Allowances and Tolerances

It's already non-conforming! It does not conform to the size limits set by the blueprint. To deviate from that standard would require written approval from the responsible engineer. Trend analysis should have been in place and practiced by the people running the machines. If this product was trending out at the end of the last run and nothing was done about it, then shame on you. Quality Control is the cop on the beat. Engineering is the judge, be it in house or customer.

I would round it up to .455 regardless and reject the part.

When in doubt reject it and let someone of higher authority override you.

Unless the part isn't that critical. Sometimes people get a little too picky when it's unnecessary, but don't make that assumption if you're not familiar with what the parts are used for yet.

Just reject it. Better safe then sorry.

Everyone on here is right in what they've said to you because they are basing their answers on what is acceptable where they are working at.

It's unfortunate if such a relative answer would be the overall conclusion of this thread because standards are absolutes and truth isn't subject to relativity.

That being said, of this thread some are right and others are wrong in regards to the truth.

My feeling is that majority of people who have posted here are of the opinion that the part was out of spec.....

Am I alone?

Not at all. I agree that the part is out of spec, I was one of the only ones seemingly on the other side of the fence, but my intention was just to show the argument, based on the standards currently set in place.

The bottom line, if that part was moving through my shop it would be rejected, without a doubt. But my shop does not comply with ANSI standards (unless requested by the customer), and we still actually have a few good ol' boys who just know how to work, and refuse to adopt any of these new (lawyer safe) standards.

The double bottom line is: If you do consistently good work, you will not need to worry about getting sued (in general), and can use common sense, logic, and experience. If you do shoddy work, you better comply with a set of recognised standards to keep your tail out of the fire.

I sent the OP a PM stating exactly that a day or two ago...

#1 Out of tolerance!

#2 Process indicator, part is on the high side.

#3 Try to get as close to 0.452 as possible!

At least it can be reworked and not scrap!

±0.002 means: -0.002000000... up to +0.00200000000.....

This is what we call precision...

So, you're right, 0.4548 is out of tolerance.....

Assuming it's an OD ...
Put it in the fridge for half an hour and re-measure...

So have we come to a consensus here? I like to know.

I feel it has to pass a go/no go gage.

I agree with go / nogo here are the tolerances for them. No where close to 0.0008

Class...................... Go........ Nogo

If I understand correctly, a class Z GO/NOGO gauge for the OP's specs would then have (assuming the values are ODs): ID_min= 0.4499", and ID_max=0.4541". Is that correct?

In any case the part is a reject!

Consensus is: In the real world, the part is a reject.

^{I'm very sorry for causing so much ANSI nonsense in this thread...}

Drop it in liquid nitrogen and try the go/no go gauge again... (put the gauge in the oven first)
Del (I'll pass that damn thing even if I have to swallow it first)

With all the controversy.

Maybe you should clear up what type of quality control you are conducting.

Is it product sampling or are you inspecting a manufactured part?

If you are sampling and it is consistent then you are seeing a deviation that needs to be corrected.

If you are inspecting a manufactured part then you have to check to see if the error is going to hurt anything because those standards are set by company policy and not ANSI. Engineers making drawings at the company do a lot of copy pasta and don't bother revising tolerances as needed per part. You might get tolerances specified for a welded on piece being applied to a shaft that is used with a pillow block bearing.

If lets say the part you are inspecting is a shaft that is measuring 0.0028" larger then the called out size to fit in a hole that was machined that also has a tolerance of +/- 0.002 and it measures -0.002" then there is a .0.0048" difference and the part isn't going to fit.

So the next step is to see if it fits where it needs to go.

Or maybe you should just get Js tables and use those tolerances.

"Engineers making drawings at the company do a lot of copy pasta and don't bother revising tolerances as needed per part."

There is a BIG difference between 'don't bother' and 'occasionally miss'. I do a lot of copy and 'pasta' in my drawings, and I do occasionally get sidetracked and forget to change a precision value. I have no problem when I am corrected by the foreman or the machinist, or anyone else who discovers the error.

But if it is true that they 'don't bother', then those engineers should be seriously reprimanded or fired!

But if it is true that they 'don't bother', then those engineers should be seriously reprimanded or fired!

Seriously fired then rehired with loss of seniority and perks...ouch!

That just doesn't happen around here.

The frame of mind is to get it out to the shop. We've had a big issue about our engineering department being a major bottle neck in getting work orders out causing us to fall behind on contract. So they generate the drawings as quickly as possible and as a result we have a lot of rework.

It's not as bad as it used to be but we still have the copy pasta bandits hard at work and very little is done for accountabilty.

Guys,

when I ran an engineering shop, and reported in at Business, Sales, Systems and Quality levels we always took the view that if a part was out of spec ( technically) the overal picture was reviewed and a concession would be agreed with all concerned when the overall picture would be understood to be acceptable at the next level and above upwards.

The future of the tolerance would be left to Engineering and Quality to review for that part. The review would be ongoing until a resolution was reached, a concession was on a one time only basis, it could not be used in perpetuity.

Sleepy

You really need to wake-up, NCR's were covered 2 years ago

Well if you used a three digit electronic vernier it would have measured .455" . To do any accurate measurements you must use the next digit to know where you are. the .4548 is out of spec....... and .4544 would be in spec

.4544 would be in spec

Only if the tolerance were called at .4524 ±0.002

The OP said "It was .452 ± .002"

If you were to make a go/nogo guage for that spec, and assuming this is a spec for an OD, how large would your maximum hole be? The gauge MUST be specified to at least one more digit of precision than the parts it is to accept/reject.

I say that hole would be Ø 0.4540.