What's new
What's new

CMM vs bore gage accuracy

RJT

Titanium
Joined
Aug 24, 2006
Location
greensboro,northcarolina
We do not have a CMM, Our customer does (Mitutoyo, don't know the model # or stated accuracy) We use Sunnen bore gages and their setting masters to measure bores or motorized height gages(Trimos Vectra with 20 millionths resolution). We have 2 pieces with a 57mm bore -.005mm tolerance. I am very confident we can make and measure that tolerance, we do it quote often (cylindrical grinder or jig grinder). Most of my customers use bore or plug gages, or air gages to measure tight tolerance bores. My question, is a CMM going to be able to measure to that accuracy and therby correlate with my bore gage?
 
We do not have a CMM, Our customer does (Mitutoyo, don't know the model # or stated accuracy) We use Sunnen bore gages and their setting masters to measure bores or motorized height gages(Trimos Vectra with 20 millionths resolution). We have 2 pieces with a 57mm bore -.005mm tolerance. I am very confident we can make and measure that tolerance, we do it quote often (cylindrical grinder or jig grinder). Most of my customers use bore or plug gages, or air gages to measure tight tolerance bores. My question, is a CMM going to be able to measure to that accuracy and therby correlate with my bore gage?

I wouldn't feel comfortable holding +/-.0001 even with my couple year old Zeiss.....

Need to find out if it is a manuall or DCC cmm. If they are only using the CMM i'd expect them to be measuring a certified ring gage every cycle to compare to the part readings before they called me and told me the parts were out of spec...
 
No experience, but I would be astonished if it could not.

Within 5 microns, certainly, imo.
Ie 57.000 - 49.995.

Quantify the measurement to 1 micron .. maybe not.
To 2 microns plus or minus .. I would think so.

I have no idea what to do in a corner case.
Suppose the CMM says its 49.994 mm.

Now what ?
And I am very interested in any experience / anecdotes in such cases.
 
I've got a reasonable amount of experience running and programming CMM's and wouldn't trust one to hold .0001''. If I need to hold that kind of tolerance I buy the ring master and the bore gage. If your customer trusts your gaging they could use their CMM as an 'overcheck' with the understanding that your inspection report is acceptable.
 
Almost all Cmm's are inaccurate as to exact size, that is a general statement, but true. I have worked, as an inspector, in several industries and we (in QA) never relied on the diameter size, just position and form. The only machines I have worked with that gave accurate bore size were older B & S and ADEA that did not use PC-Dimus software. With those you could manually adjust the probe diameter to match a gage ring (usually XX); readings would be good only for that one size. I myself would consider your Sunnen gages adequate.
 
Depends on the model.
In imperial that is a +/-.0002 R&R window if you accept only "0.0000000" parts as good.
This puts you into $100,000 machines
Real world and accepting parts near the limits of the tolerance you need a +/-25 millionths gauge repeatability here.
This puts you into 1/2 million dollar and up CMMs depending on part size.
A cmm for checking auto engine blocks is an easy million not counting the room for it to go in.
Sunnen gages and their like are the preferred method as they will test down in these numbers. They wont give "form and fit" across the entire bore.
80% of the cmms in the field would have a very hard time competing with a new Mitutoyo digital mic.

People blindly trust cmms for some reason. Perhaps because a computer spits out the numbers.
A 3-D gauging system is full of little errors that add up and make for lots of sleepless nights for the designer.

Basic to any measuring tight is knowing through real world test the 6-yuck error band. Most people will shit a brick once they know these numbers.
Then you have to subtract the error band from the print tolerance if you want all good parts.
No person who understands this world uses print tolerances when measuring parts. Process numbers have to be under the print by your gage's guesstimate.
You can never run or machine to the edges of the tolerance. If you do do you will ship bad parts and eventually end up buying them back or not getting paid for your work
Bob
 
I'm also in the "that's awfully thin ice" camp regarding trusting a CMM at that tolerance level. The expectation that a CMM will be perfect at what is essentially its performance limits relies too much on perfection in the hardware, the setup of the machine and its calibration status, and the expertise of the operator/programmer.

Correlation of your gaging masters and the results from the CMM is a process that needs to be done, well and carefully, before you and the customer make any assumptions about the parts. If the CMM is a modern scanning-probe machine, that may be of some advantage. OTOH, if it is a point-probe unit, it will be absolutely necessary to make sure the measurements contain enough points to really show the condition of the part. I think your bore gaging is a substantially better approach to this specific application.
 
People blindly trust cmms for some reason. Perhaps because a computer spits out the numbers.

A 3-D gauging system is full of little errors that add up and make for lots of sleepless nights for the designer.

Bob

Ouch... There is a reason, it's money. I hope this doesn't make you sick - but you'll know in your heart it's true, as I did.

A friend of mine recently retired & we had a chance for a conversation. We had worked & crossed paths a number of times over 40+ years and I count him as one who has "good game".

In the conversation he recalled a time shortly before OMC closed in Galesburg, where he had a bit of a run-in with a CMM. He went with the last of the toolroom because he'd still do the reamer sharpening & such by hand against the finger (those who know this, know it - those who don't - please don't speak) all the wannabes & the feather bedders left the toolroom early.

They had a CMM with little to do, so - tools were sent up to qually... He got called up on a tool# that was out of spec. He gathered a mic & gage stack and went there (it was an even tooth multi-fluted tool).

The person showed him the report & the results from the inspection (looked like a store receipt back in those days). He measures the stack with the mic & measures the cutter & says I'm good by this - see what you get because I don't know what's going on.

The response from the inspector was... "I don't know how to use a mic"...

He said "could you run that again for me?". He did & it wasn't the same but was still out of spec (a bit less). Ron said "I don't know how to run your machine, but I think you're gonna have to learn to run a tool & cutter grinder because you have my best" & walked out - nearly left the craft but came back a few years later in a different place.

About the messaging of the linear math (base 2, 10, 16 etc) vs angle of rotation - they are different bases & often come up with very inconvenient ratios that "will not" reduce for better precision but only "scale up" for "basic" solution. So - when someone says "if you can't find a number between two numbers so there is no number, and they are the same" it's for algebra & number manipulation to get some solution, it is not always a "count to the integer solution".

I'm really tired now,
Matt
 
Ouch... There is a reason, it's money. I hope this doesn't make you sick - but you'll know in your heart it's true, as I did.
.....

Yes, you drop 20-$50,000 on one of these and expect the ability to measure much much better than your 2-$500 mic.
That kind of makes sense. I spent a bunch. We want a bunch from it.
What you bought was the flexibility to do all kinds of stuff not accuracy or repeatability.
A cmm can't run with a 1960's supermic. It just can't, the math does not add up to put you in this range.

We sometimes make "master inserts" for toolholder making people Crazy tight numbers and crazy high price tags.
Then someone will call and say "Your master is no good". We asked "Why,... what did you check it with".
The answer is always my nice new cmm. At this point we know the only answer is "buy a better gauge".
Worse by miles on round tools which are just this side of impossible to accurately check on a cmm.

The nice thing about a very high quality cmm on bores is that you can map the entire bore not just one three point circle and know if it is actually a cylinder.
Hard to get this from a bore gauge
It's a slow process, not suited to production control.
Bob
 
My scanning CMM has taught me a lot about machining processes....
It is a handy tool to have in the tool box. Have to remind a lot of people its a tool and not THE ONLY TOOL.
:smoking:

I still double check super tight tolerances with a hard gage to give myself a better feeling when the CMM gives me the green light.
Running some parts today of .001 TP blind bore to blind bore 180* apart. CMM says first part was within .0003 TP. Scratched my head and go no way and head straight to the surface plate to verify.

Part of the problem is people only question the CMM when you get the "RED" result....
 
We do not have a CMM, Our customer does (Mitutoyo, don't know the model # or stated accuracy) We use Sunnen bore gages and their setting masters to measure bores or motorized height gages(Trimos Vectra with 20 millionths resolution). We have 2 pieces with a 57mm bore -.005mm tolerance. I am very confident we can make and measure that tolerance, we do it quote often (cylindrical grinder or jig grinder). Most of my customers use bore or plug gages, or air gages to measure tight tolerance bores. My question, is a CMM going to be able to measure to that accuracy and therby correlate with my bore gage?

What is the class of ring gage, X, XX ,plus, minus, master...? Are you climate controlled? Are they climate controlled? 2 tenths is usually 'fightin' territory if you are trying to compare numbers from two shops and two different measuring techniques IMO.
 
good question. The cmm guy will tell you it is always perfect and cmm makes no mistakes. In real life things happen that cant be repeatability check with the cmm do to what ever. The cmm is right from point a to point and it is x amount, but how that works with the part as a whole and functionality of it is an whole different ball game. Good old school hand buy out vs cmm buyout, fixture points clamp points vs points picked up in cmm. Part restrained vs not restrained. Gets really complex at times.
We generally notice over the years the cmm have problems on smaller the holes, threaded holes, twisted and bow, all adds to the more possible differences from the cmm vs hand check. We have no luck checking dia smaller then 1/4" with repeatability with are cmm. we go by no/go pins and bore gauges as final say.
 
We do not have a CMM, Our customer does (Mitutoyo, don't know the model # or stated accuracy) We use Sunnen bore gages and their setting masters to measure bores or motorized height gages(Trimos Vectra with 20 millionths resolution). We have 2 pieces with a 57mm bore -.005mm tolerance. I am very confident we can make and measure that tolerance, we do it quote often (cylindrical grinder or jig grinder). Most of my customers use bore or plug gages, or air gages to measure tight tolerance bores. My question, is a CMM going to be able to measure to that accuracy and therby correlate with my bore gage?
.
what many old timers do is keep ring gage to set bore gage in machine getting coolant sprayed and at same temperature and orientation as part. as a part vertical and a part horizontal can measure different. then bore gage it used and different spots looking for tapered bore and out of roundness.
.
i have seen a dry unclamped part measure different that a wet part in the machine when using a plug gage. sometimes as simple as when vise released a round bore is no longer round. wet and dry not sure why different but plug gages can give different feel when wet and when dry
 








 
Back
Top