Runout or Perpendicularity?

[ModulusCT]

So, please if you will, see the first image attached below. We're looking at a misguided attempt to control 3 separate parts which are mated together (Press fit) and need to be mated with other coaxial parts. The two inner shafts you see are meant to mate with two separate grommets. Because of this, runout is being used (only locally mind you, at 1.20 and .236 - how we're supposed to know this for sure I have no idea - I guess datum targets are not understood here).

Now, the point is to insure that the two shafts, once pressed into the mating part are perpendicular enough, to datum feature D to insure assembly and proper functioning of the mating seals, which you cannot see in these pictures. I guess I'm wondering why perpendicularity isn't used instead?? The circularity of the part should be controlled at the part detail level to insure proper fit and function.

Datum D is a fiction. This is a cast part, and so, there is no flat surface for datum simulator D to sit on. No, someone would have to make contact with the edge of that feature somehow to simulate D. I'd rather use the bottom of the cast part.

If runout is OK to use here, I believe the implementation is off. I don't think you can point to a single spot and say only measure runout there. I believe it should look something more like the second picture I've included.

Please, your thoughts. And I'm sorry for any gaps in my understanding of these parts or what was originally intended with the first drawing. I've asked, and I get the good old "it's always been like that" as a justification. That, and I've only been here a month and a half. I was told when hired that they needed my help with their GD&T... Uhh, yes.

Please, please, please, use the correct terms!They're Datum Features NOT DATUMS!!! AAAAAHAHAHAHHAHAHAHHAAAAA" -- Don Day

 

[dgallup]

I've never uses an orthogonal datum for a runout specification (your datum D or H), what purpose does it serve?

As to whether the entire surface or just a location on the surface needs to be controlled depends entirely on the design and function of the grommets which you haven't shown us. Are they seals? Do they contact in just a small area or on the entire surfaces? If they contact on the entire surfaces then you may need to use total runout.


----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[ModulusCT]

OK - Well as you can see with the image below, the grommets are beaded where they contact the shafts, and so, they're only contacting at a line (ignoring deformation due to fit and pressure). Even still... Does it save anyone anything by specifying what is essentially a single line where runout applies, rather than the entire shaft? And is it even practical to specify a single line to apply a geometric control like that? If so, I think that at the very least, datum target points should be used to specify where the measurement is to be taken, and they should be located with basic dims. Right now, it looks like someone just sort of made up their own method for doing this.

The orthogonal datum is there to control perpendicularity of the shaft to that datum feature (D or H). And from what I've learned, that is what's truly important to the design. The circularity of the shaft is something that I think should be defined on the shaft print, not an assembly drawing.

In any case, regarding total runout... Yeah, well these shafts slide through the grommet up to a point, and so, I imagine the entire shaft should be controlled to some degree. So you may be right on that score.

The interference is intentional. In reality, these rubber grommets will deflect to form a seal on the shafts. The contact points are radii as you can see.

Thanks for the input!

Please, please, please, use the correct terms!They're Datum Features NOT DATUMS!!! AAAAAHAHAHAHHAHAHAHHAAAAA" -- Don Day

 

[ewans]

Hi ModulusCT,
As you may be aware, per Y14.5-2009, it is acceptable to apply a runout tolerance to a only a portion of a surface. This is designated using a chain line adjacent to the surface, dimensioned using basic dimensions.

Cheers,
Ewan

 

[greenimi]

I agree with Ewan.

I just only would like to mention that (because of the title of the thread) one of the differences between the total runout and perpendicularity (both of them could end up with the same mathematical solution and are mathematical equivalent) is that with the perpendicularity you can use the datum shift. MMC. LMC (I know might not be applicable in the case presented, but I am talking about in general) and with total runout you CANNOT (by definition must be RFS).

Man, that was a long sentence!!

 

[3DDave]

Perpendicularity cannot affect location. The alternative to a total runout control is a position control.

 

[greenimi]

Dave,
Thank you for keeping me honest.
I was talking about a surface nominally perpendicular to a datum axis.

 

[ModulusCT]

Guys, thanks very much for the info. This has helped a lot.

One last question. If I can't convince someone that we need to either use position (my preference) or at least mark out a zone for runout, and they instead insist on calling it out at a specified POINT on the part, what do you think would be the best way to show that on the drawing? The method on the first drawing (my reference) is not quite correct (no basic dims for one thing - also, how do you indicate that runout is to be measured at only one point?? I don't think it was ever meant to do that. I believe that it was always meant to indicate along a distance of some kind)... A chain line normal across the shafts with FCF leader pointing to the line?

If you can, please let me know if this is even legal, and if so, how one might indicate it. Being new here, I can spot many problems with application, but I can't really argue intent as I don't fully understand this companies processes yet.

Thanks again,

Mod

Please, please, please, use the correct terms!They're Datum Features NOT DATUMS!!! AAAAAHAHAHAHHAHAHAHHAAAAA" -- Don Day

 

[ewans]

Hi Mod,
If you opted to use runout at discrete cross-sections (vs. over a partial length), I would say that your proposal to use a chain line normal to the axis, located with basic dims, is a good approach, albeit perhaps not completely compliant with Y14.5-2009. For clarity, I would add a note such as "runout tolerance applies at indicated cross-section only" under (or next to) the feature control frame.

You definitely should not use a datum target symbol unless you're intending to make the shaft a datum axis feature.

Cheers,
Ewan

 

[ModulusCT]

Ewan, yes, thank you. I just corrected my previous post as I intended to say "FCF leader pointing to the line" instead of "Datum Target". You're right.

In any case, thanks... This covers everything for me. Awesome. If someone else happens to know what 14.5-2009 says about this and can point me to the section (I can't find anything about specifying a particular cross section only being measured for runout), please don't hesitate.

Have a good week guys.

Mod

Please, please, please, use the correct terms!They're Datum Features NOT DATUMS!!! AAAAAHAHAHAHHAHAHAHHAAAAA" -- Don Day

 

[dgallup]

It should be done with a chain line PARALLEL to the axis (not normal) as ewans posted in his graphic above. Basic dimensions should define the zone.

I think the benefits of specifying the runout only in the grommet contact regions are 1) reduced inspection as it takes more time to measure along the entire surface 2) simplified gauging as the indicators can be fixed rather than traveling and 3) less scrap as you are not throwing away a part that has higher runout in an insignificant location. Depending on how these parts are assembled, it may be possible to "dial in" the runout at a specific location and it will be higher at other spots.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[ModulusCT]

Hi dgallup, thank you for the pragmatic explanation. I had considered some of these ideas before making my changes. I do not want to disrupt the current way of doing things, only show the requirements on the drawing, per the spec.

Therein lies the problem. I still have no good ideas about how to show runout being applied at ONE single cross section on the part. How would you show this graphically? Besides the suggestion above that is... Ewans suggestion is a good one, and I will use it if I can't find the official way of doing it. I just prefer not to use local notes when I can avoid them.

So, yeah, a way to indicate a runout measurement at a specified location (NOT over some distance)... That's what I may need. I wonder though... If runout, or total runout are sort of, confined, to a single cross section, are they really doing their jobs? Especially as two datums are called out in the FCF. I suppose it's OK. The secondary, othogonal datum would simply add it's perpendicularity error to the circularity measurement, right? Or might it prevent a valid measurement to be taken at all? If the part is wobbling due to perpendicularity offset then it seems to me that the dial indicator will not be taking the measurement at the required point (at least, not entirely). Man I love GD&T... Can Really get your noodle twisted.

Please, please, please, use the correct terms!They're Datum Features NOT DATUMS!!! AAAAAHAHAHAHHAHAHAHHAAAAA" -- Don Day

 

[dgallup]

The only use of a second (orthogonal) datum is as a basis for the basic dimensions to where you are measuring the runout. It's skewness will have no influence on the actual runout measurement. Only one point of contact will establish the datum plane.

I think the only change you need to make to the existing drawing is to make the .236 and 1.20 dimensions basic (boxed) and add a note to the runouts saying to measure them at the gauge (basic) locations shown.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.