Tool Capability vs. Customers Requirements and Specifications (Ethics)

[Abelethan]

All,

Myself and my colleges have been having an on going debate as what are legal/ethical requirements are regarding following a customers spec. The tool we use to measure this customers spec reads out to 3 decimal places. The customer has only requested one. Case in Point; Customers spec is 160.0±0.2 half of the group says 160.201 is out of spec while others say 160.249 is in spec. What are your thoughts on this manner?

Thanks

 

[pkost]

Re: Tool Capability vs. Customers Requirements (ethics)

My thoughts are that the specification is not clear enough and that clarification should be sought.

From the information you have provided the part is out of spec. +/-0.2 is an absolute...even +0.20000000001 etc is out of spec.

The lack of clarity in the specification is how accurately the manufacturer is required by the customer to measure the spec. The manufacturer chose to measure it to 3d.p and made a rod for it's own back.

 

[somashekar]

Re: Tool Capability vs. Customers Requirements (ethics)

All,

Myself and my colleges have been having an on going debate as what are legal/ethical requirements are regarding following a customers spec. The tool we use to measure this customers spec reads out to 3 decimal places. The customer has only requested one. Case in Point; Customers spec is 160.0±0.2 half of the group says 160.201 is out of spec while others say 160.249 is in spec. What are your thoughts on this manner?

Thanks

Welcome to the cove Abelethan ~~~
A simple mathamatical reasoning says that 160.201 or anything higher is out of the specification. The customer spec is based on his requirement which is 0.4 in total across the nominal of 160, ie 159.8 (159.80000000000000000) to 160.2 (160.2000000000000)
You may not know what happens to his assembly if it is out of this. Hence if it is outside of your 3 decimal reading more than 160.200 kindly check with the customer. Perhaps he may give you a waiver. BUT meet specification OR ask.

 

[Wes Bucey]

All,

Myself and my colleges have been having an on going debate as what are legal/ethical requirements are regarding following a customers spec. The tool we use to measure this customers spec reads out to 3 decimal places. The customer has only requested one. Case in Point; Customers spec is 160.0±0.2 half of the group says 160.201 is out of spec while others say 160.249 is in spec. What are your thoughts on this manner?

Thanks

I would have considered this to be a "no brainer," but that just goes to show that assumptions have no place in issues like this. We have a long standing aphorism in business and in life in general:
Break "assume" into its principal components and you have
an ASS with U and ME.

The fastest solution to this is simply to ask the folks whose decision determines whether you get paid for parts or for scrap, NOT a conglomeration of experts and amateurs (like us in the Cove), because we aren't going to make the decision whether you get paid or not because the parts don't meet the specification.

Incidentally, I also believe ANY hairline past the stated tolerance in either direction is nonconforming. Think of those tolerances as having a million zeros after the last significant digit in the decimal, rounding is NOT the rule.


The number of decimal places your instrument reads is not quite as important as its accuracy and repeatability time after time. Typically, we want an instrument that reads at least one decimal place past the number of decimal places in the specification or tolerance. But don't assume, ask the guy whose decision counts!

 

[Jim Wynne]

All,

Myself and my colleges have been having an on going debate as what are legal/ethical requirements are regarding following a customers spec. The tool we use to measure this customers spec reads out to 3 decimal places. The customer has only requested one. Case in Point; Customers spec is 160.0±0.2 half of the group says 160.201 is out of spec while others say 160.249 is in spec. What are your thoughts on this manner?

Thanks

As others have pointed out, the answer is pretty simpler. Is 160.201 greater than 160.2? Therein lies the answer. This shouldn't be a matter of contention, but the way to avoid it is staying away from the extreme limits of the tolerance.

 

[bobdoering]

Also, consider this: To be capable you should be within approximately 75% of the tolerance - due to sampling error and the other factors of the total variance equation. You can only be at the specification if you are measuring 100% and assume no gage variation (that would be a tough one to prove). Production does NOT own 100% of the tolerance - only the customer's dock does.

 

[Abelethan]

Thanks all for the replies. Its as I figured. Its the typical battle between quality and manufacturing. Thanks for your input.