Position Tolerance - Hole Pattern 1

[jimbod20]

Have a look at the attached interface drawing.

Datum C in the machine definition will satisfy both location and orientation as defined in the customer interface drawing. Machine definition of Datum C (orientation and location) WRT datums A & B is defined by a position tolerance of .002. Customer interface defines the orientation and location of datum C WRT datums A & B as .014.

The axis of the 4 through holes in the machine definition will match the customer interface defined via the position of .014 WRT datums A and B. The customer interface defines the position of 5 mount holes as .014 WRT datums A and B.

The 4 vs 5 holes leads me to the question.

Do I need to modify the FCF on the machine definition to remove or modify for the impact of the datum shift allowance at datum C to ensure the interface will match the customer definition? Will this allow the 4 holes in the machine definition (Datum C shift) to move beyond what is allowed in the customer interface definition? Would I just change Datum C to be RFS vs M or would I include an additional single segment FCF modifier?

 

[Burunduk]

Assuming ASME Y14.5:
If you are unsure about the effect of the difference from the customer's drawing when using the clocking datum, you can always add another position FCF with a value of .014(M) referenced to A and B(M) under the "4X" holes callout. This will create a simultaneous requirement with the bottom single segment of the datum feature C hole, effectively forming a pattern between all 5 holes, just like in the customer's drawing.

 

[drawoh]

jimbod20,

Your Datum[ ]C feature controls rotation only. Datum feature[ ]B controls[ ]X and[ ]Y. The [Ⓜ] on Datum[ ]B is probably not required if the feature of size is accurate, as it appears to be. Since datum[ ]C's feature is one specific hole, your hole pattern's positional tolerance should call up[ ]C, probably at [Ⓜ].

My primary concern about holes shifting would be to verify your holes do not interfere with screws and bolts. Your fasteners are a keep-out.

--
JHG

 

[3DDave]

Why are there two different definitions for the same features?

Parts that are accepted per the Customer Interface Definition may be rejected under the the Machine Definition.

 

[jimbod20]

Burunduk,
I recognized this option. I appreciate the guidance/feedback. Please review these notes.

3DDave,
The customer interface definition matched the machine drawing early in the design process. The machine definition is our standard machined casting interface definition. My customer changed the customer interface to be consistent across the final assembly. I/we provide one part installed on the final assembly.

My concern is parts that I produce (machine definition) will be rejected by my customer (Customer interface definition). The customer interface does not allow the datum C shift tolerance allowed in the machine definition. I need to fix my machine definition.

I see two options.
Single segment FCF add as Burunduk outlines.
Can I just remove the Datum C position shift allowance on the machine definition CⓂ? The tolerance allowance on the machine definition is then the same as what exists on the interface definition? I think this is the easier way to make the machine definition match the customer interface definition.

 

[3DDave]

The customer definition is less constraining as it doesn't control clocking of the group at all, so it cannot reject on it. The customer definition may result in accepting parts that are unusable. I don't see why "F" is applied to only one hole; they all need some orientation and if "F" is good enough for one, it is good enough for all.

 

[jimbod20]

Thanks Dave. Appreciate the help.

All five holes in both views are located by the same basic dimensions. Datum B in both views is an assembly pilot diameter. I provide a sketch to illustrate. My apologies. I'm trying to mask the part.

Datum F is the tertiary cast datum. A is the primary machine datum (interface mount surface), B is secondary machine datum (mount pilot diameter) and C is the tertiary machine datum (clocking). D-E-F are the cast datums.

 

[3DDave]

C is clocked from F. Why add more tolerance buildup?

 

[Burunduk]

Here's a thought:

Does it even make sense to define a tertiary datum for clocking at RMB after a secondary datum at MMB that is supposed to allow a shift? It seems that in order to enable the shift, you first need to find the best-fit orientation that potentially approves the holes while using the shift from the secondary datum, and only then attempt to involve the simulator of the tertiary datum. Otherwise, the involvement regardless of the material boundary of the tertiary datum simulator, will negate the possibility of using the shift relative to the secondary datum, and may also override the precedence order related to the constraints of degrees of freedom.

 

[3DDave]

Great thought.

Yes, it does make sense, as long as the tertiary is not an FOS it will be fine. Just to be sure it could also use the Datum Translation symbol if it is an FOS.

 

[Burunduk]

I think if the tertiary is not an FOS it may be a planar face at some location from the secondary datum axis, and when it's referenced RMB (and without the translation modifier), it would effectively still "block" some of the datum shift relative to the secondary datum when the tertiary simulator is engaged. Could it be otherwise?

 

[3DDave]

What basis have you to think that?

 

[Burunduk]

Datum simulation rules.

 

[jimbod20]

The casting drawing controls cast surface profile to the cast datums D-E-F. The primary machine datum structure A-B-C is established with respect to the cast datum structure. Datum C is held to datum F. Machined features are then referenced to A-B-C. I stack from the cast to the machined datums to individual machined features to ensure I understand where the casting surface is with respect to individual machined features. The tolerance build up is cast to machine.

 

[3DDave]

jimbod20,

This is machining features relative to the cast datum features. Those machined features will limit the orientation of the cast features with respect to the mating part.

So,
D-E-F <- A
D-E-F <- B (one of them perhaps replaced by A)
A-B(M)-F <- C (one clocking hole)

but then the rest of the holes in the pattern are, expanded, [D-E-F | D-E-F | A-B(M)-F] <- Holes, rather than D-E-F <- Holes

The features that go directly to D-E-F can have one setup and less tolerance stack error introduced by the intermediate steps relative to the cast surface. A case where this would not matter is for a bracket that aligns between machined features but these appear to locate and orient the cast surface relative to another part.

Adding more steps makes tolerance analysis much more difficult, increases allowable variability in the result, and often does not even reflect how the part is used or manufactured.

If all the holes act as a group they should get a single control.

Burunduk,

"Datum simulation rules."

Which ones; in what order?

(see Figure 7-34 Planar Datum Feature Constraining a Rotational Degree of Freedom: Secondary Datum Feature RMB in ASME Y14.5-2018 for guidance.)

 

[Burunduk]

"but then the rest of the holes in the pattern are, expanded, [D-E-F | D-E-F | A-B(M)-F] <- Holes, rather than D-E-F <- Holes"

Does [D-E-F | D-E-F | A-B(M)-F] suggest the same "common datum feature" as primary and secondary in the same feature control frame?

 

[3DDave]

No.

Do you understand the example I gave the figure number for? Maybe lead off with figuring that out before working on the advanced material.

 

[Burunduk]

"No"

Then you will have to explain this line, otherwise what you wrote looks like nonsense.

 

[3DDave]

Me: "Do you understand the example I gave the figure number for? "

Burunduk: I wanna change the subject.

Me: Do you understand the example I gave the figure number for?

Burunduk: Uses another thread with a different example.

 

[Burunduk]

This will hijack the thread, I started a new one for this problem.