True Position Problem in 3.7MR3

[dccfreak]

I wrote a program against a CAD model yesterday in 3.7 MR3, and I noticed that a 3 hole set that was visibly measuring high on the model, came out perfect after I did the true position dimensioning. I usually say yes to carry nominals back to the features in my programs, but, this one actually changed the measured values!!!???

I can see clearly that in the feature itself, the actuals are correct, and if I dimension them simply using the location feature, they are correct. I haven't seen this issue before, but, I am using read position on the holes as they vary quite a bit, and find hole is just causing me headaches. Could the read position action be doing this? I have tried answering yes and no to the save new values as target dialog, but, it doesn't make any difference. Is this a 3.7 thing or is it the read position?

TIA

[Matthew D. Hoedeman]

I'll bet you are using DATUMS in your TP, if so, STOP doing that, it does not do it correctly (as far as 99% of the users are concerned). It will shift and slide the feature AND the datums to get the best looking feature it can by using the MMC of the datums. If both datum features are 0.2mm big, then it will shift the internal alignment as far as it can to keep the datums as if on solid pins and get the dimensioned feature as close to zero as it can.

SO, that being said (twice), make an alignment BEFORE you TP a hole to the datums you need for the TP callout, then just do the TP, NO DATUMS.

[dccfreak]

Thanks Matt, but, I need to use datums as I have to replicate the assembly for this part so I want the shift. The problem is not actually relative to the shift though, but in how high the holes are from the the datum surface. The holes in question are on a perpendicular workplane to the datum structure. I could deal with it if the shift were the culprit, but, I haven't known there to be much shift in surfaces with datums on, but, I have never noticed it either. If I don't choose to carry the nominals back to the feature when I type them into the dimensional text, then this doesn't happen, but I like the theo values to match my nominals, guess I will have to live with that.

I also noticed it doesn't do it if just measure the hole in guess mode rather than using an auto circle??!!??

[Wes Cisco]

You are aware I hope that if you use the datums with a material modifer on the datums you are only getting an attempt at replicating a hard gage, that is a go or no go answer and not real measurements. There is a document I will attach that explains this. HTH

[dccfreak]

Thanks for the reply Wes. I have that particular document, but thanks for the help. The assembly that the datums mate with is lined to the mating part by the use of tapered draft pins similar to what might be used on a hard gage, but the shank of the pins are long enough to pass through and then the assembly is welded. Bolts are then passed through the assembly at a later stage in the process (kind of a strange set-up).

At any rate, I experimented with this feature and received the same situation. If I do not select to carry the nominals back to the feature after typing them into the nominal in the TP text, then the actuals are equal to that of the standard x and y location feature. After trying that, I simply turned off the read position feature in the auto feature dialog and set the theoreticals to where the hole actually was rather than the nominal values, and now when I type the nominals into the TP text, whether I select to carry the nominals back to the feature or not, the actuals remain correct. I can't figure out for the life of me why this is, but, it seems that the read position feature was the cause of my trouble. I think I am going to stay with 3.5 for awhile longer until I can figure this one out as it didn't matter what I did there, everything worked the way I needed it to.

Thanks again,
Brad

[dwade]

I just ran into this very issue last nite and it followed thru to this morning. I have 2 bores in line with each other and am calling out TP from one bore to the other applying MMC from the feature and the datum diameter. The measurement results were 0.0014 in the Y and .0004 in the Z and when TP was calculated it was showing .024"!! When I removed the MMC modifier for the datum it reported .0026 which is the correct calculation. Our FCF is | TP | dia0.000M | H | Gm | so I need to include as much tolerance as I can get.

[#2#]

I just ran into this very issue last nite and it followed thru to this morning. I have 2 bores in line with each other and am calling out TP from one bore to the other applying MMC from the feature and the datum diameter. The measurement results were 0.0014 in the Y and .0004 in the Z and when TP was calculated it was showing .024"!! When I removed the MMC modifier for the datum it reported .0026 which is the correct calculation. Our FCF is | TP | dia0.000M | H | Gm | so I need to include as much tolerance as I can get.

Wait, your TP tol. zone is a diameter of .000?!? So, the only way a part can be acceptable is if the bore dia. is less than or equal to LMC and greater than MMC?

[craiger_ny]

Wait, your TP tol. zone is a diameter of .000?!? So, the only way a part can be acceptable is if the bore dia. is less than or equal to LMC and greater than MMC?

Yes that is not uncommon, I don't like it but it is legit.

[dwade]

Medical field and I see that a lot. That and concentricity at 0.000m. Most of the holes/bores are .002" or .005" tolerance so it is close.

Any ideas as to why the coords changed? I changed about everything I could think of... work planes, measured cylinders vs. constructed cylinders, remeas diameters, no luck with any of it. Without MMC from the datums it comes out fine, with MMC it jumps around out of spec and will not repeat even in direction.

[d.evans]

comment and a question

I'll bet you are using DATUMS in your TP, if so, STOP doing that, it does not do it correctly (as far as 99% of the users are concerned). It will shift and slide the feature AND the datums to get the best looking feature it can by using the MMC of the datums. If both datum features are 0.2mm big, then it will shift the internal alignment as far as it can to keep the datums as if on solid pins and get the dimensioned feature as close to zero as it can.

SO, that being said (twice), make an alignment BEFORE you TP a hole to the datums you need for the TP callout, then just do the TP, NO DATUMS.

I've noticed that when I attempt to "Use Datums" in the TP dialog, a lot of the time, the measured values have 0.0000" deviation in all axes on all the holes TP'd that way. It screams that there is something wrong. So, I've gone back and removed the "use datums" and used an alignment equivalent to the callout (even when the callout has a MMC modifier on one of the datums) and so far, everything has come out OK.

What I'd like to know is, what is the mathematical calculation of MMC on both the feature and a datum?
I know how to do it using the MMC on the feature itself, but not sure on when a datum has a modifier as well. I have an excel spreadsheet that has the TP calculator on it, and I'd love to modify the formula to account for MMC or LMC on a datum.

[d.evans]

also, I've downloaded version 4.1 but I haven't installed it yet. Does it handle true position with modified datums accurately? Is it worth the trouble to go with 4.1? Or will I introduce a new set of problems to my job if I do so?

[dccfreak]

I tried this in 4.1 and it seems to handle it okay, but, like I stated earlier the issue in my case is that the deviations are being skewed coming from the plane, not the datum hole axes. I know that this issue does not present itself to me when using 3.5 MR2, as I went back to using that version as soon as this problem came up.