Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations waross on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Secondary datum feature does not meet a geometrical requirement wrt primary datum

Status
Not open for further replies.
pmarc

pmarc

Mechanical
Sep 2, 2008
3,227
This may be a stupid question, but I will try:
Let's say a part similar to the one shown in fig. 4-26 of Y14.5-2009 has to be inspected and there is a physical datum feature A & B simulator available in order to define proper DRF and to mount the part on it during a check of positional callout on two M20 threaded holes. There is only one part manufactured and will be no more. Imagine also the datum feature simulator is perfect, so there is no chance that bad part will fit on it and that good part will be somehow rejected.

Knowing all of that, how should inspector proceed if it occurred that as-produced part did not fit to the simulator, or in other words if the secondary datum feature pattern of 4 holes did not meet positional requirement wrt primary datum A? Should the inspection process be stopped immediately and the part sent to re-work (if possible)?

Any suggestions, thoughts to share?

PS.: This has nothing to do with the threads about extention of principles and IT# in ISO FCF's. At least at the moment I see nothing common :)
 
Replies continue below

Recommended for you

Sort by date Sort by votes
If the part did not meet the positonal tolerance to become datum B, then one does not have a datum B and other features that reference datum B should not be checked. The part would be rejected.

Dave D.
 
Upvote 0 Downvote
I agree with Dave. Since we are assuming a perfect simulator and the part does not fit on it, then the simulator is doing its job.

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II
 
Upvote 0 Downvote
Perhaps ... perhaps not. How do you rework? Randomly enlarge the holes until it fits the fixture? Ideally, you would use the two M20 features to establish where the four Datum-B holes should be, mark the locations and then rework. If the value of the part is minor, don't know if this makes sense. if the part is expensive (think a titanium workpiece), then it's worth the effort.
BTW, this kind of thing used to happen a lot at the engine factory for one of the Big-3. At one point, something like 6 out of 10 machined castings didn't pass inspection, and were scrapped. An observant quality/inspection person noted that everything seemed shifted/rotated wrt the datums (i.e. an issue with setup error using the cast datum features). He proposed a method of adjusting machined datum features in a rework, but was rejected by the OEM. He went on to privately develop the software to determine the compensation and licenses it to various suppliers now, including the OEM he used to work for.

Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
 
Upvote 0 Downvote
I agreee that the part would be rejected, or reworked. Keep in mind that pmarc said that only one part is to be made, so reworking it shouldn't be a huge deal.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
 
Upvote 0 Downvote
Guys,
Thanks for replies. There is no doubt that the part should be rejected, however I am thinking of something else:

Think of following scenario.
1. Assume that re-work is possible, but the less optimization loops the better (due to cost, timing, etc.)
2. Only one of 4 holes within datum feature pattern B is out of positional tolerance, but that is enough to be unable to mount the part on the simulator.
3. Imagine that there is much more features than just two threaded holes controlled by positional callouts to |A|B(M)| and some of them look pretty fine, some not - I mean, if the component could be mounted on the simulator, part of the features would meet their positional tolerances without any problem.

But how can an inspector verify which features are OK and which are not and require re-work together with the fourth hole of datum feature pattern, if he is not able to mount the part on simulator? Should the re-work be done on the fourth hole first and in the second loop (after loosing time and money) on the other features falling outside their tolerances?
 
Upvote 0 Downvote
This is where your MRB process comes in, if you have one where you work. In my opinion, if the simulator is what determines whether or not the part passes, then the inspector really has no way of determining what is actually out of tolerance and the part should be sent to MRB. It would be up to them to disposition the part from there.

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II
 
Upvote 0 Downvote
How do you rework? In short – paper gage.

Inspector takes longer, more complicated, labor-intensive way of checking part with general purpose tools like granite table and height gage. The found hole positions are placed on the diagram. After he is done, he will know which hole is off and how much.

Checkers…
:)
 
Upvote 0 Downvote
Okay, so I will rephrase the question to something more general:
Should the features positinally related to DRF |A|B(M)| be inspected at all, if the datum feature B is out of its geometrical tolerance wrt A? Regardless of inspection method. Regardless if the part could be re-worked or had to be manufactured from the very beginning.
 
Upvote 0 Downvote
I am not quite sure where are you leaning here.
There are politics. Your boss may override your decision and render your pain and suffering irrelevant.
As a perfect inspector in a perfect world you should turn part back to machine shop as soon as part didn’t fit into the fixture. Let the shop decide what went wrong and if they can fix it for less than making new part from scratch.
If you are looking from purely theoretical point of view, trying to reverse your framework, turn features into datums and datums into features and then hope design intent will be preserved; I personally wouldn’t go there. Too complicated, too dangerous and may take more time than making new part from scratch anyway.
So I really see no benefits to continue checking the part with important datums missing.
 
Upvote 0 Downvote
pmarc,

No, they should not be inspected since there is nothing to inspect to. You DRF doesn't exist because the secondary datum feature is out of spec.

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II
 
Upvote 0 Downvote
And this is the kind of statement I was waiting for. Thanks.
 
Upvote 0 Downvote
Powerhound is correct in my opinion. If the positional requirement that creates datum B is not in specification, you do not have datum B and features relating to datum B should not be confirmed. I don't know how "it depends on the value of the part"?

Dave D.
 
Upvote 0 Downvote
pmarc,

Are you really asking a general question on whether or not a geometric characteristic can be evaluated (for conformance and actual value) if one of the datum features is nonconforming?

Evan Janeshewski

Axymetrix Quality Engineering Inc.
 
Upvote 0 Downvote
Dave,
If the value of the part is high enough, then you can sort of backwards-process to get datum-features that work (i.e. rework them). It's maybe a complicated undertaking but it may be more cost-effective than starting from scratch. Reject parts were typically scrapped in my background by company policy, and we ate a lot of costs for that reason. If we could have salvaged a $20k plate by reworking datum features to accommodate the rest of the features, we could have saved a lot of money. It is a process that is done more frequently than one might expect.

Jim Sykes, P.Eng, GDTP-S
Profile Services TecEase, Inc.
 
Upvote 0 Downvote
Yes, Evan. This is my question.
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

M
  • Question
Primary Datum Parallel to Hole Axis 1
Replies
9
Views
2K
pmarc
pmarc
SeasonLee
  • Question
Datum callout question 2
2 3 4
Replies
75
Views
12K
mfgenggear
mfgenggear
MPSU
  • Locked
  • Question
Primary Datum on two planes
Replies
7
Views
1K
MPSU
MPSU
nmine
  • Question
Datum on hole pattern
Replies
6
Views
3K
Burunduk
Burunduk
CTengIS
  • Locked
  • Question
Irregular Feature Of Size as Datum Feature on CMM
2
Replies
20
Views
9K
CTengIS
CTengIS

Part and Inventory Search

Sponsor

Back
Top Bottom