Complex Angular Tolerance Spec and Analysis

[boon erman]

Hi!

First post here, alot of good info, so I'm excited to read and contribute!

I'm in a situation where I'm designing a new product...none of my prints are released for production yet. Anyways, I'm working on my print package and I'm having difficulty coming up with a tolerancing scheme that makes sense and can be measured.

The product consists of several components that are axially aligned, but can shift angularly as a torque is applied. Attached is a simplified view of what I'm dealing with.

Basically the variables I'm looking at are the following:

1. Angular tolerance of the position of tabs on part 1
2. Size tolerance of tabs on part 1
3. Angular tolerance of the position of the holes on part 2
4. Size tolerance of the holes on part 2
5. Angular tolerance of the position of the slots on part 2
6. Angular tolerance of the position of the curved surface on part 1
7. Size tolerance of the curved surface on part 1
8. And of course the actual assembly position of the two components in relation to each other (i.e. the can shift)

So you can see this is abit of a complicated problem. My initial thought on it would be to set up datums on parts 1 and 2 and then specify profile of a surface around each feature that contributes to that angular variability. The nominal CAD file can be considered to be the basic surface. The varibility could then be analyzed by actually modeling the MMC and LMC surface envelopes and assembling the components.

Any thoughts, recommendations, or experience with this type of desing specs/analysis?

Thanks,

Boon

[Kelly_Bramble]

As always, having a print to write on would be easier. Also, I do not know functionality; materials or target manufacturing process and these do effect optimal dimensional specifications.

Just for the record I'm not sure how part one fits into part 2 with the four protrusions - must be made of plastic.

As I have already said, my suggestions are being made with minimal knowledge of fit, form and function (FFF).

Part 1:

External diameter assigned Datum A, maybe a cylindricity, circularity or a tight limit tolerance to control form and size.

Option a - The top two tab width features assigned a limit of size then position tolerance relating the tab features back to Datum A. Declare the both tab features as Datum B. Datum B will give you rotation or an angularity reference for the remaining features
Option b - Define one or two (opposed to each other) of the protrusions as Datum B and use position tolerance to locate back to Datum A.

I think the protrusions are a more obvious choice for datum B as they relate to the protrusion cutouts on Part 2. Remember that I have very limited information on FFF. I don't know the relative mating importance of the tab feature though I suspect they matter within the target assembly.

Option a - Position tolerance the four protrusions back to the top two tabs Datum B secondary and external diameter A primary.
Option b - Position tolerance the two tab features back to Primary Datum A and then secondary datum B protrusion or protrusions.

You could use Profile of Surface to locate, orient, size and form control the entirety of the top tab features and either declare the two exclusive top tab features as Datum B or simply relate back in entirety to the protrusions defined as datum B.

PART 1

For Part 1, I recommend datum matching or mirroring relative to part 1. I assume you know that I'm talking about...

Since the cutout feature on Part 2 are likely to be the secondary datum B, relate the four threaded holes (I think I see screws in the threaded holes) back to Primary datum A internal diameter and secondary datum B cutouts.

FYI, your question was very significant for a discussion forum.

[boon erman]

Kelly,

Thanks for your thoughtful response. I have read up on using tolerance of position for the radial position of non-cylindrical features (page 158 of the Geometrics book "MMC with respect to a center plane and related to a datum feature") - and I am abit confused about how to apply TOP to my features of concern. For instance, my part 1 is a very simplified version of what I'm working with...in the real world the assembly tabs are heavily drafted and include radii. I'm uncertain how to actually put dimensions down (say for the width of the tabs) and relate those to datum center planes (to control the angular position of each feature).

Boon

[Kelly_Bramble]

(page 158 of the Geometrics book "MMC with respect to a center plane and related to a datum feature")

Yes, I think your referring to Mr. Foster's book which was written to the ANSI Y14.5-1982 standard. I would prefer to have this help session referrinusing my book "Geometric Boundaries II" - ASME Y14.5-2009. - hint, hint..

The example on page 158-159 has only a single datum reference "A". Rotation datum control is not included in his example because the tab features are equally spaced at 120 degrees and rotation control is not required.

I could not find an example in Lowell's book that included a datum reference specification that would control rotational location and orientation for non-symmetrical evenly spaced features. If you happen to have my book - Geometric boundaries, II I have an explicit example to your application on page 3.18 titled "Parts with Angular Orientation".

To be simple, when you have a pattern or individual features that are oriented about a datum axis and are NOT equally spaced or require an orientation specification relative is each other - then one or more of these features will need to be defined as an orientation/location reference for the other angular located features.

If using limit angle tolerance you would simply apply an angle tolerance with the require +/- angle tolerance. Be aware that this method is not as precise as a geometric tolerance control and that as the features move outward radially that the sweep tolerance increases.

To use a geometric datum reference between the pattern of features simply decide which feature is that most important relative to FFF or alignment in the target assembly and then define one or all of that pattern as a reference datum. That reference datum will be used to orient and location control the other pattern of features to the datum feature (s). I recommend using geometric position tolerance relative to datum "A" axis, and then Datum "B" the orientation control datum you select based on FFF.