# Tolrance interpretation

Six Sigma – iSixSigma Forums Old Forums General Tolrance interpretation

Viewing 10 posts - 1 through 10 (of 10 total)
• Author
Posts
• #31497

Never Mind
Participant

This can seem a stupid question, but we are having problems with it.
We’ve got a part with a hole that is specified Ø10±0.1. How would you define the group of parts that meet that specification?
The easy answer would be “a part whose hole is a cylinder with a diameter not smaller that 9.9 and not larger than 10.1”. The problem with that is that the surface of the hole is not a perfect cylinder.

0
#83244

Never Mind
Participant

0
#83246

A.B
Participant

What do you mean by “perfect” cylinder ? What is your objective ? If you can answer both questions then you can set ” preliminary ” tolerance specs.
It seems that you are not satisfied with the tolerance interval +/- 0.1 , that means you still have not defined your final specs.
Right ?
A.B

0
#83250

Never Mind
Participant

The spec ±0.1 is set by our customer, so it is the final spec. By a not “perfect” cylinder I just meant that the shape of the hole can not be described just with one diameter, to check if it is in tolerance or not. Let me explain better what I mean.
The product is an assembly from which some of the parts are buyed to subcontractors, including the part with the hole. It is a plastic injectied part. Becasue of the irregular shape of the part and the injection conditions, the shape of the hole, which is almost perfect in the mould, gets deformed (somehow ovalized and conic). According to the Engineering people, that geometrical errors are allowed as long as they fall within the tolerance for the diameter. The supplier says that the part is Ok because they check it with a min-max plug gage, and when you check it that way it looks ok. The fact that the min gage Ø9.9 is fully accepted means that no part of the hole goes below the LSL. The fact that the max gage Ø10.1 is not accepted means that at least some parts of the hole do not go beyond the USL, but not that no part of the hole goes beyond it. In fact, if you measure the hole with a caliper you will find some ” diameters” beyond the USL. But the suppliers still says it is ok because not only the go-nogo test is passed, but also the “average diameter” is between 9.9 and 10.1. That may be true, but my view is that is not enough to say that the part meets the specification.
For me a good part will be “a part where the whole hole can be contained between two imaginary cylinders of diameters 9.9 and 10.1”, and the part does not meet that, but the supplier says it is a wrong interpretation of the specification. My point is that an oval shaped hole with minor diameter 9.9 and major diameter 10.3 would meet the supplier’s criteria, but for me it would be clearly out of tolerance (the major diameter is 0.2 above 10.1).
Please do not give pseudo-answers such as “Have you checked the measurement variation?” or “Why not to perform a DOE on the injection to fix the problem in first place?”. I just want to know if my interpretation of the specification is Ok, if the supplier’s interpretation is Ok, or if they are both wrong and there is another one correct.
Any help will be appreciated.

0
#83256

topher
Member

From your explanation, you’re interpretation is correct.
By only checking with a go-nogo plug gage, the supplier cannot make the statement that the “average diameter” is in spec. The only thing the go-nogo plug can verify is that the minimum diameter of the oval (or other configuration) hole diameter is between 9.9 and 10.1. Depending on the degree of deformation, the average hole diameter could easily be larger than the USL and still pass the go-nogo plug inspection.
There are a number of measurement methods and techniques available to provide a more accurate representation of the true minimum or maximum hole size. I have used various ways to address similar situations in the past, such as: air plug gages, vision systems with software to calculate min/max/ave diameters, placing the part in a plain ring gage to bring it back into “round”, etc.
Without knowing the specifics of the part in question, I can’t recommend what type of measuring instrument and possible fixturing would be most appropriate for your application.
Good luck.

0
#83263

Never Mind
Participant

Thanks Topher.
Now, suppose that the average diameter is actually in tolerance, is that enough to say that the part is in tolerance? Example:
Diameter spec: Ø10±0.1
Go-nogo plug: Ok (the minimum diameter of the oval is between 9.9 and 10.1)
Minimum diameter of the oval (measured somehow): 9.95 (Ok) (this explains that the go-nogo was Ok).
Maximum diameter of the oval (measured somehow): 10.15 (not Ok)
Average diameter: (9.95+10.15)/2=10.05 (Ok).
Is the part conforming the spec stated at the beginning? How would you define the universe of parts that meet that spec?

0
#83274

A.B
Participant

My opinion is the ambiguity lies in the fact that you ( or your customer ) have not defined the geometry of your tolerance : in the case of a hole you can face a circular tolerance as well as a rectangular or a square tolerance. Obviously the circular tolerance is the best for you ( keeping symmetrical intervals around the centre of the hole ).
As long as your customer or supplier do not agree on the type of tolerance , then you’ll have always some diameters out of the +/- specs ( which is a fact ).
A.B

0
#83284

Never Mind
Participant

However, I am just trying to understand how the specification should be interpreted. I imagine that there must be some standarization on that. Forget about this particular case, my customer, my supplier, and my part. The question is very stight:
What is Ø10±0.1? What falls inside this specification and what falls outside? (given that a part can NEVER be a perfec cylinder, regardless of how perfect may it look in the drawing)

0
#83291

Participant

Dear Gentlemen,
I do not know how much this will help but I would like to share this with you…
As per ISO 1101 – 1983 article 1.1; 3.2;3.3; 3.4 and 3.6 it says the tolerance feature can be of any form or orieantation within this tolerance zone, unless a more restricitive indication is given. This tolerance applies to the whole lenght or surface of the considered feature.
Depending upon the application, we can decide the measuring tools.
I take this oppurtunity to get some feedback also, it is open for comments and suggestions..
Surprisingly, many shop floor engineers say that the tolerance should be with in 1/3 rd the tolerance. which I do not agree. If there is no indication, there is no need to produce “over quality”.as it costs money!
Is there any other standard which indicates limits of 1/3 of the tolerance?
regards,

0
#83295

Never Mind
Participant

So many people from manufacturing, quality, shop floor, engineering, etc… and nobody knows what Ø10±0.1 means? (me included) How will all we do the Cpk, defects count, MSA, improving process, etc if we can’t tell right from wrong?
Let’s make it like a survey. Please answer how do you use it, even if you are not sure that it is the right way. Maybe we can find a “best practice”