machining accuracy

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Loose nut

Well-Known Member
Joined
Dec 31, 2007
Messages
366
Reaction score
2
What do you consider to be an acceptable level of accuracy for model making? I have always strived for .001" as a standard where it counts, some times I make and ......

In the golden olden days of Model Engineering when the main measuring tool was a ruler marked out to 128 Th's and parts where made to fit each other, 5 or 6 thou was considered good but we now have much better tools today.

The pros. keep talking about finishing parts to a couple of tenths of a thou. which is beyond my capability so what is a good and reasonable level of accuracy to try and achieve.
 
Loose nut

Invite any one of the pros to your shop to show you how they consistently hit that thenth
of a thou. size. I doubt you'll find many who have the time to come and show you. ;)

If it's straight and on size within .001 for me, it's Dead Nuts on!

If it actually needs to be closer than that I'll intentionally cut it +.001 and carefully hand
polish it to the perfect size.

Rick


 
Loose Nut: Along with machine capabilities, there are a lot of other factors.
How fussy you are and how much pride you take in your work enter into it.
Do you have to have parts made to .001" if they work just as well made to .010"?
Will your friends be impressed if you tell them every part in this engine is made to .001" tolerance?
My friends just look at a running engine and say "That's nice, what does it do?"
My practice is that when fitting a bearing race, .001" is no where near close enough but when making a 3" diameter flywheel .250" is close enough if that is the material I have.
Everyone makes his own choice. Mine is the journey, not the destination.
 
i have very good friend that started building a small up right steam engine about 15 years ago and he is about half done.
i asked him why it is taking so long and he said that every thing is machined to .0001" tolerance :eek:

i said why would you work to that tight of tolerance and he said that that is the tolerance he works to at work.

i then proceeded to explain to him that you have to look at the parts you are making and determine what the part does and does it have to be that precise.

like stan said in the previous post that if the flywheel is plus or minus .250" who cares if it will do the job then it must be ok.

anyway 2 weeks latter my friend finished the engine and now he is working on another steam engine model of some sort and he also told me that he is now having fun making his models. i guess it is because now he is not treating it like his job.

chuck
 
Loose Nut,

On general blueprints, if no tolerances were shown you would make the part +/- 0.002".

In little engines, you would aim for getting it as close as you can, and if the engine runs, stuff the tolerances. Lives are not dependant on little engines, trying too hard can spoil all the fun.

If the bore size on the drawing says 0.500", and you made it 0.0510", by the time you made the piston to fit the hole, it is almost guaranteed to run.

Unless the drawings call for super fine fits, just do your best on fits and finishes.

John
 
Bogstandard said:
If the bore size on the drawing says 0.500", and you made it 0.0510", by the time you made the piston to fit the hole, it is almost guaranteed to run.


John

I'd say he would have to go abother .449" to get there. ;D Or make a darn small piston of .050"

Sorry John, I just couldn't resist. :-*

Bernd
 
Sorry about that, it is the middle of the early morning here, half asleep.
I will leave the error in, just so your post means something.

What a gentleman I am.

John
 
In the golden olden days of Model Engineering when the main measuring tool was a ruler marked out to 128 Th's and parts where made to fit each other, 5 or 6 thou was considered good but we now have much better tools today.

they were making them perhaps without aid of a micrometer but the does not mean their standards of fit were any less than today. There are many comparative technique to create extremely accurate fits. On the contrary, i think you could make an argument that the more gadgets the newbie is equipped with today the lower the standard of work....why? they never learned the basics, filing, scraping, general bench work, layout etc whereas is less equipped learned many of these skills because there was no other way to do things. I haven't done it, but don't think it would be too difficult to fit a piston to a bore with less than a thou clearance without a micrometer.

Note that in making a one-of, its fit that matters not hitting a nominal dimension. if the the goal is 1 thou clearance between piston and cylinder that's what counts not whether the final bore is 1.000" or 1.005". In fact even with all of our fancy equipment today, its still good practice to turn the piston to the bore.

Invite any one of the pros to your shop to show you how they consistently hit that tenth
of a thou. size.

and

If it's straight and on size within .001 for me, it's Dead Nuts on!

there's, imo, several stops along the accuracy track between these two statements. Within a thou isn't good enough in many situations, yet you better be running a hardinge or monarch in temp controlled environment and have some very nice measuring equipment to consistently claim to hit a tenth. My best digital mic measures in tenths - if tenths is its best resolution, how can someone claim to hit a tenth without a mic that measures to even a smaller increment ...and does so accurately. show me a guy in his garage or even the average job shop running around with mic that accurately measure to a fraction of tenth ::)

However when it really really matters, like getting the proper press fit on something or housing or shaft for a roller element bearing, you need to machine to better than a thou.... a couple of tenths is a reasonable target imo....when you have to (as per chucks point).... also we tend to work with smaller stuff so clearances and tolerances are smaller.

My German tool and die maker turned high school machine shop teacher would unleash a tirade for filing in the lathe and emery was for finish not fit. Granted he was trying to train us, but the point is valid, once you start what that stuff you've no guarantee the part is round or straight. I do use fine emery to take off the minute burrs and ridges if its a moving part, but I've never lost what that very skilled and disciplined man taught us.....and those abrasive techniques are useless for bores unless you want a broken finger

Chuck point is right on - far more germane to us that how accurate you can machine to is understands when and what accuracy is required. There are reasons model engineer drawings don't usually have tolerances or even clearance specified. Its up to us, acting as engineers rather than by rote following a tolerance schedule to understand what is required and machine to that - i like the engineers expression: Good enough is good enough! (though good enough is often in the eye of the beholder when we're talking model engines)
 
Note that in making a one-of, its fit that matters not hitting a nominal dimension. if the the goal is 1 thou clearance between piston and cylinder that's what counts not whether the final bore is 1.000" or 1.005". In fact even with all of our fancy equipment today, its still good practice to turn the piston to the bore.

Its up to us, acting as engineers rather than by rote following a tolerance schedule to understand what is required and machine to that - i like the engineers expression: Good enough is good enough! (though good enough is often in the eye of the beholder when we're talking model engines)

YES!!!

When in industry, if it was required I had to have a tolerance finer than 0.001", it was then ground to dimension, much more controllable.

I have actually seen, when I visited The Dowty Rotol Company, flat faces being reworked by disabled people (they seemed to have a knack of patience and tolerance for this sort of work) for use in the aircraft industry, using precision glass blocks laid on the surface with a light shining thru it from the end. The light was distorted slightly by the uneveness of the surface. The offending area was given a very light rub on a piece of paper to minutely wear away the surface until the light showed parallel lines running thru the glass block. The glass block was the rough stuff, they ended up under DTI's that had a face diamenter of about 24". Their tolerances were 0.00001" or better. Slip gauges were just chunks of 'near enough' metal compared to what they were doing.

John
 


Thanks for the input, I have been making things on a fit to fit basis but have wondered if I should have been more conscientious of tolerances, a matter of prospective I guess.

I know that model loco. builders make everything to a rattling good fit they, have to be a bit loose to run properly.
 
At work this morning I finished up an order for 20 pieces of a part.
The OD of the parts was 9" something with a .0005" tolerance and a
finish requirement of 16RMS.

These are OEM parts and ANYTHING out of tolerance is immediately scrapped.
Scrap enough of them and YOU may be scrapped!
Add into the mix you need to meet the quoted time for the job.
Often those times are estimated by someone who has never...........
Oh I'll leave that dog sleep... LOL

They were all good this time around.
At least two of them were brought to size and or finish with emery cloth.
I don't know how many of the pieces my buddy made needed to be finished
in that manner.

I don't bring that kind of stress home with me.
There's no need to.
When your building an engine or anything else for that matter, you know what has
to fit and in what manner.

If you miss a size the mating piece size can be altered to accept it.
If you miss that one as well, pick the easier of the two to make and try it again. ;)

Rick


 
My Taig and Unimat are capable of .0005" accuracy. (That word seems to have different meanings to different manufacturers.) My watchmaker's lathes are specified to at least .00005" and one's half that. Do I work to that level? Heck no! I may be crazy & I may be insane (and I have the paperwork to prove both!) but I am not masochistic. How close do I go? Whatever it takes and no more. I may push a bit to get as fine as possible but that's just to push myself, not for any practical reason.

Bogs, that light refraction technique sounds cool. .00001" ain't too shabby either. :)

Not anywhere near in the same class, I often use glass plate salvaged from an assortment of sources for the final polish on some things, starting with "scary sharp" for things that need to be sharp to a bright finish using diamantine and other fine abrasives.

What's this have to do with machining accuracy? Aside from a high polish, this approach can be used to "adjust" that last teeny tiny bit for whatever purpose. For me, it's part of the self-challenge, hence it's play.

And Bogs said ...

In little engines, you would aim for getting it as close as you can, and if the engine runs, stuff the tolerances. Lives are not dependant on little engines, trying too hard can spoil all the fun.

That's pretty much it ... keeping track of the dismal point - er, decimal point - in the process. ;D

Best regards,

Kludge
 
I would just like to amend the quote that Kludge picked up on.

Most of the time, the above rule can be applied, but there are times that tolerances have to be kept very tight. In those situations, it usually takes a bit of experience in fits and finishes, coupled with a bit of tedious elbow grease. Good examples are flame lickers and Stirlings.

Other than that make it so that it runs.

John
 
What annoys me the most is having a part that is 2 or 3 thou. over size and taking another cut without changing the settings on the cross side, just a cleanup cut to take out the spring of the tool and bring it to size and it usually ends up 2 or 3 thou. under size.

I usually true up the part in question while it is still 15 or 20 thou over, mike it and start the finishing to size but that last little bit.............!!!!!!!!!!
 
Kluge:

My watchmaker's lathes are specified to at least .00005" and one's half that.

I have never worked with watchmaker lathes so I am curious as to how the accuracy is determined. Do the handwheels have less backlash than .00005" or will it turn a shaft sticking out of the collet a couple of inches to that tolerance?

Many years ago, I bought a Boxford (British Southbend) that came with a final inspection sheet that listed the deviation from zero in all possible movements of the tools. Does the same type of sheet come with a watchmaker lathe?
 
Loose nut said:
What annoys me the most is having a part that is 2 or 3 thou. over size and taking another cut without changing the settings on the cross side, just a cleanup cut to take out the spring of the tool and bring it to size and it usually ends up 2 or 3 thou. under size.

I usually true up the part in question while it is still 15 or 20 thou over, mike it and start the finishing to size but that last little bit.............!!!!!!!!!!

Lose nut, sounds like you're close, but taking a little too much depth of cut too late in the process. Back up so you have a couple shots taking very light cuts to get there. You'll want to experiment with your lathe to see what DOC it can be rigid enough to be accurate at might be. For my lathe, I shoot to take heavy passes until I'm within 0.010", and then 2 very light passes after that. It's nearly always dead on at that point. You need one spring pass that you know isn't going to take you over so that you've got an accurate read on where you are.

Cheers,

BW
 
Assuming that there's no free play in your slides Loose nut, do you give the tool a quick lick with the stone before the final cut just to get the edge back ?
Regards Ian.
 
Circlip,
Please expand on that. What are your procedures in sharpening and maintenence of the cutting tool?
Thanks.
Regards,
Fred
 
Stan said:
Do the handwheels have less backlash than .00005" or will it turn a shaft sticking out of the collet a couple of inches to that tolerance?

I've never seen a cross slide with that little backlash (but then, all the ones I've owned have been very used and slightly abused) but the spindle sideplay is limited to that and is fairly easy to adjust since the watchmakers used to do so fairly regularly to keep them in true. In theory, that should include the collets which, in theory, should be bored true. I've never seen anything to support that they are or that they aren't. I do know that they are capable of a lot more than I can imagine using. Take a look at a mechanical movement for a lady's watch sometime and remember that they used to be hand made and repaired using these machines.

Of course, with any lathe, the effects of the sideplay are multiplied the further you get from the headstock so, as a rule, everything that could be done at the headstock was done at the headstock, and with a graver rather than a cross slide-mounted tool as often as not. (Reflective thought: probably more often in the original intended use.) Note that the cross slide on a watchmaker's lathe has no lead screw to move it the length of the bed since they were never intended to be used like the cross slide on a larger lathe.

Many years ago, I bought a Boxford (British Southbend) that came with a final inspection sheet that listed the deviation from zero in all possible movements of the tools. Does the same type of sheet come with a watchmaker lathe?

All of my machines came out of estates and closed shops so there was no paperwork. I've never even contemplated having the idea of fantasizing about a new lathe due to the expense, although there is a Chinese one that has my eye even though it's on something closer to a Geneva bed rather than a more sturdy WW bed. (I'm actually more interested in some of the accessories since this machine probably can't do anything more than what my 6mm can handle due to the bed design.) In any case, I have no clue what comes with them out of the factory.

Best regards,

Kludge
 
Sorry Fred, I'm " Old School " HSS tooling, not carbides. When you've done your "Cleaving " cuts, before you do the finishing cut you give the tool tip a quick hone with the tool still in the toolholder with a fine sharpening stone or one of those fancy DMT diamond hones to get a keen edge back on the tool.
PS. if you can get hold of some carbon steel tool bits, they keep a good edge for most of our applications.
Regards Ian
 

Latest posts

Back
Top