Another Rupnow 1 by 1

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.
Bob - slightly off topic - but you did say that statistics was your middle name.
I once visited Fichtel & Sachs and queried a Vickers microhardness check which they had placed on a report as 1052.1 - I said that such accuracy wasn't possible.
Turns out I was wrong - they calibrated each and every one of their inspectors against known standard samples - so they had a statistical offset and error range for each inspector.
By having 5 persons taking a reading they could then statistically masturbate the data into a more precise answer.
Trust the Germans to be so precise as to "calibrate" their people.
Hence my comment - and K2's valuable suggestion to calibrate yourself against a known standard.
Regards, Ken
Edit-P.S. as a reference use a ball-bearing of the same I.D. they're generally pretty accurate.
 
25 years setting standards in Quality has rubbed off on me then. We calibrated all our inspectors against each other and international other plants and Companies. Most suppliers had calibrated inspectors as well. We checked data and regular test results during audits, and I only ever found 1 plant (Bosch = German) that was rubbish! Never Italians, French, Spanish, etc. - they always told the truth! The Germans wanted to sack the poor sod, I said we would sack the as a supplier with that sort of attitude, so the MD agreed "re-training" and "re-calibration" was the answer - for the inspector, his supervisor and the Quality Manager! Zeig! ...
re: my valuable suggestions - please re-train me when I get it wrong!
re: "use a ball-bearing of the same I.D. " - use a bore from a ball race as a calibration bore as well...
K2
 
25 years setting standards in Quality has rubbed off on me then. We calibrated all our inspectors against each other and international other plants and Companies. Most suppliers had calibrated inspectors as well. We checked data and regular test results during audits, and I only ever found 1 plant (Bosch = German) that was rubbish! Never Italians, French, Spanish, etc. - they always told the truth! The Germans wanted to sack the poor sod, I said we would sack the as a supplier with that sort of attitude, so the MD agreed "re-training" and "re-calibration" was the answer - for the inspector, his supervisor and the Quality Manager! Zeig! ...
re: my valuable suggestions - please re-train me when I get it wrong!
re: "use a ball-bearing of the same I.D. " - use a bore from a ball race as a calibration bore as well...
K2
I don't quite get this.
 
Calibrating inspectors against standards, and across companies, makes all the products the same - irrespective of the plant.
The Germans didn't like that idea, if a guy met the standards - OK. if not they sacked him. They believed they were the best - but were not - according to their own results compared to others!
Manufacturing guys have to make 99.997% of stuff right. Inspectors equally have to find 99.997% of faults in products. So during a shift, they would be fed "known" faults at random and had to be 100% OK on finding them. E.G. during visual inspection of dye penetrant crack detection of forged parts.. It worked and the warranty failures dropped to Zero.
C'est la Guerre..
K2
 
Statistics is practically my middle name, but I'm not comfortable using them here. Let me show you why. Two batches of ID measurements at the non-chuck end of the cylinder.

snip

Since the bore is obviously close to 0.950 and my target is 1.000, I can take off another .040 and then repeat all this. I just don't like the spread in values test to test. Also, this design calls for a Vyton piston ring instead of cast iron. I tend to believe that makes the real application a bit more forgiving.
Hmmmm - - - - you're overthinking this - - - agonizing over tiny amounts of stuff that -- - - - well - - - at this point - - - just don't matter.

Rough cut to -0.050" of desired size.
Usually this roughing is done in 0.100 to 0.200" dia cuts (radius of 0.050 to 0.100" have done up to 0.200" radius but that's on more stable machines!!).
Get a quick measurement +/- 0.001 is quite good enough!!!!!!!
Now you are finish cutting - - - may need to change your tool - - - - I did (roughing and finishing are 2 different steps!!).
Do one light cut - - - say of 1/2 the dia left to cut (0.010" is maybe too light but not more than 0.025" radius).
Now is when you measure the crap out of the piece.
If you have a taper - - - note it.
Mark out what you're going to do (mark on the lathe bed for boring every 0.001" change or on the piece for turning).
Cut leaving 0.001" of material doing the cut as planned!!!
Now you have some fun sanding (or using a flap wheel on an air die grinder).

Doing this my worst project had the lathe giving me 16 thou change and I was within tenths at 0.002" too big.
Should have shot for 0.001" too big - - - there was a pile of sanding.
Bearing fit over 14" was the result with -0.0002 to -0.0004" on final size - - - just what the boss wanted.
 
How did you cut it? Did you use a reamer? I never completely trust those telescoping ID gauges. They require the use of two tools, each tool representing another layer of error. Do you have any direct reading micrometers for inside? Since you still have a way to go, why not cut out .030 and then ream it?

I don't trust reamers!!!!!!!!!!!!! I've even had one cut a thou under size - - - over being more common.
It is possible to trust telescoping gauges - - - - but only with regular use.
Developing the feel first takes practice and to maintain the skill also takes practice.
So practice makes perfect.
Direct reading mics are fine if your bore is short - - - if the bore is say 3.50" rough and 8" deep your direct reading mics - - - - well they're nice ornaments.
You can buy yourself a dial bore gauge - - - - but if you're only doing occasional work - - - - they're not worth the $$$$$$$ and they ain't cheap!!!
 
I don't trust reamers!!!!!!!!!!!!! I've even had one cut a thou under size - - - over being more common.
It is possible to trust telescoping gauges - - - - but only with regular use.
Developing the feel first takes practice and to maintain the skill also takes practice.
So practice makes perfect.
Direct reading mics are fine if your bore is short - - - if the bore is say 3.50" rough and 8" deep your direct reading mics - - - - well they're nice ornaments.
You can buy yourself a dial bore gauge - - - - but if you're only doing occasional work - - - - they're not worth the $$$$$$$ and they ain't cheap!!!
Yes, you are rite. however, it's better to be a single thou under than a thou over or maybe even worse, two thou under. Course you might try a second reamer. I read that a brand new reamer often is half a thou OVER but eventually gets down to the correct size, so a used reamer may be better than a new one. Such is life.
 
HI CFLBob !
Now what is the error between cylinder diameter measurement positions ?

Most accurate answer: I don't know.

I had chores to do yesterday and didn't work on it, but here's every measurement.

Diameters.jpg
 
Sorry, Minh Thanh

The cylinder is going to be 2-3/8" long when it's finished.

The column that says outer end is where the cylinder head sits, the very top of travel.

The 2" deep in cylinder isn't really 2". It's more like 1.6" to 1.7" mostly because my telescoping gauge won't reach all the way in. The piston has a 1" stroke and is 0.96" long so it's around where the bottom of the piston ends up at the bottom of its stroke.

I wanted to measure top to bottom of the piston stroke, but my telescoping gauge isn't long enough.
 
Hi !
Have you checked the vibration of the chuck?
Maybe the bearing is worn, what bearing does the chuck shaft use ?

The holiday preparations here have overtaken the work in the shop and I haven't spent a minute on it since the original post.

I can see or feel no looseness in the bearings, although I haven't tried to measure it. My lathe is turning six years old in January, so it's not that old and gets light use. I'm not saying the bearing can't be bad, just that I think the problem is coming from me, not the lathe.
 
The holiday preparations here have overtaken the work in the shop and I haven't spent a minute on it since the original post.

I can see or feel no looseness in the bearings, although I haven't tried to measure it. My lathe is turning six years old in January, so it's not that old and gets light use. I'm not saying the bearing can't be bad, just that I think the problem is coming from me, not the lathe.
THis begs the question: did you check the levelness yet?
 
My technique is to lay this on the side of the cylinder so that the head is on top of the last circular flange, then grab the knurled part of the handle with a pair of long needle-nosed pliers (it's just left of the non-knurled part on the end). Then I put it in place in the cylinder still holding it in the pliers, with the pliers flush with the end of the cylinder, and unscrew the clamp holding the movable pin in place. Jiggle it a bit and then tighten down the screw to hold the position - all the while trying to preserve that angle so that the measuring head is perpendicular to the walls of the cylinder.

Bob, you are using the wrong technique, and will definitely get variations in your readings. Ken I's technique below is correct - you deliberately "cock" the gauge off square and let the ends contact the cylinder walls. Tighten the clamp, then tilt the gauge through square and beyond, and remove it. The clamp mechanism allows the fingers to be compressed as this happens, but does not allow them to spring back out.

Bob, This is the way I use a telescopic gauge :-
Insert it deliberately skew and very lightly nip it up - then tilt it into alignment (going to be a bit more difficult with your "bent" gauge) - the telescope will collapse - when you feel it is a good fit - tighten more solidly.
Re-check the fit in the bore - a very light interference - tilt to remove and check it with a micrometer.
Repeat several times at the same spot to "calibrate" yourself.
Regards, Ken
 
Bob, you are using the wrong technique, and will definitely get variations in your readings. Ken I's technique below is correct - you deliberately "cock" the gauge off square and let the ends contact the cylinder walls. Tighten the clamp, then tilt the gauge through square and beyond, and remove it. The clamp mechanism allows the fingers to be compressed as this happens, but does not allow them to spring back out.

Thanks, Andy,

Crossing my fingers that I get a couple of hours tomorrow to go back out to work on this. I'm all but sure I'll have time on Christmas day.


Bob
 
Hi Bob, I just tried a quick demo of the idea for using a side-lever DTI for a comparitor against the micrometer. Easier to show pictures...:
The internal calipers I use, with a large diameter thumb wheel for very fine feel. I find these much better than bore gauges for less than 1" bore.
View attachment 132038
Another alternative to your bore gauge:
A side-lever DTI with a scriber held on the back with a rubber band. - If properly fixed, and adjustable, this could serve as a bore gauge comparitor for use against the micrometer.
View attachment 132039
View attachment 132040
The next picture shows how you would set the micrometer to the size to be measured, then set the DTI Bore gauge to zero aganst the mic.
View attachment 132041
Then it can be used as a comparitor to measurethew undersize or oversize of the bore. = becomes a direct reading bore gauge!
While I have only shown the scriber affixed with a rubber band, there is no adjustment of the gauge to various dimensions. If you made a small clamp-on fixed point that was adjustable with a screw, and locknut, then you would have a more secure but versatile measuring gauge. - This DTI measures to 0.0001" per division, so you can even use it for honing to size, where you may only be removing a fraction of a thou per pass of the hone.
But I would recommend you try and reset the lathe to be closer to a parallel bore before you finish boring the cylinder, anyway.
Cheers!
K2

K2

After messing around with my telescoping gauge a bit more, I figured it was time to get a replacement. I could get another used Starrett gauge or an internal caliper like you showed (except I'm guessing you made that knob yourself). I found a used Starrett internal caliper on eBay, and got it yesterday.

This is going to be a whole new learning curve. While it seems easier to put in the center and find the diameter rather than some other distance, when I adjust my micrometer ratchet, it closes the jaws and wipes out the measurement. I've always heard to use the ratchet knob to prevent that, but I don't see how that's going to work.

More study.
 
K2

After messing around with my telescoping gauge a bit more, I figured it was time to get a replacement. I could get another used Starrett gauge or an internal caliper like you showed (except I'm guessing you made that knob yourself). I found a used Starrett internal caliper on eBay, and got it yesterday.

This is going to be a whole new learning curve. While it seems easier to put in the center and find the diameter rather than some other distance, when I adjust my micrometer ratchet, it closes the jaws and wipes out the measurement. I've always heard to use the ratchet knob to prevent that, but I don't see how that's going to work.

More study.
I've always had that problem with my telescopic gauges. As a result I don't use the ratchet, instead I had to develop a 'feel' for the pressure on the gauge so as to measure repeatably. If you collapse the caliper even slightly you should perceive a sort of 'mushiness' in the feeling of twisting the micrometer.
 
Back
Top