling and learned to sharpen and set it up correctly, the results were shocking to them. They actually got great results nearly every time. As they practice and get confidence based on their successes they often come back to carbide, but with the knowledge that if they run into difficulty they can change tooling and get great results with HSS. Many like the low cost aspect of HSS too. The main reason I steered them toward HSS was that in many cases the carbide tooling or the POS low cost import lathe or mini lathe that they had are not fit for purpose. Rather than spend energy helping them fight with the two main cheapass uncooperative vendors here in Ontario Canada I thought it would be better to show them how to use HSS to get usable satisfying results on whatever barely functional low quality lathes they have. Again, just my $0.02 worth. Your mileage may vary. I've used carbide in industry with great success including specially designed carbide composite tooling for high volume water valve body manufacturing, so I'm no stranger to it. But that experience was with good quality industrial grade carbide tools, and large high quality lathes, not some of the rubbish the hobby vendors sell people here (I've seen some good lathes from these vendors, but some really awful ones. All the little $750 red mini lathes I've seen from one particular vendor are absolute crap and needed a lot of work before turning anything.). On my own old 9" South Bend lathe I have used carbides with some success, and do for semi hard steel work, but I find the speed and power on my lathe is really too low to get the best out of carbide tooling, so I find myself mainly using HSS still. The South Bend works great with HSS and a simple lantern toolholder gives huge flexibility and fast setup. Yep, it's old fashioned and crude, but it works for me. I'm sure lots of people will find fault with this, as usual. I get results though, and am happy with it.

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[John ReeseParticipant @johnreese12848]

Once you have it assembled no one will be able to tell if things are off size as long as the mating parts fit properly.

Oh, wait! You told about t here.

[Sam StonesParticipant @samstones42903]

I can’t be sure of the story that James Watt checked piston to cylinder clearance with the thickness of a gold (half?) sovereign.

However, years ago at a model engineer club meeting near Melbourne, one old chap declared …

“I don’t know about all this tolerance stuff."

"When I was a lad, we had to get it right.”

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Sam Stones on 27/09/2018 00:36:30:

…

“I don’t know about all this tolerance stuff."

"When I was a lad, we had to get it right.”

.

What an absolute gem, Sam

Thanks for sharing that

MichaelG.

[Sam StonesParticipant @samstones42903]

I thought so too, Michael.

My apologies to Dave C, for not being all that helpful.

Sam.

Sunshine in Melbourne for a change.

[not done it yetParticipant @notdoneityet]

Posted by Sam Stones on 27/09/2018 00:36:30:

I can’t be sure of the story that James Watt checked piston to cylinder clearance with the thickness of a gold (half?) sovereign.

However, years ago at a model engineer club meeting near Melbourne, one old chap declared …

“I don’t know about all this tolerance stuff."

"When I was a lad, we had to get it right.”

The point is, If you are making one-offs it doesn’t matter if the parts are slightly under- or over-sized. BUT, the part made second must fit the part made previously so machining needs to be done to a tolerance for clearance. Things can be made to work (less than ideally perhaps, if the piston is undersized) but not if it is of larger diameter than the cylinder it should fit in.

The machinist obviously needs to plan which part is easier to adjust – piston could be reduced a tad or the cylinder enlarged, but one component may be needed to be remade. If the fit is beyond redemption, it is a case of deciding which is the more difficult/expensive to replace and not doing other work, on a piece that could be rejected, before being sure the important dimension is satisfactory.

Piston rings must fit the bore diameter, not necessarily the piston, to provide a satisfactory seal so a slghtly undersized piston could be acceptable (but obviously not ideal).

If serial production is required, both parts need to be made such that any combination of parts will fit without more turning, at least (parts may need some minor adjustment(s) by the fitter). That is where tolerances are more important. But this thread was less of tolerances, but more of how to achieve the final size. Creeping up on it with successively smaller cuts is not the best way to achieve this, practically. Yes, it works, but is more time and resource consuming. Three equal cuts, confirmed by careful measurement after each, should result in the final cut being a good one.

The turner may find the first finishing cut is not quite the same as the second because of the roughing cuts affecting the surface finish. Temperature is also important and should not be ignored – cutting a hot work-piece carefully to close tolerance may result in an under-sized part when cool.

All these things need to be considered and planned for. Experience will eventually lead to confidence. There is no substitute for experience.

[Author]

Posts