Not quite getting how one applies a position to a width of a non-FOS. Seems like more of a profile thing.

I agree, use profile. The angles and corner radii size will affect the width.

ctopher, CSWP
SolidWorks '19
ctophers home
SolidWorks Legion

TheTick and ctopher,

If he applies a tolerance to the width, it becomes a feature of size.

The width should be something like 20×30.2±0.2. If the size tolerance is is critical and tight, the problem is solved. If the width is sloppy and not critical, then profile is the right way.

--
JHG

Hi, JHG

From the drawing, this dimension 30.2 is applied to virtual sharps. I don't think this is a feature of size.

Best regards,

Alex

That internal square hole looks like a FOS. Since it likely engages a snap. it seems more relevant, as well.

Proposal:
Use position for the hole. Control outer profile relative to each respective hole.

AutoDesign123,

Do you have a view in which you could label the tabs with numbers? If yes, you could change the 30.2 to a letter, say D, and then tabulate the widths to show them as D-1, D-2, D-3, etc. You could do the same thing with slot dimensions if they also vary.

Thanks everyone for your responses. I've added an image to show everyone a bit more detail what I'm hoping to check. I need to make sure that the tab needs fits into the slot and there are 18 of these tabs of varying widths (31.9mm dimension in my example - Option 2). I've included what I believe are two options based on the discussion.

Option 1:
- use a surface profile between two points on both sidewalls of the tab feature
- would it be acceptable to use a view that shows all tabs and indicate each one with a note "TAB X". Then create a detail view of one tab and use this notation to express that the requirement applies to "ALL TABS LABELLED TAB X"?

Option 2:
- updated the callout to include the material thickness which makes it a FOS and appropriate to use the positional callout.
- negative of this approach is that I would need to dimension each tab feature
- the suggestion provided by pmarc (label the dimension with a letter and create a table) is great and would only require me to make one view

Anything that I've missed or that's wrong with either option?

No. Neither option is good.

A better approach is the one TheTick proposed. Use position for the hole. Control outer profile relative to each respective hole.

Why do you want to control the minor fillets? There is no mating features that come into contact with the fillets.

Dimension 31.8 is not clearly defined. It is also not a feature of size.

Tab location is "largely" defined by the sides of tabs rather the fillets.

Best regards,

Alex

Quote (jassco)

Why do you want to control the minor fillets? There is no mating features that come into contact with the fillets.

The sidewalls of the tab at its base (including fillets) define the position (left/right) of the tab in the mating feature (see image below). The hole engages with a snap but that snap does not dictate left/right position of the tab.

Quote (jassco)

A better approach is the one TheTick proposed. Use position for the hole. Control outer profile relative to each respective hole.

I'm struggling in understanding how to implement this. I understand that I can define the position of the hole relative to the datums. To "control outer profile relative to each hole" means to me that I would need to create new datums using the hole and then use surface profile to control the tab walls back to the datums created by the hole?

Quote (jassco)

Dimension 31.8 is not clearly defined. It is also not a feature of size.

This has been mentioned lots so I did some reading of ASME Y14.5 to try and improve my understanding.

Quote (ASME Y14.5 - 3.35.2 Regular Feature Of Size)

"...one cylindrical surface, a spherical surface, a circular element, or a set of two opposed parallel line elements or opposed parallel surfaces associated with a single directly toleranced dimension"

In the image above, the wireframe view shows the two lines in orange that I've dimensioned. These lines are parallel ["two opposed parallel line elements"]. The tolerance isn't in the dimension directly but is covered by a default tolerance from my title block. If you exclude the lack of a direct tolerance can you help me understand why this would not be considered a regular feature of size?

autodesign

why is the tab tapered?

combine option one and two and use gage points mid point of the tabs to measure the width. also the angle of each side will have to be specified. then use a true position or profile of surface. way easier then measuring a theoretical internal sharp corners.
to measure gage points a cmm will be required. but the tolerance should be liberal. eg you could make d view and e view

examplehttps://res.cloudinary.com/engineering-com/image/upload/v1650661595/tips/tab_kam18d.pdf

• https://files.engineering.com/getfile.aspx?folder=25001581-e852-4f06-9c15-e

AutoDesign,
In your first image, there is an MMC modifier that follows the position tolerance value. It would actually make sense to apply it at the base of the tab so that the width includes the entire feature that is being engaged with the slot - including the two radii (according to the assembly depiction image you posted later down the thread). Then theoretically it is possible to inspect using a hard gage made as your worst case mating part. Easier and functionally more appropriate than having to engage the top tangency lines between the angled surfaces and the radii (the ones you've shown in orange).

Root corners are NOT FOS. Maybe... under the most extreme definition of the term, but not under the common competent understanding of GD&T.

Chasing down root corners that intersect an organic surface is a fool's errand. Ditto for measuring off a tangency.

In cases where I can not measure a boss diameter or width at the end, I specify a basic distance from the end and measure at the section. CMMs can do this no problem. Also gage-friendly and repeatable.