True Position? 1

[metaldork]

So I'm drafting up some process prints for a run of new tombstones we have in our shop. Aside from the surfaces being machined we are putting a dowel hole pattern on the faces. We are looking to control the location of these holes to the bottom and to the center of the tombstone, which I believe will be our primary and secondary datums. My question is what would be ideal for our third datum?

 

[pmarc]

Frank,
What makes you think that ISO does not like simultaneous requirement? Do you have any reference to the standard?

It is not that I am defending the concept. I do think it is less intuitive and I mentioned it only to show that there is a possibility of living without tertiary datum feature, however I just wanted to know where did this opinion come from?

 

[CheckerHater]

Simultaneous requirement has nothing to with creating your datum reference frame. It simply means that all the features that reference the same DRF make a pattern. You establish your DRF long before you decide if simultaneous requirement will apply. And without clocking feature your DRF (in this case) will not fly.

 

[pmarc]

CH, since all features referencing the same DRF create a pattern, they are all tied together regardless of how many degrees of freedom were constrained by the DRF.

There is this Tec-Ease tip showing idea of the concept. Though I have some remarks to this, it also offers interesting comments about additional considerations that should be taken into account before applying it.

http://www.tec-ease.com/gdt-tips-view.php?q=126

I am pretty sure there was quite hot debate about it way in the past on the forum.

 

[CheckerHater]

The part in Tec-Ease example can be allowed rotational freedom because it is "spinning" in real life.
How will you inspect "the tombstone" if you allow it to spin?
There is even better illustration of simultaneous requirement. Look at ASME Y14.5-2009 Fig.4-39.
Do you notice that part on the picture has TWO HOLES as secondary datum, while OP's example has only ONE HOLE for secondary datum? The devil is in the detail.

 

[fsincox]

Pmarc,
As I understand it is not their default condition, it seems it would have been easier for all concerned to just do it the ASME's way, if it made sense to them. I certainly can imagine there is dissention in both camps, though.
Frank

 

[pmarc]

Sorry, but I don't get you. I do not see a difference between Tec-Ease's example and OP's except the number of datums referenced.

As for 4-39... Well, this is quite a nasty example IMO, exactly because of TWO holes being the secondary datum feature. Please explain me how could the part's geometry look like if the profile callout was specified as a SEP REQT. Where would be the difference between having and not having simultaneous requirement for this particular part?

 

[pmarc]

Frank,
For your information - excerpt from ISO 5458:1998 "Positional tolerancing" (page 3):

http://files.engineering.com/getfil...-4e69-ad8f-4005bceb75f8&file=sim_reqt_ISO.JPG

 

[CheckerHater]

pmarc,
It's really sad that you don't see the difference.
The idea of using DOF is simple: you constrain part until it stops moving OR the movement of the part has no effect on measurement.
For example: you put flat part on the table - create primary "plane" datum. The part can slide and spin, but you still can measure parallelism because it only requires one datum.
Or the part you make on the lathe: when you rotate it, it still look the same, so you can allow rotational degree of freedom in your DRF.
4-39 is nasty, but legal. Part is going nowhere.
With SEPT REQ for profile part will look the same. It's just, if you decide to make gage to check the part, you can build two separate gages - one for profile and one for 4-hole pattern.
This is the best I can do to present my point of view. Feel free to disagree.

 

[fsincox]

Pmarc,
Is it your understanding that it is as broad as the ASME understanding?, this says "shown on the same axis", which is a good point for me to remember, but my whole contention is these things are part of the standard, but not even known, or followed, in most shops.
Thanks,
Frank

 

[axym]

CH,

Pmarc is right. It is not necessary for the datum feature references to constrain all 6 degrees of freedom, regardless of the part geometry. Parts can be inspected even if rotational or translational degrees of freedom are left open.

During inspection, any open degrees of freedom can be optimized. In other words, the part can be arbitrarily rotated/translated within those degrees of freedom. This isn't always convenient or easily repeatable, so a common practice is to arbitrarily choose additional alignment features provide stable constraint for inspection.

pmarc,

Regarding 4-39, as CH says there could be one gage for the profile and one for the 4-hole pattern if SEP RQTS were specified. This would change the accuracy requirement for the part, and allow the holes to be shifted one way and the profile to be shifted another way. The original SIM REQ would not allow this, as the entire part would have to be evaluated on the same gage (in effect, both the profile and the holes would have to be evaluated in the same DRF).

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Evan,
You are right about the datum shift effect. I forgot about that. Thank you.

 

[CheckerHater]

Evan,
I understand you are CMM kind of guy who believes that it's just enough to put part on the table for inspection.
Would you please clarify; when you "arbitrarily choose additional alignment features", do you in fact add missing datum feature(s) possibly not specified on the drawing?

 

[pmarc]

Frank,
Point for you on whether simultaneous requirement is implied for features shown non-coaxial on ISO prints or not. It seems it's not. I just looked to Henzold's book (figs. 20-29 and 20-30) and he says that in such situations clear indication "simultaneous" or "separate" must be given. Though it is just only his idea of dealing with this, not supported by any standard, I am inclined to buy this explanation. Thank you.

 

[axym]

CH,

Short answer, yes.

To inspect a feature on the CMM, a fully constrained "part coordinate system" needs to be created by aligning to features on the part. If the datum features in the FCF do not constrain all 6 DOF's, then additional alignment features are added. Since these additional features are arbitrary, it is quite possible that they are not specified on the drawing. These additional alignment features should ideally be convenient to measure and have a good chance of constraining the open DOF's in an optimal way (or at least close). Often the considered feature itself is chosen as an additional alignment feature.

Many (but not all) CMM softwares are able to use this arbitrarily constrained coordinate system as a starting point, and then "best fit" within the open degrees of freedom from the FCF.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[dtmbiz]

metaldork

Here is something to talk about. Pictures help.
I know you say square but hey, its a tombstone

 

[fsincox]

pmarc,
I actually wouldn't mind it either; my main objections come from the fact that, I can never use it in the real world. Just like 0 position at MMC and the envelope all fine concepts if the other guy plays by the same rules but I do not find any shop or inspection people (really the only ones are the ones here) that understand it to be the case. This puts me, as a responsible engineer, in the position of not having it anyway. I would rather start from the same position as they do since I have to anyway.
Frank

 

[fsincox]

dtmbtz,
you are thinking of the same tombstone I was thinking of!
Frank

 

[CheckerHater]

Thank you Evan.
Appreciate straight "yes" as always. There is one thing I still don't understand.
You say "To inspect a feature on the CMM, a fully constrained "part coordinate system" needs to be created by aligning to features on the part."
At the same time you support pmarc's opinion that creating fully constrained set of datums is "wrong".
DOF is mathematically precise way to create comprehensive datum framework. I understand "mathematics" may be a dirty word among some part of 14.5 crowd, but you in your line of work actually use and appreciate DOF, so I feel confused.
At least you made me realize that DOF concept is not widely accepted, even among "the best and the brightest"; Genium manual, for example, ignores it completely.
Well, nobody is perfect :-(

 

[fsincox]

CH,
I do not believe pmarc is saying it is wrong to have a full datum framework just not necessary "under the law" which I am forced to agree is correct.
Frank

 

[CheckerHater]

Frank,
pmarc actually used the word "wrong".
Technically it still legal to use plus-minus dimension for location, but I do not say using position is wrong.
ISO "likes" idea of using DOF pretty much. We could discuss it sometime.
BTW, I misspelled your name in other post. I am sorry. :-(