API610-11, Table 16, note b) reads “cumulative tolerances not applicable”.

[fabienw]

thread407-400572

Hi,

supplemental discussion from a clarification posted by

It is to prevent someone from arguing "I'm 2% high in head, so that should make my power 1.5% higher than rated. Since my power is 5% high from the rated point but only 3.5% high from the +2% head point, it meets the tolerance.

tolerance on head is +3% / -3%
tolerance on power is +4%

how to interpret "cumulative tolerances not applicable" in the following examples?
a/ head +3% and power +6% : fail. Even if one could argue 6% of power "includes" 3% of head, so if head would be 0%, power would become 3%, and hence acceptable.
> in this case, trim the impeller to 0% head which should yield 3% power (give or take), then pass.
b/ head +3% and power +4%: pass.
c/ head -3% and power +4%: pass. Even if one could argue 4% of power "includes" -3% of head, so if head would be 0%, power would become 7%, and hence not acceptable.

 

[shashankTurbomachine]

your question is not clear.

what do you mean by cumulative

A well phrased question has answers in itself!!!

 

[moon161]

Think of a hole pattern, 10 holes evenly spaced @ 2.000 ± .010 . If the tolerance is accumulative, then some wise guy is gonna say the 10th hole is 20 ± .10 from the first because 10x .010 is allowable when you probably don't want that. You need a way to say 'YOU KNOW DAMN WELL WHAT I MEAN'.

Two ways to specify this safely would be '10 HOLES SPACED 2.000 ± .010 TOLERANCES ARE NOT ACCUMULATIVE', another safer more explicit way would be to dimension each hole and specify the tolerance on each dim or with a note/block tolerance but that begins to suck and make for an ugly drawing for big hole patterns.

In OP's question, there are two non accumulative tolerances.

The pump head is to be 'nominal head ± 3%'

The power usage is to be 'nominal power ± 4%'

One tolerance may affect the other but they are taken individually as acceptance criteria.

 

[moon161]

So if tolerances not cumulative:

a/ head +3% and power +6% : fail. Even if one could argue 6% of power "includes" 3% of head, so if head would be 0%, power would become 3%, and hence acceptable.

Yes, this condition fails

> in this case, trim the impeller to 0% head which should yield 3% power (give or take), then pass.
b/ head +3% and power +4%: pass.

Yes, I think do this.

b/ head +3% and power +4%: pass.

Yes, pass.

c/ head -3% and power +4%: pass. Even if one could argue 4% of power "includes" -3% of head, so if head would be 0%, power would become 7%, and hence not acceptable.

Pass. Each tolerance is taken individually.

 

[moon161]

I should ad that I don't know if this is a valid code interpretation, but how I would interpret it.