True Position of a Group

[John Kugler]

The PC-DMIS help file defines TP of a group as follows:

1. Create a feature set of the holes from the Set menu option (see "Constructing a Set of Features" from the "Constructing New Features from Existing Features" section for instructions on creating feature sets.).

2. Select the True Position menu option. The True Position dialog box appears.

3. From the True Position dialog box, select the created feature set as the input feature, and select any other options used for the dimension.

4. Select the Textual check box from the Analysis area.

5. Access the Edit Window toolbar, and select the Report Mode icon. You are now able to see the deviation of each feature and the translation and rotation required to bring the features into tolerance. Results are always shown using Regardless of Feature Size (RFS) on Datum to see the maximum deviation.

My question is this, is this right?? I always had trouble believing the "Set" was actually the center of the group. I have two holes with the TP callout back to a center hole datum and I have to TP them as a group. Just wondering if this is the best way or should it be one hole to the other?

Thanks in advance.

[tking]

Hey Kev,

He just handed you a soapbox! You're on!

TK

[Johnson]

My question is can you use that as best fit too?

[William Johnson]

The only time that I have constructed a feature set was for profile tolerances.
So I don't really know how what you are doing works. When I want to do a true position of a hole pattern that is lets say (t.p. .005 A) A being the face perpendicular to the holes I use the alignment window and do a bestfit with translate/ rotate. Open the alignment window select bestfit, then select the holes in the pattern then ok ok. Alignments can be either two or three dimensional. If there is a (t.p. .005 to a,b,c) then you will need to select translate only or rotate only depending on what datums a,b,c are.

There is a very significant difference between a 2D and a 3D best fit.

A 2D Best Fit alignment requires an initial alignment. The alignment is created in the specified workplane and is based on the current alignment.


A 3D Best Fit alignment uses the raw (machine) data and correlates it to the theoretical values. It does NOT use a previous alignment but creates a completely new one.

This might be what you are looking for.

[Jan d.]

I presume you mean true position of a pattern (of holes).

I was told that any PC-DMIS version before 4.0 can not evaluate TP of a pattern to datums correctly. This is one of the reasons I use 4.0 now. And I am sure it does work in 4.0, although I think there are a few more kinks to work out by Hexagon.

Actually, I really like the new way to deal with GD&T in 4.0. I think that is a great step forward. Reporting I am not so sure. But the dimensions is great!


Jan.

[John Kugler]

Yes I know about the TP of a group to datums issue. I have my alignment at the datums and am trying to do a group back the alignment/datums.

At a User's Group meeting last week, B+S explained the TP issue in length and it makes sense to me now. 4.0 is the way to go.

[kbotta]

Please post a pic of your drawing, and I would be glad to help.
Kev

[kbotta]

Honestly, it would help if you posted a pic of your drawing. it always makes it easier - in determining the correct procedure...
With that in mind:
I beleive you are correct in questioning Pcdmis's help file for creating a feature set. Is your FCF controlling the axis created be the pattern? Doubtful. Do you have a composite FCF? Or a 2ssFCF? (again...Drawing...) My suggestion, w/out knowing the above variabes - would to be to dimension each hole ind.

Ok...Soap Box starts here>>>>>>>>

I really wish that the powers that be (whomever creates the algorithms for Pcdmis) would get with the committee members (or visa versa) and harmonize some of the GD&T used in pcdmis - to make them more compatible with what is on TODAY'S drawings. We shouldn't be driven by ISO req's either (or din, jis etc...). There are too many issue;s with the way Pcdmis reports 'GD&T' - that if you do not understand the results you are getting, you could be reporting something completly wrong. I've heard too many times - "well I hit the (insert GDT Char. here) button, so I figured it HAD to be right....
Well, I'm here to tell ya - your wrong....or could be wrong...
There are issue's. Period. You need to understand the limitations of 1) cmm's, and 2) Pcdmis.
I love this software. Best I have used by far - and support them (pcdmis) to the end. But...it can be better. And this is where we come in. Spread the word, or talking to tech. support. "We", can make a diffrence...
Peace.
End of soapbox>>>>>>>>>>>>>>>

Kev

[John Kingston]

Ok...Soap Box starts here>>>>>>>>

I really wish that the powers that be (whomever creates the algorithms for Pcdmis) would get with the committee members (or visa versa) and harmonize some of the GD&T used in pcdmis - to make them more compatible with what is on TODAY'S drawings. We shouldn't be driven by ISO req's either (or din, jis etc...). There are too many issue;s with the way Pcdmis reports 'GD&T' - that if you do not understand the results you are getting, you could be reporting something completly wrong. I've heard too many times - "well I hit the (insert GDT Char. here) button, so I figured it HAD to be right....
Well, I'm here to tell ya - your wrong....or could be wrong...
There are issue's. Period. You need to understand the limitations of 1) cmm's, and 2) Pcdmis.
I love this software. Best I have used by far - and support them (pcdmis) to the end. But...it can be better. And this is where we come in. Spread the word, or talking to tech. support. "We", can make a diffrence...
Peace.
End of soapbox>>>>>>>>>>>>>>>

Kev

"And there was much rejoicing!"