What's new
What's new

Scientific paper on accurate (proper) measurement of tight tolerance bores?

Liberty_Machine

Aluminum
Joined
Feb 28, 2017
I'm looking for any reputable argument for the accurate measurement for the proper method of inspection for tight tolerance bores.

Backstory; I'm working on a part that has two holes on one face. The tolerance of the holes are thus: .1251" to .1255". As any seasoned machinist or QC person would know, you can't fit a .1251" Deltronic gage pin into a .1251" hole. As such, I've been using a .12500" pin along with a Comtor Gage to verify hole size (hole is .250" +/-.005" deep). End customer says "uh-uh, needs to take a .1251 pin to be in tolerance".

Yeah yeah, customer is always right....right? But, I still feel like having this argument with them that they are incorrect. Losing battle?
 
Good to find the checking method the customer will use and match that. likely a simple DRO height gauge with a probe would do. I guess that device would be about $3,000.or so (?) common checking ability is:
• Resolution: 0.00001"/0.0001mm
• Maximum permissible errors (µm): 2+L(mm)/400
• Repeatability: 0.00004"/0.00008"• Motorized carriage displacement speed (mm/s): 150
with these you need accurately center to the bore..

The old hack way is to assume .0002 larger than the tried gauge pin, but the customer has to check the same way for that to be approved.

Air gauge is very accurate and can be used for many sizes, but most often you need buy a gauge for each size.
 
Well, yeah, customer is always right. But...
a .1255 pin won't fit in a .1255 hole either so just look at it as if the size is .1252/.1256.
 
@michiganbuck, the question isn't really "what can I use to check this", but: what is the proper (if there is such a thing) way to check this and is there any "white paper" on this from a body of credited manufacturing engineers?
 
What is the budget, what is the frequency, how many do you need to measure, and how long can you take for measurement? How quickly do you need to be up and running? Old school would have been air gage, I'm sure there are some digital solutions now. Where is CarbideBob?
 
Isn't this just basic gauge pin stuff? You get a class XX 0.1251 plus pin and a 0.1255 minus and there are your decision points. Sure you lose 40 millionths of your tolerance zone, but at least you can be comfortable with what you send out the door. And ease of use is hard to beat. As are costs.

Not familiar with the Comtor Gage and how that plays out referencing an off size pin.
 
I'm looking for any reputable argument for the accurate measurement for the proper method of inspection for tight tolerance bores.

Backstory; I'm working on a part that has two holes on one face. The tolerance of the holes are thus: .1251" to .1255". As any seasoned machinist or QC person would know, you can't fit a .1251" Deltronic gage pin into a .1251" hole. As such, I've been using a .12500" pin along with a Comtor Gage to verify hole size (hole is .250" +/-.005" deep). End customer says "uh-uh, needs to take a .1251 pin to be in tolerance".

Yeah yeah, customer is always right....right? But, I still feel like having this argument with them that they are incorrect. Losing battle?

Well it's your (the customers) money so... BUT I agree with you a .1251" pin WILL NOT fit into a .1251" hole. even if the hole is perfect, some clearance has to be allowed. I would ask them "what are the pins you would like to see fit "comfortably" into the hole?" .1252-.1254"...???

edit: No matter what research you bring to the table... it's their part, so do the best you can with whatever numbers they give you
 
I mean if it was me I would use go/no go pin gages at the drawing tolerance sizes and a bore gage to make sure there isn't any taper or a barrel shape. If you know your customer expects to use the .1251/.1255 gage pins, then you are good to go.
 
This is way outside of my area of expertise, but I Think the obvious questions would be; What does the Gage show the hole size as being, how accurate is the gauge, and how accurate is the standard used to calibrate it? If it reads exactly 0.12510 or even 0.12512 Wouldn't that mean the hole could be either in tolerance or out of tolerance depending upon which way the margin off error swings?

Also, how does curvature of the hole play into the measurement? If the hole has 0.0001" of curvature, is the "true" size of it an average based on the projected cross sections? If this were the case, the measurement of any given cross section would check ok and be round, but the hole would still actually be undersized wouldn't it? I genuinely don't know, some of this stuff is fascinating and out of my league.

I would think that opening up the hole a little so the go/nogo method can be used is the way to go because it takes you out of the realm of theoretical. From my point of view, the end goal is establishing the measurement with the highest level of certainty possible in the least amount of time

:EDIT

To answer your original question, I know brown and sharp literally has a published book about accurately checking hole positions, I think it includes info on checking size as well. If I can find out the name, I will post it.
 
A split ball (AKA split anvil) gauge of the type made by Diatest, but also sold by Mitutoyo, Mahr and perhaps others, certainly has the resolution you need. Diatest also makes an accessory that looks like a miniature arbor press that holds probe and that raises the part on a platform up to the split ball. It only would take a few seconds with a 0.1250" ring gauge to check the zero of the indicator prior to every part, followed by the part, with a reading on the indicator of anywhere from 0.0001" to 0.0005" passing, and anything outside that range failing. Further, if there was a variation of the diameter along the 0.25" length of the part, it would be apparent with this setup.
 
If the ball says pass and the pin does not who is to be the king of the world?
Here we get into the rabbit land of diameter and what is a hole.
Only fools or rookies run on the edge of a tolerance because you do not know better.
Bob
 
If the ball says pass and the pin does not who is to be the king of the world?
Here we get into the rabbit land of diameter and what is a hole.
Only fools or rookies run on the edge of a tolerance because you do not know better.
Bob

Straightness of the bore factors into the ball/pin question as well. A GO ball will go into more holes than an equal size GO pin.
 
We do work that often has .0002 tolerance total, including IDs / Bores. On larger sleeves that might be -0 / +.0003.

Our stated procedure on ID measurements that can take a Go / No-Go Pin check is as follows:

Class X Gage Pin or better. (required)
Go pin is - .0001 of low end, No-Go is +.0001 of high end.

A .1250 pin will NOT fit in a .1250 hole. A .1249 however will, with careful insertion because .0001 clearance is pretty small. If a .1249 pin goes into a hole, the physical reality is that the hole is something larger than .1249. Meaning, it's going to be REALLY close to the low end of the tolerance. IMO you have nothing better to make that check if you are using a Class X (40-millionths) or Class XX (20-millionths)Pin.

This is the method we use on sleeves and custom work with high-tolerance holes that have to be checked to .0002 total tolerance. We've never had a return or complaint on features checked this way. However, this is in STEEL, high-precision work. If he expects a .1251 pin to go in a ".1251" hole, he's just making sure nothing is actually smack at the low.

That's OK too. He just doesn't want to float smack at the fence-line where almost anything could cause a fit issue / binding.

On some of our jobs what comes into play: I'll be checking parts and Go Pins are binding, or having an interference fit. Parts get washed out, cleaned of all debris. Sometimes we find they have shrunk a tad overnight, or a temperature change is now causing a (for instance) .1249 Go pin to bind up in what should be a .1250 hole.

In some cases we'll have those parts re-honed on the Sunnen honing station relaxing the Go Pin selected, and use a (for instance) .1250 Pin in what's supposed to be a .1250 Hole, because we know at that point we are for sure up off the very bottom of the allowable tolerance. No-Go still has to be true of course.

As Carbidebob has suggested, doodling around right at the edges of allowable tolerance is always chancy. At high-toleranced features you can get pooched with nothing more than stress relief or temperature change.
 
My years of experience have shown me that bore gauges, mics and pins can't be mixed when measuring tight tolerances. If you want the customer to buy the parts you have to measure them the same way they measure them. What ever size the customer thinks it is is the correct size.
 
A bit more food for thought:

1. Let's not mix up single line diameter (gauge ball) with envelope on a cylinder (gauge pin). The latter requires the hole to be straight.
2. Air gauges are good, but impacted by surface roughness. If your feature roughness is too much different than your master gauge roughness you will get variation.
3. With perfect form you can get a pin to go into a hole with less than 1 micron clearance, but it's hard to get it started. At some point I had a gauge pin with ~just under 2 microns of barrel error (wider in the middle than the ends, but good roundness). I also had a bearing inner race with an ID in the middle of that size. With good alignment you could get the pin to freely slip into the race. Once on, it was smooth and had the tiniest bit of lash at the end, none a bit later, and had to be barely forced through the middle. About 1/3 of the people I showed it to were unable to get the pin to start, but agreed that it was freely fitting if I started it for them.
4. I tend to think that people aren't nearly as repeatable on dial bore gauges as they think they are. Even those who get consistent results usually miss by a few microns if I ask them to measure a certified ring gauge with the size label covered.
 








 
Back
Top