Help with datum feature on a slot with non opposing edges 3

[sheevaraam]

Hi Everyone,

My questions pertain to the attached figure.

-I have a sheetmetal part.
-I have a face that goes against another part, and so am using that plane as Datum A. This face constrains Translation in Z, and rotation about X and Y.
-I have six staggered dog-bone type feature (see detail view A for what I call the dog bone feature).
-Four of these constrain translation in Y and rotation about X. So The secondary datum needs to be based on these features.
-Two other staggered dog-bone like feature constrain the remaining degree of freedom , i.e translation in X. So the tertiary datum needs to be based on these two features.

My questions are
1. What is the best way to select the Datuming scheme? Presently I have A as the primary datum, B-C as the secondary datum, and C-D as tertiary. With B being defined by two points on the dog bone, and C two points on opposite . Is this correct? Do you have a better for selecting the secondary on tertiary datum? Is there a way of using Composite datum so that I can use all four dog-bone features for secondary datum, and two vertical dog-bones for tertiary datum?
2. How do I tolerance the slot width (Detail A) I wan that dimension to be 1.3+/- 0.08. But I recall that in ASME14.5, the opposite edges should have some overlap to be dimensioned with size control. So what are the alternatives to the wway I have dimensioned it, since the opposing edges on the slot are staggered and not overlapping?

Thank you everyone in advance for answering!
Siva

 

[sheevaraam]

The title should read "....slot with non-overlapping opposing edges"

 

[sheevaraam]

Another Typo, Presently I have D-E as tertiary datum (not C-D). Thank you!

 

[PaulJackson]

1. What is the best way to select the Datuming scheme?

The first consideration in answering that question is found by discovering how this part is functionally constrained for location and orientation with other parts. Which features are engaged in what sequence with what fit (clearance/interference)?

Secondary considerations can identify substitutes for functional features to account for things like measurement instability, process complexity or process politics e.g.(fixtured assembly, inseperable assy processing, semi-finish purchases, etc.)... but they should only substitute if the stacked variation between the functional DRF - the substitutes and other features controlled is equivalent or less than the variation without them.

You are going to have to describe the mating parts and assembly sequence in detail for any of us to give you a good advice or just use that information yourself to decifer what degrees-of-freedom are limited or totally constrained in what order in the assembly.

2. How do I tolerance the slot width (Detail A)

Two of three rotations must be stopped to make that measurement on any of those features.

You could stop rotations for each individually reflecting the punch contour and breakout variation by doing an all-around profile of each slot to [A] but I doubt that that is a functional option.

Or... you could identify which actual surfaces functionally constrain the secondary and tertiary degrees-of-freedom from the assembly sequence or whatever... and define the slots in sequence.

First each slot for its "all-around" profile to [A].

Next those slot surfaces identified as sequentially stopping the most remaining rotational and translational degrees-of-freedom as possible for their orientation and location to one another via a pattern profile refinement to [A].

Next likewise for the tertiary surface(s) to [A|B].

Finally any remaining to [A|B|C].

Paul

 

[sheevaraam]

Hi Paul,

Thanks for your reply. In my original message I have actually described how the part is assembled and what the mating features are. May be I didn't do a very good job of it, so let me try again.

-The sheetmetal plate mates against a U channel (also sheetmetal) on the face that has been identified as Datum A. This will constrain one translation (Z) and two rotation (about X and Y).
-The mating part has 6 rectangular slab like feature. These rectangular features are 1.2mm X 9mm. They all go into the 6 dogbone like punched features (one of which has been enlarged in detail A)
- 4 of these rectangular features that mate into the punched features are horizontal, so they take away another translation (Y and rotation about Z)
- Two of the other mating rectangular feature take away the final translation X.
- The datuming scheme I have chosen is
Primary - A
Secondary - B-C
Tertiary - D-E

Please take a look at thhe picture and kindly comment on the above datuming scheme if it is the right way to do it.


As for the punched features themselvle. The dimensionio i am worrieed about is the 1.3mm width. Can I call out a profile locally just at that place instead of an all round profile. Because the other things in that feature are not that important. Thank you all again for your time!

 

[drawoh]

sheevaraam,

I can sort of see what you are trying to do, but I cannot see why you are trying to do it. You need datums that will fully constrain your part.

I assume datum[ ]A is the top or bottom face of your part.

Your datums[ ]B1 and[ ]B2 locate in the vertical, and prevent rotation of your part. Your datum[ ]D1 controls the horizontal position. All your other datums are not required.

The datums on opposite sides of your slots make no sense to me unless you have a Z[ ]shaped bracket.

You can define the width of a slot as a datum.

JHG

 

[KENAT]

Not delving too deeply 'cause I don't have time but-

"Can I call out a profile locally just at that place instead of an all round profile"

Yes, so long as you make it clear what portion of the profile the tolerance applies to. If just pointing to the surface without the 'all around circle' on the FCF leader isn't clear see ASME Y14.5M-1994 figure 6-13 for an example or how to be explicit. Also, don't fall into the trap that profile is only for 'tight tolerances' it's not. You can have a 'loose' profile tolerance. For relatively complex shapes with a lot of dimensions used to form them surface profile is often better.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484

 

[sheevaraam]

Thanks Drawoh and Kenat!

Drawoh, I was considering putting a datum across the width of the slot. But since the opposing planes of the slot are staggered, I read in a book that it should not be size tolarnced unless there is some overlap. THey say that only if there is overalap, I can call it out, so that the dimensioning can be done from the axis between two opposing planes. So I was looking for other options if that is not allowed. But if it is allowed, I will go ahead and dimension the width of a slot and call out a size tolerance.

 

[PaulJackson]

If all six tabs 1.2 X 9mm are fixed for position on the U channel mating part and enter all six slots simultaneously then both are a pattern on their respective parts.

On the plate I would specify the gap between the two opposing surface areas with a basic dimension then attached to one of the leaders of that basic I would put a pattern profile control 6X [profile-surf|0.08|A] to control all six gaps for their pattern integrity. Below that control I would hang the datum feature identifier making all six gaps (twelve surfaces) datum feature "B".

I would tolerance everything else on the plate to [A|B] including the remainder of the six slots that are supposed to be in clearance from the U channel's mating tabs.

I would employ a similar strategy on the U channel's mating tabs except I would use a limit dimension rather than a basic and a position rather than a profile if the tabs actuall had opposing surfaces.

From what you explained to me I think that is the way they function... but I have one question - Do the tabs get twisted or bent over in assembly? Just curious!

Paul

 

[drawoh]

sheevaraam,

Dimensions and datums extending across the widths of your slots are valid if you have something that (nominally?) contacts both faces at the same time. I would not absolutely say that the two faces must overlap, but it makes the concept way easier to implement, hence my remarks about Z[ ]brackets.

JHG

 

[sheevaraam]

Thanks guys! That was helpful. I guess for this case I would just put a size tolerance across the width and call it as my Datum. B on horizontal ones, and C on vertical ones.

Paul,
No , luckily we don't bend any tabs, because they are pretty short and have generous chamfers and fillets on the edges where they insert the punched feature. However, yes, it does look pretty overconstrained! In my defense, I did not design this part, am just making the drawings for the person who left.


Thanks everyone again!
Siva

 

[PaulJackson]

Sheevaraam,

It is grouping all of the gap features together in a pattern via the "6X" specification that enables the "conservation of documentation".

It is the pattern designation itself that links basic location and orientation of the pattern members to one to another and the profile that tolerances those linked basics. If you split up that single pattern into two... "4X" and 2X [C]... then you are going to have to individually identify which gaps are and [C], tolerance the 4 gaps to [A] and then tolerance the 2 [C] gaps to [A|B] then tolerance all remaining features to [A|B|C].

Paul