MMC on Composite Tolerances

[[Ry...]]

Hello,

When using composite tolerancing we are seeing a large spike in deviation when applying MMC.
My guess is that it's changing the computation of the cylinder, based on the data we've captured.

Can someone explain exactly what the software is changing in the background that could cause this deviation to change so drastically?

These are also probed features, which lends me to believe it has to do with how we've probed the hole.

Any information on this would be greatly appreciated.

Thank you.

[[Ti...]]

Please sign in to view this username.

In my opinion they need to improve the composite tolerance capabilities. More importantly, they need to create an online video to specifically explain how to do this. Composite tolerancing is not well understood and it requires a video explanation. They like to point you to the help section, but this is a different animal. I believe somehow you have to create an 'element group' and a few other steps.

I don't see a need for composite tolerancing in your example, b/c the lower FCF(feature control frame) is only using 1 datum which is only perpendicularity. When the lower segment in a composite has 2 datums, it controls rotation, and that is when you need composite. Also, both your FCF's above are not the same, one has MMC for the feature and the other does not.

 

There is also a bug in the software in which I uploaded a video to the GOM team...if I remember, if you make a change to a positional tolerance using MMC it will not impact the calculation, UNLESS, you then edit the feature, and SIMPLE close the feature, do nothing...just edit it and close it. I documented this in a video explanation to the GOM team but have not heard back and this was 50 days ago.

Have a read here on composites:

https://forum.gom.com/topic/592-defining-composite-tolerance/

 

Please sign in to view this username.

maybe you can help this user

 

The good thing is composite is only needed to be ran as a composite if the lower segment, what is called the FRTZF(Feature-Relating Tolerance Zone Framework) has 2 datums in the FCF(feature control frame). With only one feature control frame present in your case, there is no need for controlling rotation, only perpendicularity. So you can just run it as a multiple single segment. I believer and hopefully

Please sign in to view this username.

can confirm even is you want to do a 'multiple single segment' for multiple features you must use the create 'element group' still.

 

It is for these multiple layers of issues it would be incredibly helpful if a GOM professional, that also understands ISO and ASME Y14.5 extremely well could make a video explaining these topics well. This is not an easy ask, but I think GOM should produce such a video.

 

 

 

Edited March 1, 2023

[[Ti...]]

Here is an excellent video on the difference b/w composite and multiple single segment and when they are the same thing:

https://www.youtube.com/watch?v=4OVvshCacxg

I'm interested in how you make our with your problem so if you can please keep us posted here. I'm just an end user that pushes the software.

[[Ch...]]

Hello Ryan, Hello Mr Jones,

 

according to your labels I would assume that you compare Material Requirement on surface with the middle axis. Of course the result is completely different. To be sure enabled the deviation plot. Otherwise please contact your support partner to get further advise. 

If datum A is a plane that is perpendicular to your Cylinder then the same behaviour should be appear without using composite tolerancing. For this case (and only for this one) Mr Jones is correct that the composite tolerances is the same a perpendicularity tolerance. Of course there are existing datums that restrict positional degrees of freedom so that it makes sense to have FRTZF with only one datum (e.g. if the datum plane would be parallel to cylinder).

And yes for all GD&T tolerances that have to apply to more than one feature you have to create an element group first.

Christoph Schult

 

[[Ry...]]

Please sign in to view this username.

 Thank you for your thorough response. In summary, I agree with everything you've established. In this case I did create an element group of the two features. I only do this for composite and not multiple single segment. I've argued for a simple perpendicularity check rather than composite tolerancing for that very same reason. Design won't budge. Since the standard allows it, they're sticking with it. (A lot of our drawings have this issue.) Also, with the top portion not having MMC and the bottom having it, that's also Design. That's how they drew it up and they'd like to keep it that way.

In this case I was able to rationalize that since it passed without MMC, the MMC modifier wasn't necessary as the bonus isn't needed. However, I fear I'm only delaying the inevitable for when it doesn't pass without MMC. We've also seen significant differences between probing 8, 9, 10, 11, or 12 points for the cylinder. It varies so drastically, that it's tough to establish a "best practice" method. That's why I'm trying to understand what the software is doing in the background.

Thank you for the resources! I'll definitely be checking these out.

Please sign in to view this username.

 Can you please elaborate on your first statement? Are you talking about scan data vs probe data? We only probe these features as the holes are too small to be scanned. 

Ryan

[[Ch...]]

It has nothing to do with data acquisition. Per default MMR evaluation is defined on the hull of toleranced feature. Please have an deeper look into the corresponding standard. ISO doesn't know a concept as axis interpretation if the tolerance under MR. And even if the ASME standards says that the surface interpretation should take precedence over the axis interpretation (ASME Y14.5-2018 par. 5.9.2). 

If you using a pattern we are providing only the surface interpretation because otherwise each element has an different tolerance zone (given tolerance value + element vise bonus tolerance). Despite from computation problems this would be results in hardly understandable result. Therefore for MR only surface interpretation is offered

For "single element tolerance" you can choose your interpretation manually.

[[Ti...]]

Just to clarify the lower segment is only referencing the bottom surface Datum so you simply have a perpendicularity tolerance, BUT, simply using a perpendicularity on each individual feature allows each feature to float anywhere inside the location tolerance set by the first composite line(PLZF). If you apply perpendicularity on an AMSE print they are separate checks. I'm not sure simulanteous callout even fixes that fact.

The second callout(FRTZF), while simply saying you must tighten up your perpendicularity, further tightens up the distance each feature can be from one another.

I have never done a composite in GOM, so maybe

Please sign in to view this username.

could explain, but if you simple use a perpendicularity tolerance on the 'ELEMENT GROUP' does that require each element to remain withing the toleranced distance from each other as well as the tolerance .025 to datum A for perpedicularity. I've always wondered this.

My guess is if you grab the group as a whole and use a .025 to Datum A it should require both all feaures remain in a circular tol zone to the CAD model(basic dimensions) as well as .025 perpendicularity to Datum A.

Edited March 18, 2023