Using 1 hole from hole pattern as a datum 1

[Tarthrin]

This situation comes up a lot in our parts. Take a look at the attached picture. It is a cover that gets bolted down to the rest of the assembly by the four holes in the pattern.

Basically, say we have 4 holes that are the same size. The engineer wants to use 1 of the holes as datum B and one of them as their datum C.

1. Can I dimension all four holes and then call out one of them as a datum?
2. Having two holes with a datum each seems they would fight each other?
3. Any better ideas?

 

[Storaker]

Hi,

I don't belie there is a reason to refer to B and C in the hole potion tolerance.

By calling out |POSTION| 0.3 |A | instead, you will be saying the same thing in a more "standard" way.. The holes have to be in position relative to each other and dtm A. Note that while this secures the hole to hole position in the pattern relative to each other, the pattern position relative to the rest of the part is not toleranced. You can off course use dtm B and C to relate the rest of the models to the holes.

Rgds. Audun

 

[KENAT]

Do you have other features that then relate to B & C?

If so maybe make the entire hole pattern a datum ala ASME Y14.5M-1994 4.5.8 & fig 4-22.

Posting guidelines faq731-376

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

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[drawoh]

Tarthrin,

I am fine with your drawing as shown. Datum[ ]B controls X[ ]and[ ]Y. Datum[ ]C controls rotation.

--
JHG

 

[3DDave]

This callout makes feature B and C self-referential; for example, if it was not a pattern, but only the one hole used as Datum B, the position callout would not be [A|B]. The way it is defined the references to B and C do not add any information or special meaning and should be removed for clarity.

It is also missing the diameter modifier on the tolerance in the feature control frame.

 

[CheckerHater]

Everything depends on what you are trying to achieve.

If you are trying to align holes to each other, datum A is enough, no need for B and C.

If you want to align holes to something else, make that something else a datum.

If you don't know what you are trying to achieve, then it doesn't matter how you do it.

 

[Tarthrin]

Thanks for all the replies.

Generally we do not treat the holes as a pattern. And they are attaching the cover to the tray. The engineers typically use the bottom of the pack as datum A and pick two holes as Datum B and C. They then dimension everything back to the main surface and two main mounting holes.

While it makes sense to make this simple four-hole feature into a controlled pattern, this is example is the simplest pack/cover combo when have. Normal pack cover/tray hole patterns are 70+ holes around a non-flat multi-angle flange. Which makes it difficult to use as a single hole-pattern. I would like to try to use this example to be able to apply it to future designs.

One of my issues is the self-referencing that happens when you try to dimension all 4 holes with the same callout. It makes a bit more work, but it would make more sense to dimension the hole for datum B and the hole for datum C seperately, and call out the other two holes on their own.

Good catch on the diameter modifer 3DDave.

 

[CheckerHater]

If holes B and C are used to reference several other features, it makes sense to dimension them separately, that is "2 holes" and "2 holes".

Look at the enclosed picture from book by James D. Meadows.

 

[Tarthrin]

That is the answer I was looking for. It looks like that scheme would work for what we need to do.

 

[SeasonLee]

I prefer to use the hole pattern as datum feature, since there are some differences between the two design scheme, and I think the hole pattern datum will more accurately represents how the part mount in the assembly.

Season

 

[Tarthrin]

I think if we try to use a 70-hole hole pattern as a datum, it might give some suppliers heart-attacks.

 

[3DDave]

The problem with the original method was that it used holes that are part of the pattern as origins for locating the pattern.

It is perfectly fine to just select two of the holes as a datum, though it doesn't make sense unless the mating parts are primarily aligned to those two holes.

In the example given from Meadows' book, there is no increase in utility over just making all feature related to A alone. The way it is toleranced is not to describe how the part functions, but how to make the job seem more convenient for the inspector, but without making a difference in how easy/difficult it really is. To be more explicit, there is no difference in the design of a gauge to verify the parts between the two datum schemes.

In both cases the pins at B & C will be .300 -.005 -.014 diameter, whether they are called datum simulators or just used to inspect the related holes.

In the case of 70 holes - if 70 fasteners hold the part in place and other features require alignment based on all 70 fasteners, then one might argue that making the 70 holes a datum is reasonable; as a datum or not the 70 holes need to have their sizes and locations verified, which a gauge fixture would help with, and once the gauge is made, it's also a datum simulator, capable of using all 70 holes at the same time. No one would be happy if they got an important item aligned to find that the last fastener would not fit.

 

[tooldeziner]

My reading of the Standard is that the B datum, when applied to a cylindrical feature, whether a hole or a shaft, establishes X and Y planes. With the addition of the A datum (perpendicular to the B datum) controlling the Z axis, the tolerancing framework is complete. The C datum shown in your picture is not only unnecessary, but creates conflict with the B datum.

 

[CheckerHater]

cylindrical feature, whether a hole or a shaft, establishes X and Y planes

And those planes are free to spin around until you clock them with third datum feature. As "tooldeziner" you should know that there are always two dowels locating and orienting the plate.

 

[axym]

Hi All,

Here are some thoughts.

-I agree with SeasonLee that making the entire pattern the datum feature will give the largest amount of tolerance for the pattern. If that is how the part fits, with all of the holes acting equally, then it is appropriate.

-There are downsides to designing assemblies with multi-hole patterns as datum features, however. If you want a good fit between the holes and their fasteners (i.e. some clearance but not too much shift), the relative positions all of the holes will have to be accurate since any of them could be involved in the constraint. Accurately controlling the relative positions of large numbers of holes can get expensive.

-Multi-hole pattern datum features also make inspection more difficult because all of the holes need to be measured in order to establish the datum reference frame and then datum feature shift must be calculated. For 4 holes these downsides are not too bad but for 70+ holes it's a different story.

-I agree with 3DDave that the example in the Meadows book does not offer any real advantage over using the entire pattern as the datum feature.

-If you want a better fit between the parts in the assembly, it is better to choose two of the holes and design them to function as secondary and tertiary datum features (i.e. to constrain particular degrees of freedom and not clash)

-The secondary datum feature B would have a tighter size tolerance and tighter Perpendicularity to A. The idea is that the B feature constrains the X and Y translations with minimal clearance and shift.

-The tertiary datum feature C would have a tighter size tolerance and tighter Position to A and B. Ideally, C should be a slot with the centerplane pointing towards B. The idea is that the C feature constrains the last rotational degree of freedom without clashing with B for the X translation. Alternatively, C could be a cylindrical hole if a diamond pin is used in place of a cylindrical fastener.

-The tolerances on the other holes in the pattern can probably be loosened, as they only need to provide clearance for the fasteners and not participate in the constraint.

-This configuration is generally easier to inspect because the datum reference frame is based on a small number of features and has minimal shift.

-The downside is that two of the features in the pattern must be produced differently than the others, so different tools or programming must be used.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Michael854]

I don't know about a 70 hole pattern, but I've seen a 55 hole pattern on a Caterpillar large mining truck wheel rim.
There are several things to look at. If we are analyzing the wheel rim spinning around at high speeds after its attached to the knuckle, the the datum feature should be the pattern.
Likewise, if we are trying to determine how much float to have at each of the 55 studs, we'd prefer the pattern.
But, if we're talking about GD&T so that a Cmm operator can do a part layout quickly...and have a quality engineer present the results of that layout in understandable form...the we would choose an arbitrary B and an arbitrary C.

 

[Michael854]

For all you guys who think its always better to have an arbitrary B and an arbitrary C...

If you change a car tire, do you really tighten the "B and C lug nuts" first?