I've been doing some GD&T training lately and one thing that seems to complicate things for me is Rule #1 (aka the Envelope Principal). It states that at MMC, a feature must have perfect form and as it departs from MMC, a form error is allowed in the amount equal to the deviation from MMC. Unless there is a form callout that further refines and limits it. Say I have to add a flatness dimension for Datum A as shown below. The flatness comes in at 0.08, which is within tolerance. BUT there are at least 6 other features on this part that are dimensioned to Datum A. If the 23 +/- 0.5 for example measures at 23.45, that only allows me a form deviation of 0.05. The flatness is now out of tolerance. But how do I get PC DMIS to account for that? The flatness is still going to report in tolerance at 0.08, even though Rule #1 has now been violated. Worse yet, (and this is just rhetorical) how do I tell the shop supervisor that I have to reject his parts even though the flatness is reporting in tolerance when it's actually not?
The flatness of Datum A is not affected by the state of the 23 +/- 0.5 surface. The "23" surface is measured from the most extreme outside the part points of Datum A, and the entire surface must be within the 23 +/- 0.5 tolerance zone.
Also, MMC only applies to features of size. Planes are not features of size.
No, a plane is not a feature of size by itself, but the distance between two opposing planes is. And I get that an MMC modifier can only be applied to flatness of a feature of size, but MMC itself and the Envelope Principal is constantly applied to everything unless an independency modifier is used. If the 23 width from Datum A to that opposing surface is measuring 23.45, then that width is only 0.05 from maximum material condition. There is excess material present at either Datum A or the opposing surface or both that is bringing that width only 0.05 away from its outer envelope of perfect form. Therefore, a measured flatness of 0.08 would bring the width outside of the envelope. If what I'm saying is incorrect, then I'm getting some wrong info from the training provider that I paid a good amount of money to.
Pasted below is what the ASME Y14.5 says concerning the Envelope Principal (Rule #1). It does say that it doesn't apply if a flatness tolerance is associated with the size dimension. However the flatness tolerance in question here is not associated with size dimension itself, but rather one of the surfaces it's composed of. And as you said, a plane itself is not a feature of size. So Rule #1 still applies. It also says that a variation of form equal to the amount of departure from MMC is allowed. In my theoretical example, the 23 dimension measured 23.45. So I would only be allowed a form variation of 0.05. The flatness callout may be 0.1, but that amount only applies if the departure from MMC is 0.1 or more. Anything less, and Rule #1 would take over and I would only be allowed an amount equal to the departure from MMC. In this case, 0.05. So again my question is how do I get PC DMIS to account for this when reporting flatness? Or is that even possible?
I'm more of an ISO user but obviously we get ASME stuff as well so here's my take:
I think it comes down to how you think about a datum.
A datum is theoretically perfect (in this case it's a perfectly flat plane)
A datum feature is what's on the actual part, and will therefore not be perfect (i.e. have form error)
A datum feature simulator, is what we use to create our working datum from the datum feature, be that a surface table, or a CMM
Obviously on a surface table we have it as close to zero form error as possible, so when we place out part on it we're using an as-near-perfect simulation of the datum (external to material). We would then measure the 23±0.5 and as long as it measures (over the full surface) within those limits we're good.
With the CMM, we measure the plane with a suitable number of points, and from that we construct the appropriate type plane (a primary datum in this case, external to material also) to use in our alignment for the evaluation.
The Flatness of datum A is evaluated independently of this. Again, to use first principals you'd measure this characteristic separately.
I wish I could show you the training material I have that suggests that particular surface doesn't necessarily get evaluated independently when it comes to the flatness tolerance. But I'm sure it's copyrighted. However, it does come from ASME Senior Level certified engineers, so I can't just automatically disregard it. I can probably point them to this thread and ask for their take on it though. Trust me, I want to be able to apply that full 0.1 flatness to that surface and not have to worry about anything else related to it. As an inspector, Rule #1 really complicates things for me because I can't just measure something and see if it's within tolerance. I have to be constantly thinking about the envelope principal.
Rule#1 doesn't apply to the example you have given. As
NinjaBadger said, the flatness is associated with datum A which is a plane. Planes are not features of size. The 23 ± 0.05 is a linear dimension to another plane - also not a feature of size. For the 23 ± 0.05 dimension to be a feature of size (a width), the planes would need to be parallel and have opposing vectors.
I'm sorry to be a pain in the butt, but I don't see how this wouldn't be a feature of size. I know it's hard to tell from my snapshot, but I don't think I'm allowed to post the whole thing since it's a controlled document and all. So here's a rough sketch instead. Datum A is on one side of the part, and the 23.5 surface is on the other. The planes are parallel and have opposing vectors as far as I can see. It's the overall thickness of that section of the part. So what am I missing?
I think, for it to be a feature of size, that 23 dimension would need to be called out to BOTH surfaces, like:
|<----23---->|
To me, it is only calling out the non-datum surface, no different really than putting a PROF A 1.0 on the surface.
I understand where you're coming from. This company (aka Big Yellow) uses what I think is called zero plane dimensioning on their prints. I believe it's just a different way of calling out dimensions. Most of the dimensions are all referenced back to a set of three intersecting planes. In this case, plane 01, which also happens to be datum A. But that doesn't change the structure of the part, just the formatting of the dimensions. Or maybe I'm just super confused.
Edit: I guess I need to understand whether using zero plane dimensioning changes anything. It's not talked about much in any training I've seen, but I deal with it in almost every print I see because Big Yellow is dominant where I work.
