Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

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

Washer madness 1

Status
Not open for further replies.
Kedu

Kedu

Mechanical
May 9, 2017
193
Washer_j_uv2lhr.jpg


New on the form. Glad I found this site.
Story: shown washer received from supplier, measured by QE department found non-conforming, disposition rejected.
Supplier yelling the parts meet the print.
Inspectors using X min and X max methods to determine the worst case.

Question: what would be the theoretical values (absolute minimum and absolute maximum)?



Supplier states:

x min.: .7285

X Max.: .7745



QE/Inspectors :

X min.: .7335

X Max.: .7665.



Parts measured on CMM are closer to the supplier values (and obviously they are defending their calculated numbers)

Algorithms used: maximum inscribed and minimum circumscribed circles (not LSQ)


Which theoretical values you agree with (if any)? Supplier or QE/Inspector?
 
Replies continue below

Recommended for you

Sort by date Sort by votes
Since you are using the max inscribed and min circumscribed methods, then it doesn't equate with measuring X, because there could be form error on the ID or OD. (Think of a sudden bubble in the shape of one of those circles.)
 
Upvote 0 Downvote
Kedu,

You failed to state the drawing interpretation standard. Assuming ASME Y14.5-2009, I believe the supplier has correctly calculated the wall thickness extremes allowed by the tolerances.

Similar questions have been discussed here quite a few times. One recent example is thread1103-433578, and that should lead you to more.


pylfrm
 
Upvote 0 Downvote
Kedu,

As per ASME Y14.5, I get xmin=0.7335. This occurs when you have the largest ID, the smallest OD, and maximum positional error. The value xmax occurs under a different set of conditions, and I am not sure I would care about it.

--
JHG
 
Upvote 0 Downvote
Belanger,
That is exactly the gist of my question: to correlate theoretical value with CMM measurements (specified algorithms) and open setup inspection measurements (x min and X max)
Then first step has been (and still is an open question): what would be the theoretical values?

Pylfrm,
A master document (used by company accords the board) specifies ASME Y14.5-2009. So, you are correct.
Thank you for related discussion.
May I ask you how you came up with the conclusion that the supplier is correct? The examples shown have no LMC combined with RFS. Should I understand it is the same as LMC on ID combined with MMB on OD?

Drawoh,
Our QE got the same x min. values as yours, but not the supplier. Looks like pylfrm, agrees more with folks from manufacturing hence a state of confusion persists.
 
Upvote 0 Downvote
Such problems with using 2-point measurements as stand-ins for functional gages combined with non-functional datum references. At the least, I am unable to envision a case where there is some mechanism that centers the washer on the OD and then is concerned with the LMC of the hole. It makes me wonder what problem the originator of the requirement was trying to solve.

The distance from the center to the maximum OD is 3.005/2 (this is the center of the RFS datum feature)
The distance from the maximum OD to the far side is 2.995 (if the part is oddly shaped, there could be a dent or flat)
The distance from the center to the LMB is (1.508+.02)/2

so 3.005/2 - 2.995 + (1.508+.2)/2 = -.7285.


No one else seems to believe in showing their work. Any reason for withholding that info?
(edit to fix typo in calculation)
 
Upvote 0 Downvote
Full transparency: here is my work:

VC for ID: 1.508 + .020 =1.528
RC for ID: 1.492 -.016 -.020 = 1.456
X min: [2.985 (RMB) for datum feature A – 1.528 (VC)] /2 = .7285
X max.: [3.005 (perfect form at MMC required) – 1.456 (RC)] / 2 = .7745

I will feel “safe” if one of the experts around this forum ratify (or rebuke for that matter) my calculated values.
 
Upvote 0 Downvote
3DDave,

I used a spreadsheet...

[pre]=(3-0.005)/2-1.508/2-0.02/2[/pre]

...and I got 0.7335.

Remember, it is LMC/B of the OD, not MMC/B.

--
JHG
 
Upvote 0 Downvote
Agree with greenimi approach.
------------------
min x = { LMC of OD - outer boundary of ID (aka VC) } /2

min x = (2.995-1.528)/2

min x = .7335
----------------
max x = { MMC of OD - inner boundary of ID (aka RC) } /2

max x = (3.005-1.456)/2

max x = .7745
----------------

VC = 1.508 + .020 = 1.528
RC = 1.492 - .020 - .016 = 1.456
 
Upvote 0 Downvote
AndrewTT,
Looks like you agree with my approach, but not with my results [surprise][ponder]
x min is different versus what I got.
 
Upvote 0 Downvote
Drawoh, it's the center of the RFS condition, which means that the basically max diameter part can have a notch or a flat. If the parts are guaranteed to be perfect form at every section at every size, then it wouldn't matter. Which is why using 2-point measurements for this task is not useful.
 
Upvote 0 Downvote
For what it is worth, I agree with greenimi's numbers (I would not just call 2.985 "RMB", but rather "LMB" of datum feature A.)

When it comes to the value associated with the OD in the stack for Xmin, it should not just be 2.995. As 3DDave said, the OD can be oddly shaped (there could be a dent or flat) and that will affect the Xmin value. We went through this kind of exercise at least a few times in last 2-3 months on this forum.

I would, however, like to ask the OP additional question. What is the purpose of this exercise? I hope the inspectors do realize that even if they check that the value of "wall thickness" falls within calculated <Xmin;Xmax> range, there can still be a non-conformance against one the three characteristics used in the calculations.
 
Upvote 0 Downvote
So are we saying that the OD can be made as small as 2.995 and then a local form error of .010 (from rule #1) can get you down to 2.985 locally? Trying to figure out how the 2.985 is showing up.
 
Upvote 0 Downvote
AndrewTT,

The local size is still 2.995, but only in a strategic spot. The actual mating envelope is at 3.005, thus the local bite/dent/flat can penetrate the AME with the full local variation allowed.
 
Upvote 0 Downvote
The more important question is whether the subject parts will work in your assembly.


Mike Halloran
Pembroke Pines, FL, USA
 
Upvote 0 Downvote
All,

Thank you for the enlightenment. Look like supplier is correct. So be it.

Pmarc,

We have these parts in inspection and are measured for size (ID, OD, -caliper, ring gage and cylindrical pins), but when comes for position at LMC no gage is available . So, consequently (should I say naturally) inspectors asked for a min/max values that can be measured with a caliper. (Another thought floating around has been to put these parts on the CMM, on a regular basis, but that “brain storming” - suggestion has been quickly scrapped and considered ridiculous for a long term)
Therefore, calipers are good enough (wrongfully or not) – since supplier is also checking them with the calipers.

The dilemma has been:
What are the values inspectors should abide to? and from here things have gotten out of control hence my original questions.

May I ask, what are the risks you foresee using this inspection method-calipers—correlated with correct x min/x max values? Accept bad parts? Reject good ones? Either one? Both?
 
Upvote 0 Downvote
Caliper inspection cannot perfectly represent a comparison against boundaries.

It's not clear what the goal of this dimensioning scheme was to begin with; there should be an engineering analysis showing the reasons for all the tolerances, but I expect there isn't one because the requirements conflict from what would be expected.

For example - the smaller the hole, the more restrictive it is in relation to whatever passes through it, but the greater the distance it will be allowed to be off-center, so it would not be able to be centered wrt to the datum feature. But the datum feature reference suggests the being exactly on center is the desired goal.

One could do this inspection optically - make a transparency that has the LMB of the hole completely blacked out and set it behind a mechanism with an iris that contracts around the OD. If any light is seen at the border of the LMB simulator then the hole is too far off center. A separate check would see that the hole ID is not out of spec.
 
Upvote 0 Downvote
Kedu,
Inspecting position tolerance at LMC with calipers is quite a simplification, but I understand why the inspectors want to take this approach (not that I agree with it, though).

If sizes of OD and ID are known it should be quite easy to roughly estimate actual positional error of ID with respect to OD based on measurement of "wall thickness" on at least one side of ID.

For example, if measured values of OD and ID along x axis are 3.000 and 1.500 respectively and measured "wall thickness" on one of the sides of ID is .730, this means (again, very roughly speaking) that the center of ID is .020 (.020 = 3.000/2 - 1.5000/2 - .730) off of its true position, resulting in total .040 of actual position tolerance error.

Notice that for this set of made up numbers OD and ID values are within specification limits, the measured "wall thickness" is within <.7285;.7745> range, yet the position of the ID is out of spec. (.040 vs. allowable .028).
 
Upvote 0 Downvote
pmarc said:
(I would not just call 2.985 "RMB", but rather "LMB" of datum feature A.)
Click to expand...

Thank you for correcting my terminology.

 
Upvote 0 Downvote
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top Bottom