Issues with dimensioning GD&T

[[email protected]]

We cant seem to understand this...
When we report out gd&t from pc dmis it comes out with major differences when say for instance
you check concentricity from an ID bore to the OD it may read .001
But if you reverse it, say the OD to the ID it will say .006!
idk but I'm hoping its just something silly on our part.
What's the deal there?
Its aligned to a model and the degrees of freedom are well constrained

[Matthew D. Hoedeman]

length of 'datum' and length of feature.

A longer datum and shorter feature gives a smaller value, a shorter datum and longer feature gives a bigger value.

[[email protected]]

I'm not understanding this?
what do you mean when you say longer or shorter datums and features?
I've never heard that terminology.

[Matthew D. Hoedeman]

wow.

[[email protected]]

Please explain this
say you have 2 measured circles on the same plane.
just say its a gauge ring.
align it properly and then create a datum for the inside and then one for the outside.
dimension concentricity from the od to the id
and then the id to the od
and you get 2 different results of the same measurement?
A person with a simple pair of calipers could mic around the id to od of the part
and then flip the calipers over and get the same
because it WOULD be the same
so why is pc dmis saying its not?
That's what "wows" me.

[UKCMM]

Are you measuring one feature as a circle and the other as a cylinder as this can cause the type of error you are seeing.

[[email protected]]

circle to circle
ID to OD

[AndersI]

Projection distances maybe - is the ID and the OD on the same level? Are they parallel? Compare to this picture of distance between two lines - big difference, depending on which line comes first, and which one is treated as the datum:
dist_li_li.PNG

[[email protected]]

Makes total sense!
Thank you for explaining it.

[vpt.se]

Are they measured in the same plane? If not, it could be the unparallelism between the planes that causes this difference.

EDIT: Just like AndersI's crude sketch above - just replace the lines with planes.

[KIRBSTER269]

Don't take this offensive, but have you ever taken a GD&T class, if not you should, (and I don't mean "Hexagon went over it, when I went to their course") here's the problem with the courses. They Jam 3 years of knowledge, into a one week class, they are throwing GD&T in there also. Ask all the questions on here you need to. You will always receive an answer, but a GD&T class in my opinion, nope let me rephrase that. GD&T should be taken before even touching PC-DMIS. We do know and understand that there are situations of companies, that don't want to spend the money or so on and so forth. Don't wait for them, invest in yourself if you have to. That's what I did. I paid for 2 of my 3 GD&T classes, all of my PC-DMIS classes, and no one can hold that over my head, except me, That's why Matthew D. Hoedeman response was WoW. Sometimes the clear picture is the one that can be seen, A good GD&T class will do that for you. Then PC-DMIS will become a less complicated software to run. Well besides all the crashes and problems, it will be easier.

[[email protected]]

@kirbster269
Very heartfelt! Thank you for that.
What you said about the wow really alerted me to my own shortness of knowledge in this field.
Sometimes we get really good at what we do and think we really know something,
but its conversations like this that allow for an opportunity of growth.
I will be attending the gd&t course from hexagon next week in Texas.
Then back again in November for the 303 class.
I hope to level up a bit from those 2 classes.
But most importantly I will like you said explore a true gd&t course to really educate myself on what's actually going on.

[KIRBSTER269]

Just to add, I have been doing this for quite sometime, but even I have been clearly wrong on subjects, even on discussions on this forum, kind of embarrassing when one misquotes or misleads someone seeking answers, I could delete those remarks, but I don't, because someone else might look at things from that perspective, and when it's wrong it's wrong. I believe that this way people who question things, we as a group cover the right answers as much as the wrong ones. I keep my ASME book close to me, because even in my mind if I know I'm correct, I still open up the book to be sure.

[Mike Ruff]

Check the circularity of both circles. I bet you will find that one of them isn't very round and that is what's causing this. Concentricity is based of the centerpoints of opposing sides of a feature. If your form is out, it can affect the concentricity reading.

[Mike Ruff]

Also, is there a reason you are using circles instead of cylinders for this dimension?