Tolerance strategy for precision molded parts

[PaulJackson]

Attached is a specification I propose for limiting the profile error permitted within a specified area... while permitting the over all tolerance to constrain the contour within the larger boundary. Naturally the lower portion of the composite is only oriented to the datum features.

Paul

 

[Belanger]

Interesting. Trying to picture this -- you say you want it controlled within a specified area, but 3 dimensions are given. Is it meant to be a volume and we're controlling the portion of the contour that passes through that volume?

Good stuff, though. I like these proposed scenarios that stretch our brains!

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[PaulJackson]

At first I had it 40X40 but then I realized that surface profile is 3 dimensional. Think of is as a narrower zone wraped around the basic contour that is permitted to translate (not rotate) within the larger overall zone (effectively controlling smaller relative local sizes, offsets, or other geometry.

Paul

 

[CheckerHater]

I think you have one 40 too many.

If you imagine tolerance zone translating along the vertical axis, it (kind of) becomes irrelevant.

 

[PaulJackson]

Certainly the "All Surfaces [Profile...] is controlling the 3D contour of a part... I think that a subset of that area would have to be a 3D volume (cube like or spherical) considering contour detail that has size, offsets, or other geometrey relative to depth don't you think?
Paul

 

[pmarc]

Paul,
Could you post a sketch or something showing application of such callout?

 

[MechNorth]

The profile control is 3-D (just as flatness is), however it can only be applied per unit area of the SURFACE (just as flatness can be refined per unit area), not per unit of volume. You are not controlling the tolerance on a volume.
Otherwise, yes, the refinement seems fine to me. Your refined zone floats within the primary zone, and is not required to be centered on the nominal, as the primary zone is. The refinement prevents sudden changes in the surface. I like it.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[PaulJackson]

pmarc,
The specification is intended to be a title block tolerance for precision moldings or castings akin to general tolerance allowances for categorical lengths. Its drawbacks are that it doesn’t have as many defined levels but its benefits are that it is defined relative to the datum reference frame and the tolerance value does not depend on how far a feature is from the chosen origin but rather how it conforms to the upper segment zone ALL OVER and how it conforms to the refined lower segment(s) zone(s) for contour oriented to the datum features (as necessary).

MechNorth,
ALL SURFACES or ALL OVER controls contour in planes XY, XZ, and YZ... don't you think?

 

[axym]

Paul,

This is an interesting idea. You're taking the concept of Flatness applied on a unit basis for planar surfaces, and extending it to apply to Surface Profile on curved surfaces. So any given "patch" of a certain size must conform to the 0.2 Profile tolerance. Is that right?

I agree with CH and Jim that it should be a 40 x 40 area and not a 40 x 40 x 40 volume. I understand that the surface exists in 3D space, but the surface itself is still a 2D entity.

Having said that, defining the 40 x 40 patches on a curved surface will be very complicated. The patches would all have the same area, but each would have a unique shape. It would be kind of like cutting out a 40 x 40 piece of tape, and then sticking it on the casting in different places. The math for this would be pretty far out, because you're taking the idea of unit-based Flatness on a planar surface (Euclidean flat 2D space) and extending it to unit-based Surface Profile on a curved surface (non-Euclidean curved 2D space). I think we might need to use Gaussian coordinates, which is a type non-Euclidean geometry used to deal with curved spaces. I wouldn't want to have to explain this to the guy at the casting supplier ;^).

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Paul,
I am with others saying that one 40 should be removed from the FCF.
I also like Evan's comment about difficulties in application of profile per unit area concept to curved surfaces. I suspect that none of ASME standards clarifies this in a way that the concept could be safely transferred to a real world application.

 

[PaulJackson]

OK... I think I see the objection the refinement area boundary is always planar but it can be viewed cross-eectionally from various perspectives xy,xz,yz,... Is that correct?

So if we drop the 40... do you think this has the potential to constrain basic geometry to a smaller refined profile tolerance in a specified area while constraining the over all to its specified boundary?

Paul

 

[Belanger]

A quick side note -- when identifying a datum target on a countoured surface, the datum simulator is supposed to be made to follow the basic contour, not merely be a flat surface. So just extrapolate that idea to this profile concept, and still use 2-D area dimensions for a contour (thus keeping only two of the 40 dims).

I think that summarizes much of our reaction to the OP, although it is interesting to imagine trying to describe the area of a really crazy contour with only two dimensions.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[ptruitt]

Rather than 40 x 40, why not 40DIA ? Is that better, worse, or the same?

Peter Truitt
Minnesota

 

[CheckerHater]

Spherical diameter?

 

[PaulJackson]

So this is the corrected concept.

 

[pmarc]

Paul,
As long as a reader of this note knows exactly how the smaller areas (40x40, 200x200, 20x20) are established on surfaces other than just nominally parallel & perpendicular to A|B|C DRF - which I think may be an issue in case of molded parts due to drafts - I think both options are fine.

 

[ptruitt]

Checkerhater:

I'm thinking of the surface patch where it is limited by a basic sphere.

Peter Truitt
Minnesota

 

[CheckerHater]

Peter,

I understood you; I am just not sure each and every shop out there will do.

I am always suspicious of forcing some mathematical trickery on people who are trying to make a living using limited set of tools.
Same with very vague or broad requirements like “profile of everything all-over” or “perfect part required at MMC”.

Yes, our drawings are supposed to be as process-independent as possible, but if you don’t know how your requirement will be measured, don’t put it on the drawing…

 

[ptruitt]

Thanks, Checkerhater,

I guess I am not sure how to check either Paul Jackson's refined approach or the variation I asked about other than employing analysis software that would analyze a point cloud. I keep thinking about the benefits of round vs. square tolerance zones, but I know that we are not talking 2D, here, such as the location of a hole on a flat surface… Nevertheless, it seems that a sphere will give more tolerance than a cube while maintaining the functional design requirements.

Peter Truitt
Minnesota

 

[CheckerHater]

Maybe in few more years when CMMs will be as common as calipers…

Right now I am not sure if you can program one to check “round” spots.