3D Math Data, GD&T and FAI - My issue lies more with Profiles

[Celtic Warrior]

Hi
Just thought I would let you know that you're not on your own with this type of issue!!
I have been in a similar discussion with a customer of mine for 4 years now over the determination of position by GD&T specifically profile of a surface. We tried various different ways to determine the 'true' position by measurement none of which were as the designer intended. In my case however despite several long and difficult discussions with the designer, he was unable to give any input as to how to resolve this point only that all proposals were not what was needed.
We have just won a new order with the same customer so I am preparing myself for the next round, as we have the same designer with the same issues and as yet no agreement.
On the positive side, we also have no complaints from the customer either, guess it wasn't as important as first thought?
CW

 

[SArun Kumar]

Being on the receiving end of drawings with comments like "all other features - refer to math model" I would recommend caution in minimalizing a drawing. While the designers and engineers at your suppliers will have CAD workstations to be able to interrogate the part design, the people doing the inspection of the resulting parts for qualification and ongoing production will almost certainly not.

My request for any customer on their drawing would be this: tell me on your drawing what is important, do it clearly, and accurately. If you have a position or profile specified on the drawing, I want to see a datum structure that restrains degrees of freedom correctly, is meaningful, and accurate to the design intent, and I want to see all required basics to be able to measure them.

By leaving required information off the drawing, you open it up to interpretation. If you have simply a profile callout, I will interpret it as requirement of flatness, straightness, form, etc., and not necessarily measurement about a true profile contstrained by a basic. If you have a position callout with no basics, I will come back to you and request that the drawing be corrected.

Just from one on the receiving end.

If I have (Profile 0.8 against Datums A /B /C).it means two parallel profile surface with displacement of +/-0.4 from mean is our tolerance zone.
Now I have max displacement observation -0.718(-0.718-0.40 = - 0.318 deviation) but when I am going making report against this (Profile 0.8 against Datums A /B /C)= -0.718 it seems ok. How could I make this report correctly?

 

[Stijloor]

If I have (Profile 0.8 against Datums A /B /C).it means two parallel profile surface with displacement of +/-0.4 from mean is our tolerance zone.
Now I have max displacement observation -0.718(-0.718-0.40 = - 0.318 deviation) but when I am going making report against this (Profile 0.8 against Datums A /B /C)= -0.718 it seems ok. How could I make this report correctly?

Did you interpret the Profile callout correctly?

You must measure and report the results relative to the true profile that is specified using basic dimensions. The default profile callout is that the tolerance boundary (zone) is equally disposed about true profile.
So you must measure deviations relative to the true profile specification.

Do you have a sketch that you can scan and attach?

Stijloor.

 

[justncredible]

What you want is called the "T-Value".

Say you have a given nominal surface point value. You want the known deviation from that nominal along its approch vector.

JMGREEN MADE THE ATTACHED DOCUMENT FOR PC-DMIS USERS.

With a cad model in pc-dmis you can report out the T-Values, it is great for complex surfaces. You produce that value from a surface plate layout. It in simple terms the devarition from nominal square to the axis. Well maybe not to simple, but that is how I understand it.

 

[Stijloor]

What you want is called the "T-Value".

Say you have a given nominal surface point value. You want the known deviation from that nominal along its approch vector.

JMGREEN MADE THE ATTACHED DOCUMENT FOR PC-DMIS USERS.

With a cad model in pc-dmis you can report out the T-Values, it is great for complex surfaces. You produce that value from a surface plate layout. It in simple terms the devarition from nominal square to the axis. Well maybe not to simple, but that is how I understand it.

Have you seen the print?

Stijloor.

 

[justncredible]

Have you seen the print?

Stijloor.

You bet, a cad model in pc-dmis I would report out the T-values, every .100" all around the entire edge. Also I would be able to give moves to a machinist making the unit. Based only on the T-values.

 

[Stijloor]

You bet, a cad model in pc-dmis I would report out the T-values, every .100" all around the entire edge. Also I would be able to give moves to a machinist making the unit. Based only on the T-values.

What I meant was: have you seen the part drawing and call out that Arun Kumar was referring to?

Stijloor.

 

[SArun Kumar]

Hi,

What is my issue is that if u go through to attachment file page 2 of 2 there are some Profile GD&T for form & trim edges 0.8 @ Datums A B C are shown in general notes. So it means two parallel profile surface with displacement of +/-0.4 from mean (3D model) is our tolerance zone. I got it checked in CMM with Point to Point comparison with 3D model.(after best fitting with 3D model) the observations of some of points are available with attachment file page 1 of 2. In these observations I have max displacement observation is - 0.6737. (Highlighted in attachment) means the actual out of tolerance is -0.6737-0.40 = - 0.2737.
But when I am going making report against this
(Profile 0.8 against Datums A /B /C)= 0.6737 it seems ok.
How could i make observations report correctly?

 

[Stijloor]

Hi,

What is my issue is that if u go through to attachment file page 2 of 2 there are some Profile GD&T for form & trim edges 0.8 @ Datums A B C are shown in general notes. So it means two parallel profile surface with displacement of +/-0.4 from mean (3D model) is our tolerance zone. I got it checked in CMM with Point to Point comparison with 3D model.(after best fitting with 3D model) the observations of some of points are available with attachment file page 1 of 2. In these observations I have max displacement observation is - 0.6737. (Highlighted in attachment) means the actual out of tolerance is -0.6737-0.40 = - 0.2737.
But when I am going making report against this
(Profile 0.8 against Datums A /B /C)= 0.6737 it seems ok.
How could i make observations report correctly?

The datum features "B" and "C" are features of size and have the Maximum Material Condition (MMC) symbol attached. This callout affects the manner in which the datums are utilized. A possible "datum shift" comes to mind. Have you taken this in consideration when programming your CMM or similar?

Any GD&T Covers who want to chime in?

Stijloor.

 

[David DeLong]

Usually, when the secondary and tertiary datums are features of size (holes or pins) and referenced in MMC, most CMM Operators and their equipment end up using the datums holes in RFS. They, of course, would lose extra tolerance derived from the difference between the actual datum hole size and its MMC size.

Checking fixtures made correctly would end up with a cylindrical pin of MMC size in the secondary hole and tertiary holes unless one of the holes or both have a geometrical requirement and have a virtual condition size. In this case, the cylindrical pin would be at its virtual condition size. This is the reason we find differences between the CMM and the checking fixture. The CMM might reject the product while the checking fixture might accept it. A checking fixture made correctly is the mating part made in the worst possible condition. A properly made checking fixture overrides the CMM.

In all the years that I have been training in this subject (21 years), I only had 1 CMM Operator state that her equipment could calculate the extra bonus from the datum holes at MMC but that is it. All other CMM Operators confess that they measure the part with the secondary and tertiary datums in RFS.

Hope this helps.