HI all,

A relatively straightforward question!

I have a characteristic that is 22 and has a tolerance of +0 -0.1. In this case what is the nominal? Is it 22 which is obviously the upper limit of the tolerance band? or is it in fact 21.95, which is in the middle of the tolerance band? :confused:

Not really had to think about that before!!!

Good question, I dont know the definitive answer but I always took 'nominal' as the big number, e.g. in this case 22, and not as the mid point.

Interesting question. I went to ANSE Y14.5 Dimensioning and Tolerancing to look for the answer.

1.3.28: Size, Nominal: The designation used for purpose of general identification 1.3.34: Tolerance, Unilateral: A tolerance in which variation is permitted in one direction from the specified dimension.

In this context it might be more appropriate to used specified dimension instead of nominal dimension (Lumber: Nominal 2x4, actual 1 ¾ x 2 ¾ )

As this is a unilateral tolerance, ie one is 0 and the other allows a variance, the nominal (or desired) figure will be the +0 as that is what you are aiming for.

HI all,

A relatively straightforward question!

I have a characteristic that is 22 and has a tolerance of +0 -0.1. In this case what is the nominal? Is it 22 which is obviously the upper limit of the tolerance band? or is it in fact 21.95, which is in the middle of the tolerance band? :confused:

Not really had to think about that before!!!

In this context, "nominal" refers to the name of something, or how something is commonly identified. 22 is the nominal, because that's how the dimension is referred to.

Manix,

If you don't mind me asking, what is the nature of your question? What context did this come up?

It's a good question; I'm just wondering what the nominal purpose is.

Thanks to all for your input,

Manix,

If you don't mind me asking, what is the nature of your question? What context did this come up?

It's a good question; I'm just wondering what the nominal purpose is.

Brad, my question arose simply because the software I use to generate my charts for Initial Capability Studies and SPC, actually uses the target as the nominal and as a result stated in the charts that the nominal is in fact the upper spec limit. If that makes sense!

I just noticed it because I don't think I have had an instance of that before, most dimensions, the nominal sits nicely in the middle of the upper and lower. It really was just that, it has no bearing on the results or how I interpret them, it is just making sure the labelling is correct, and of course, to help me know what I am talking about when ever I mention "nominal".

Brad, my question arose simply because the software I use to generate my charts for Initial Capability Studies and SPC, actually uses the target as the nominal and as a result stated in the charts that the nominal is in fact the upper spec limit. If that makes sense!

It does make sense. The confusion stems from the fact that we are used to dealing with equilateral tolerances, where the nominal and the target are the same value. With unilateral tolerances, the target will usually lie equidistant from the nominal and the extreme limit. In such cases, the target is not expressed explicitly on the drawing, while the nominal always is.

"Thanks to all for your input,
Brad, my question arose simply because the software I use to generate my charts for Initial Capability Studies and SPC, actually uses the target as the nominal and as a result stated in the charts that the nominal is in fact the upper spec limit. If that makes sense!"

I agree with Jim. My understanding is the value that the tolerance is applied to is the "nominal". The target - if the data for that dimension is normal - would be in the middle of the tolerance zone - whether unilateral or not. The target - if the data for that dimension is not normal - well, that depends...

The fact that the tolerance is stated as unilatereral is essentially a clue to the supplier as to the needs of the product function. It is handy info, but rarely affects processing control. For unilateral form tolerances, and in metric tolerancing, all bets are off. :cool:

I agree with Jim. My understanding is the value that the tolerance is applied to is the "nominal". The target - if the data for that dimension is normal - would be in the middle of the tolerance zone - whether unilateral or not. The target - if the data for that dimension is not normal - well, that depends...


But think about lumber. In the US, you buy nominal "two by four" lumber, but it is really 1.5" x 3.5" (give or take something like 1/16"). So the entire tolerance band is well below the nominal value.

Or pipe. Nominal 3/4" copper pipe is 7/8" OD (and varying ID).

But think about lumber. In the US, you buy nominal "two by four" lumber, but it is really 1.5" x 3.5" (give or take something like 1/16"). So the entire tolerance band is well below the nominal value.

Or pipe. Nominal 3/4" copper pipe is 7/8" OD (and varying ID).

These days that stuff is beyond tolerance, anyway. :cool: That's a whole different misuse of the term "nominal", IMHO. Right up there with "nominal" names of fish in restaurant dishes. :notme:

But think about lumber. In the US, you buy nominal "two by four" lumber, but it is really 1.5" x 3.5" (give or take something like 1/16"). So the entire tolerance band is well below the nominal value.

Or pipe. Nominal 3/4" copper pipe is 7/8" OD (and varying ID).

That's a perfectly valid use of the word "nominal," which in this case means "by name." A 1.5" x 3.5" piece of lumber is 2x4 by name, or nominally. The same usage applies to blueprint dimensions, only in a slightly different sense; the "nominal" in this case applies to the "named" dimension.

The same usage applies to blueprint dimensions, only in a slightly different sense; the "nominal" in this case applies to the "named" dimension.

At least the original question has been answered. :rolleyes:

" For unilateral form tolerances, and in metric tolerancing, all bets are off. :cool:

Metric tolerancing? What is so odd about the metric system? Or have I missed the point:D

Metric tolerancing? What is so odd about the metric system? Nothing much, imo (I have been using it all my life), but I think we tend to use equilateral tolerances more often here in Europe?

/Claes

Metric Tolerancing makes more sense to me than +/- 1/16th!!!!!

Yeah, yeah, it is an age thing here in europe, but was there not a reason for moving away from the Imperial system?:notme:

Being 26 and here in Europe I can't for the life of me work out why anyone wants to use feet and inches, but then I suppose I do drink pints and my speedo on my car still says MPH! :rolleyes:

The metric system rules! :notme:

Many metric prints tend to have whole number nominal, at the cost of centered bilateral tolerancing. So, one should not assume the same impression for unilateral dimensioning as one does on english prints when you come across it.

As far as metric versus decimal english....just pick one. Mixing them on a print is so irritating. I have seen prints call out 8 mm X 8 mm X 30 feet. The only time I have seen fractional english these days is when people are using Quality Level TCE. But, no need to digress - again... :cool:

I use different words to describe these terms. Nominal is the middle of the specification versus set point that is where you set your process. So 22 is the setpoint, but 21.95 is nominal.

I use different words to describe these terms. Nominal is the middle of the specification versus set point that is where you set your process. So 22 is the setpoint, but 21.95 is nominal.

Not sure what you mean by "set point." Why would you want to "set your process" at the upper tolerance limit? The nominal (using the conventional definition) is a "set point" (or should be) only when the tolerance is bilateral and equilateral. The nominal callout should describe the ideal condition, but not necessarily the target. The target should be a condition that takes the variation of the process into consideration. With a unilateral tolerance, the only time the nominal should be the target is when you're absolutely sure that there will be no variation beyond it to the high or low side, depending on which direction the tolerance allows.

Not sure what you mean by "set point." Why would you want to "set your process" at the upper tolerance limit? The nominal (using the conventional definition) is a "set point" (or should be) only when the tolerance is bilateral and equilateral. The nominal callout should describe the ideal condition, but not necessarily the target. The target should be a condition that takes the variation of the process into consideration. With a unilateral tolerance, the only time the nominal should be the target is when you're absolutely sure that there will be no variation beyond it to the high or low side, depending on which direction the tolerance allows.

I assume you have a +0 -0.1 tolerance on a process with tool wear or other factor that causes the dimension to decrease over time. The target or set point is 22 in this case since you want to set the process at the upper specification if the dimension decreases over time. But the nominal value in my world means the mean value for the process. Therefore the nominal value is 21.95 because on average the part will have a value of 21.95.

I assume you have a +0 -0.1 tolerance on a process with tool wear or other factor that causes the dimension to decrease over time. The target or set point is 22 in this case since you want to set the process at the upper specification if the dimension decreases over time. But the nominal value in my world means the mean value for the process. Therefore the nominal value is 21.95 because on average the part will have a value of 21.95.

If this thread proves anything, it's that we need to make sure that our terminology doesn't cause confusion. In the end, it doesn't make much difference if the correct denotation is used if the connotation is different to the person we're communicating with. If someone says "I want you to aim for the nominal" the best response is, "What do you mean by 'nominal'?"