GD&T Position

[[Ji...]]

Hello everyone,

We are frequently encountering measurement‑related questions. Many junior customers keep referring to various international standards such as ISO standards.

 

Our Calypso software has already been configured for ISO‑compliant evaluation.

I would like to clarify: when evaluating the position tolerance of a hole, is Gaussian (least‑squares) fitting the default method according to ISO requirements?

 

Which ISO standard and specific clause defines this rule?

[[Ma...]]

This is ISO compliant.

Datums are tangential, but measured feature is Gauss.

I think only difference about ASME and ISO will be in that feature's evaluation.

[[Ji...]]

Please sign in to view this quote.

Sorry, I would like to know which part of the ISO standard defines this? I have searched for a long time but have not found any clear specification.

[[Ma...]]

Please sign in to view this username.

Hi and sorry, i have no document about this.

But years on forum and work on machines can give a clue how it is.
My main knowledge was with using ZEISS Inspect, where i can easily visually see how is ISO, ASME and GEOMETRY behaving.

Many hints were from

Please sign in to view this username.

 and his drawings and simulations.

[[Jo...]]

You can have a look here for the applicable standards.

https://dgq.de/fachbeitraege/iso-gps-matrix-modell/

 

[[Er...]]

From ISO 1101:2017

Please sign in to view this quote.

Extracted integral feature: the real, tangible, non-perfect feature as probed by the CMM.
Essentially the probed point collection belonging to that feature (a hole, a plane, etc.)

Extracted derived feature: the feature computed from one ore more integral features.
Examples are real (non-perfect, i.e. curved) median lines (centerlines) and median surfaces (like the median plane of a slot).

Optionally, it is possible to specify that the associated (fitted) feature, for instance a Gaussian fitted plane or the perfect centerline of a Gaussian fitted cylinder shall lie inside the tolerance zone. If the Gaussian fitted element (feature) must lie inside the the tolerance zone, rather than the extracted element, this is indicated by adding (G) in the second compartment of the tolerance indicator. See picture below.

In many cases, getting the Gaussian associated plane inside a tolerance zone is way easier than the extracted surface itself.

So, answering your question, no, the Gaussian (least‑squares) fitting is not the default method for the the feature to be measured.
By default, the extracted (real, probed, or calculated from probed points) feature must be measured.

 

 

Edited 4 hours ago