How to Measure a Diameter - Average or Maximum/Minimum?

D

dna_leri

I am sure this is a basic question but I do not have a reference standard for measurement to give me the answer.
We have a number of plastic moulded parts with a diameter dimension on the drawing. As part of the validation, our design group measures the diameters using a CMM which gives an average diameter measurement. If this meets the drawing spec, then they approve the part. However is we take a min and max diameter, one or both of these may not meet the specification - obviously because the part is not round. There is no roundness specification on the drawing or the location to measure the diameter is not specified.
Is there a standard way of measuring diameter (average or max/min) specified in a recognised measurement standard? My instinct is that average diameter is wrong, but I need some back-up.
 
Last edited by a moderator:

SteelMaiden

Super Moderator
Trusted Information Resource
Re: How to Measure a Diameter - average or max/min?

I am not sure what the standards (specifications) are for your product. I know that for steel bar there are diameter tolerances (based in part on sizes) for finished steel it is nominal minus some dimension (one, two, three thousandths, for example). So if the out of round condition was bad enough to go outside the spec limits, that bar was out of spec.

As far as the actual measurement to record, again I would suspect that it would depend on product/industry standard. We did SPC so we had a fair number of measurements, but basically in a bundle, or comparably a lot, we just took a measurement and recorded it. We knew that everything was supposedly within tolerance according to sampling and SPC. We also checked out-of-round condition and used SPC on that.

Does this help you? I am sure that are lots of differences between our two industries.
 
B

Britman - 2012

Re: How to Measure a Diameter - average or max/min?

There are a number of ways to control the parameters of a diameter, however the first consideration is application, the current method could be acceptable for it’s’ intended end function – what is the function of this controlled diameter, could help in answering the question.

Review its intended use and tolerances / measurement methods / records as applicable; do not over spec. if not required, will increase costs.

A good reference to Geometrical tolerances –

www.roymech.co.uk/Useful_Tables/Drawing/draw_geom.html
 

Ron Rompen

Trusted Information Resource
Re: How to Measure a Diameter - average or max/min?

The commonest method for evaluating a diameter is also the simplest; use a GO/NOGO gauge. Any method which takes a discreet number of points (such as a CMM or optical comparator, or a bore micrometer) is subject to error based on several factors (method of calculation, number of points taken, operator skill, etc).

The 'best fit pin' method is frequently the most applicable, since it matches what your customer intends for the function of the feature.

If all else fails, you can always ask them to more clearly define what it is they THEY feel is important.
 
D

Dean Frederickson

Re: How to Measure a Diameter - average or max/min?

When using a CMM I always used "average" as long as it was in tolerance (+/- .003 or higher) it was good. Now if your tolerance is so tight (+/- .002 or less), that roundness may play a part in fit and function, then I would probably use a gage pin to check the diameter size. This is not to any standard it's just what I always went by.:2cents:
 

Stijloor

Leader
Super Moderator
Re: How to Measure a Diameter - average or max/min?

There are a number of ways to control the parameters of a diameter, however the first consideration is application, the current method could be acceptable for it’s’ intended end function – what is the function of this controlled diameter, could help in answering the question.

Review its intended use and tolerances / measurement methods / records as applicable; do not over spec. if not required, will increase costs.

A good reference to Geometrical tolerances –

www.roymech.co.uk/Useful_Tables/Drawing/draw_geom.html

Hello Britman,

Thank you for the information.

Allow me to point out that there are still (significant) differences between the ISO (world) and ASME (USA) GD&T standards.

Stijloor.
 
B

Britman - 2012

Re: How to Measure a Diameter - average or max/min?

Hello Britman,

Thank you for the information.

Allow me to point out that there are still (significant) differences between the ISO (world) and ASME (USA) GD&T standards.

Stijloor.

Good point Stijloor.

Cheers
 
M

M Greenaway

Re: How to Measure a Diameter - average or max/min?

Yeah I think the basic answer is that your measurement method must suit your desired application. If your CMM gives you a result but the parts still dont fit properly then you have to change your method. Although roundness might not be specified it is implied by the max and min material conditions.

Micrometers only measure across two points, CMM's can measure as many points as you want, but as others have said only a ring gauge can truly ensure that the diameter is acceptable as it looks at the whole diameter and in effect mirrors the application - assuming that you are talking of a diameter that fits into a hole, if it flaps in fresh air I wouldnt worry too much and just use the CMM.
 
T

True Position

Re: How to Measure a Diameter - average or max/min?

In my experience on a CMM using the least square best fit for a diameter will give you more accurate results to the part functionality compared to the maximum inscribed or minimum circumscribed calculations. Both of those other fits, while if accurate, will function as the physical pin/ring, they tend to fluxuate and have much much lower repeatability due to say a small error in finish, dust, debris, etc causing the entire diameter and position to move. Something avoided with the LSQ.

If you're really concerned about out of roundness, can your CMM take a lot(1000+) of points quickly and show you the roundness and generate a form plot?

If you're using a Zeiss this is very quick and easy, not so sure about PC-DMIS.
 
G

Geoff Withnell

Re: How to Measure a Diameter - average or max/min?

In my experience on a CMM using the least square best fit for a diameter will give you more accurate results to the part functionality compared to the maximum inscribed or minimum circumscribed calculations. Both of those other fits, while if accurate, will function as the physical pin/ring, they tend to fluxuate and have much much lower repeatability due to say a small error in finish, dust, debris, etc causing the entire diameter and position to move. Something avoided with the LSQ.

If you're really concerned about out of roundness, can your CMM take a lot(1000+) of points quickly and show you the roundness and generate a form plot?

If you're using a Zeiss this is very quick and easy, not so sure about PC-DMIS.

I wonder how one can talk of "part functionality" without knowing the design and use of the assembly? In any case, dimensioning and tolerancing on most drawings are defined by ANSI Y14.5. Under Y14.5, any deviation of the part beyond the concentric circles of max and min diameter is out of tolerance. There is a implicit roundness spec of the difference between max and min diameter.

Geoff Withnell
 
Top Bottom