Impact of toleranced datum feature to other dimension controlled by datum feature

[mkiragu]

Hey,
I have read through threads, videos and literature but cant seem to get a straight answer so seeking some help. It is correct to say that it is not wise to have a part that has Datum feature tolerance larger than the positional tolerance of a hole?

Background:
I have a drawing of a part where the datum itself a profile of a surface has a tolerance of 3mm and the holes, which are present on opposite sides of feature have a positional tolerance of 1.0. Problem is the part is subject to twist and pitch on datum A, so throwing hole positions way out.

Am I wrong is assessing that the datum tolerance should be tighter and that the part needs to be constrained to remove twist to ensure it reflects machining? Part is large composite panel.
Thanks

 

[chez311]

It is correct to say that it is not wise to have a part that has Datum feature tolerance larger than the positional tolerance of a hole?

You are correct. A good rule of thumb is to have your datum features an order of magnitude more precise than the features which reference it. A tight tolerance which references a sloppy datum feature will cause your inspection department headaches for the reasons you mentioned. Theres no restriction against it in the standard, but it certainly has the potential to create issues.

 

[3DDave]

Not knowing the relationship between the surface controlled by profile and the hole location, it's not clear that there should be a problem. 3mm on a 10,000mm wide surface is likely good enough.

If the feature is used RFS then it matters less as the position is verified relative to the datum feature simulator/true geometric counterpart and not from the part feature.

If the free part distorts relative to the installed condition, then it's likely it needs to be constrained for validation of the tolerance requirements.

 

[mkiragu]

@3DDave
The part is only 1200mm width and 3100 long, so the 3mm surface profile significantly impacts the part when you get a twist. I will be recommending the part get constrained as inspection does not replicate machining and assembly requirements, but wanted to be sure before I made recommendations. Attached a crude sketch of how the part kinda looks like in attachment.

 

[drawoh]

mkiragu,

In general, you need to be careful about using features of size as datums. A crude rule of thumb is that fixtures should be ten times as accurate as the feature being fabricated and/or inspected. This applies to your FOS[ ]datum. If you use a sloppy feature as a datum, your part is not inspectable unless you specify tolerances at the datum feature's MMC.

If your datum feature is a 3mm profile, I don't think your [⌀]1mm positional tolerance is inspectable. Perhaps your drafter can apply datum targets!

--
JHG

 

[3DDave]

The rule of thumb is a little bit dumb. It should be that the variation allowed for inspection is subtracted from the tolerance allowed for the feature.

This example does not appear to have any features of size as datum features. The part should be placed into contact with the datum feature simulator - the location of the holes is relative to that simulator, not the part feature. If the part is deformed from it's installed shape when it is in the free condition, that's a separate problem to deal with.

 

[drawoh]

3DDave,

The rule of thumb allows me to decide whether or not to ignore the tolerances of my FOS datum. Anything less than an order of magnitude is trouble.

--
JHG

 

[mkiragu]

3DDave,

If your datum feature is a 3mm profile, I don't think your ⌀1mm positional tolerance is inspectable. Perhaps your drafter can apply datum targets!

I feel this is correct, but it isnt my/our drawing to change. I am trying to simplify the issue to explain to customer and all involved. there is an added complexity as Datum A is not even a geometrical shape/feature, it is an arc and plane from side profile, but that is another issue all together.

Question that am trying to understand is when a datum is generated from a datum feature that is meant to be flat but is not flat, lets say is concave, when it generates the plane will it be best fit of the extreme limits? That is the center plane of all the inspected points? This will help explain the impact of twist of datum feature generation.

 

[pmarc]

mkiragu,

This seems like a problem related with the fact that the part is subject to variation/deformation in free state, but the drawing doesn't properly capture that.

 

[3DDave]

mkiragu, You quoted drawoh, not me.

 

[jassco]

Hi, mkiragu:

Well, it depends on how the part is used. How is the datum feature A is used in the upper level assembly?

Best regards,

Alex

 

[drawoh]

...

Question that am trying to understand is when a datum is generated from a datum feature that is meant to be flat but is not flat, lets say is concave, when it generates the plane will it be best fit of the extreme limits? That is the center plane of all the inspected points? This will help explain the impact of twist of datum feature generation.

Yes, that was me you quoted.[ ]

Your datum features must immobilise your part. All the versions of ASME[ ]Y14.5 explain how planes, edges and points function as datum features. A plane contacts at three points. An edge contacts at two. If your feature is an unpredictable curve, you need datum targets.

--
JHG

 

[jassco]

This is what ASME Y14.5-2009 says on Non-rigid parts. Also see Fig. 5-14 "Specifying Restraint for Nonrigid parts.

4.20 RESTRAINED CONDITION
Unless otherwise specified, all tolerances apply in a
free-state condition. In some cases, it may be desirable
to restrain a part on its datum features to simulate their
function or interaction with other features or parts. To
invoke a restrained condition, a note is specified or referenced
on the drawing defining the specific requirements.
See Fig. 4-42. This figure illustrates a part that
should be restrained until sufficient reinforcement is
added to retain its design shape. In this illustration,
the restraint must be per a document referenced on the
drawing. In a restrained application, it is permissible to
use as many datum targets as necessary to establish the
datum features.

 

[3DDave]

A nominally planar feature may or may not contact at 3 points, even as the primary feature, hence the math standard goings on about stabilizing them.

 

[mkiragu]

Thanks Guys, its helped in confirming my thoughts that part needs to constrained on a dedicated tool and potentially look at the datum A tolerance values with customer.

 

[Woosang]

mkiragu,

There may be OEM requirements so called "fit up force", which allows you to apply some amount of force during assembly and/or inspection.
For example, maximum xx pounds per yy inches interval, or just simply a 'finger force'.
This requirement might be defined in a process specification which is invoked on your drawing.