PR 240078 Dimension Angle Between no longer allows/support the TRUE/COMPLEMEN

huangyhg

Dimension Angle Between no longer allows/support the TRUE/COMPLEMENT toggle
<< Guillermo Nevado  --  04/25/06  10:30:07>>
In PCDMIS 3.7, The Edit Window (in Command/Summary Mode) allowed the toggle of the dimension angle calculation from True to Complement.  This, option however, was never available via the dialgo box (F9), but was usefull in forcing which of the two possible angles the user wanted.
Steps to reproduce:
Can only be duplicated by comparing the program in 3.7 to 4.0 because 4.0 does not even show the toggle as an option.
3.7 programs open up with what seems to be the right calculation, but I have not confirmed this.
In 4.0 there is absolutely no way to create the same angle dimension and choose the mode. Anylonger!
Results:
In 3.7 The (TRUE)/(COMPLEMENT) toggle is available (Edit Window Command Mode)
DIM ANGL1= 2D ANGLE (TRUE) FROM LINE LIN2 TO LINE LIN1 ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH
AX    MEAS      NOMINAL     +TOL       -TOL        DEV      OUTTOL   
A     45.0000    45.0000     0.2500     0.2500     0.0000     0.0000 ----#----
DIM ANGL2= 2D ANGLE (COMPLEMENT) FROM LINE LIN2 TO LINE LIN1 ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH
AX    MEAS      NOMINAL     +TOL       -TOL        DEV      OUTTOL   
A    135.0000   135.0000     0.2500     0.2500     0.0000     0.0000 ----#----
In 4.0 it is not.  This is a pasted copy of the 3.7 program
DIM ANGL1= 2D ANGLE FROM LINE LIN2 TO LINE LIN1 ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH
AX       MEAS    NOMINAL       +TOL       -TOL        DEV     OUTTOL
A     45.0000    45.0000     0.2500     0.2500     0.0000     0.0000 ----#----
DIM ANGL2= 2D ANGLE FROM LINE LIN2 TO LINE LIN1 ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH
AX       MEAS    NOMINAL       +TOL       -TOL        DEV     OUTTOL
A    135.0000   135.0000     0.2500     0.2500     0.0000     0.0000 ----#----
I have attached 3.7MR3 and 4.0 Release versions of same program.

Expected Results:
Ideally work same as it used to, and if changed, provide proper documentation for said change.  Even in 3.7 the toggle was not part of the dimension angle dialog box, so it really was a functionality difficult to locate unless the user was told to go to toggle from Edit Window.
<<END>>


<< Changes made by Neil Kay -- 02/25/09  15:24:04>>
Status:  RESOLVED to CLOSED
<<END>>
<< Jerry Naylor  --  08/13/08  10:55:25>>
wasaddedtoreadme42B
<<END>>
<< Changes made by Neil Kay -- 07/23/08  10:29:17>>
Action:  David Petrizze to Guillermo Nevado, Status:  REVIEW to RESOLVED
<<END>>
<< Don Turcotte  --  01/02/07  13:04:34>>
This seems to work correctly and properly takes into consideration the vectors of the features.
<<END>>
<< Changes made by Don Turcotte -- 01/02/07  13:04:46>>
Action:  Don Turcotte to David Petrizze, Status:  MOREINFO to REVIEW
<<END>>
<< Changes made by Tim Wernicke -- 12/06/06  10:26:53>>
Status:  PENDING to MOREINFO, Priority:  Stop Rel. to Critical
<<END>>
<< Guillermo Nevado  --  12/01/06  09:48:59>>
Don:
Because I don't really agree with the current method of choosing the angle to compute based on the nominal value, I would appreciate if you could verify Paola changes for the 90 degree case.  I like how it works, but still contend that feature nominal vector direction should be the overriding rule for which angle is true versus complement even if we choose to report one versus the other.  I worry about the case where a part has a flange, for example, that has to be at 90 degrees.  If it is over bent by say 1 degree, and the angle is on the spec print reference to the inside of the part, then 89 would be the appropriate answer.  The opposite case would be 91.
I am not sure that Paola's changes will assure that the reported value 89 versus 91 will be consistent with feature definition (vectors). I supplied a sample program where I faked a 1 degree  positive and negative deviation for the 0, 45, 90 and 180 degree cases.  I believe 90 will not work correctly every time in the sense that the output to the user will not necessarily tell which way the part is either over/under bent.
<<END>>
<< Changes made by Guillermo Nevado -- 12/01/06  09:49:18>>
Action:  David Petrizze to Don Turcotte, Status:  REVIEW to PENDING
<<END>>
<< Changes made by Don Turcotte -- 11/20/06  08:24:33>>
Action:  Don Turcotte to David Petrizze
<<END>>
<< Wagner Lee  --  11/17/06  14:20:14>>
ok, it works. thanks
<<END>>
<< Changes made by Wagner Lee -- 11/17/06  14:20:16>>
Action:  David Petrizze to Don Turcotte
<<END>>
<< Don Turcotte  --  11/17/06  11:21:20>>
As Paola explained below, in version V40+ there is no toggle for true/complement.  The meas calculation follows the nominal, so if you set the nominal to 135, the meas should be ~135.  If you set the nominal to 45, the meas should be ~45.
<<END>>
<< Changes made by Don Turcotte -- 11/17/06  11:21:56>>
Action:  Don Turcotte to David Petrizze
<<END>>
<< Wagner Lee  --  11/17/06  08:20:48>>
Tried in 4.1RE, and 4.2 Beta(10/07/06), and the toggle for true/complement is not visible.
<<END>>
<< Changes made by Wagner Lee -- 11/17/06  08:20:50>>
Action:  David Petrizze to Don Turcotte
<<END>>
<< Don Turcotte  --  05/22/06  09:19:03>>
Reviewed.
OK
<<END>>
<< Changes made by Don Turcotte -- 05/22/06  09:19:20>>
Action:  Bret Naylor to David Petrizze
<<END>>
<< Changes made by Don Turcotte -- 05/22/06  09:19:03>>
Action:  Don Turcotte to Bret Naylor
<<END>>
<< Paola Pallo  --  05/18/06  10:10:34>>
Done.
For review :
V41R\DIMENS\DIM_2D_A.CPP
V41R\DIMENS\DIM_3D_A.CPP
<<END>>
<< Changes made by Paola Pallo -- 05/18/06  10:10:44>>
Action:  Paola Pallo to Don Turcotte, Status:  OPEN to REVIEW
<<END>>
<< Tim Wernicke  --  05/17/06  11:33:27>>
Paola, please add your change to V41R.
<<END>>
<< Changes made by Tim Wernicke -- 05/17/06  11:33:31>>
Action:  Don Turcotte to Paola Pallo, Status:  REVIEW to OPEN, Priority:  Critical to Stop Rel.
<<END>>
<< Paola Pallo  --  05/17/06  17:35:24>>
Having a phone call with Guillermo.
Actual algorithm is correct, but as already written in my previous note, now the nominal value is extremely important and changing it you should get different results.
I verified a weakness in Angle Dialog management that could lead to confusion : clicking F9 on an existing Angle Dimension, even if you changed the nominal , it was ignored and you got always the first calculated MEAS value.
Now this is fixed. For review :
V41B\DIMENS\DIM_2D_A.CPP
V41B\DIMENS\DIM_3D_A.CPP
V42\DIMENS\DIM_2D_A.CPP
V42\DIMENS\DIM_3D_A.CPP
<<END>>
<< Changes made by Paola Pallo -- 05/17/06  17:35:32>>
Action:  Guillermo Nevado to Don Turcotte, Status:  RESOLVED to REVIEW
<<END>>
<< Paola Pallo  --  05/16/06  10:56:46>>
Guillermo, please give me a call to clarify this.
<<END>>
<< Changes made by Paola Pallo -- 05/16/06  10:56:57>>
Action:  Paola Pallo to Guillermo Nevado
<<END>>
<< Guillermo Nevado  --  05/16/06  00:55:47>>
I am refering to the angle_test program not the original one in question.
<<END>>
<< Changes made by Guillermo Nevado -- 05/16/06  00:55:58>>
Action:  Guillermo Nevado to Paola Pallo
<<END>>
<< Guillermo Nevado  --  05/16/06  00:54:34>>
I have double checked it.  Please check the attached program.  It shows problems with angle dimensions in 4.0  I wrote a program (read comments from me dated 4/27 and explaining the program).
The program was written in 4.0
<<END>>
<< Paola Pallo  --  05/15/06  16:06:00>>
We confirm that v40 design is not supporting anymore TRUE/COMPLEMENT option, since the internal algorithm was improved  to calculate the four possible angles, and to return the closest one according to the specified feature's directions and to the nominal  angle value . This last condition is mainly important, so please double check it.
<<END>>
<< Changes made by Paola Pallo -- 05/15/06  16:06:21>>
Action:  Paola Pallo to Guillermo Nevado, Status:  OPEN to RESOLVED
<<END>>
<< Paola Pallo  --  05/11/06  18:04:03>>
Contacted Tim and Wade via e-mail
<<END>>
<< Changes made by Tim Wernicke -- 05/05/06  13:57:09>>
Priority:  High to Critical
<<END>>
<< Changes made by Robert Fischer -- 05/01/06  13:55:20>>
Action:  Guillermo Nevado to Paola Pallo
<<END>>
<< Robert Fischer  --  05/01/06  13:54:54>>
Fulton Precision just called in with the same problem - Angles are now trashed...wanted to know if it was a bug.....Very happy to hear that he could SAVE AS and go back to an earlier version.
<<END>>
<< Changes made by Robert Fischer -- 05/01/06  13:54:55>>
Action:  Paola Pallo to Guillermo Nevado
<<END>>
<< Changes made by Tim Wernicke -- 04/28/06  09:05:17>>
Action:  Tim Wernicke to Paola Pallo, Status:  RESOLVED to OPEN
<<END>>
<< Guillermo Nevado  --  04/27/06  23:55:44>>
Read from previous comment from me up.
I just spent time writing a PCDMIS 4.0 program (note no test on migration) that calculates angles between planes 0/180 apart, 90 apart and 45 apart, and two pair of lines 45/135 apart.
I faked out a +1 and a -1 deg angle as measured while still mantain true nominal (theo).   Please review the PDF report attachment and sample program.  I found issues with evaluating w/ nominal 180 on the nearly parallel +1/-1 planes.  Note that it was alwasy 179 regardless.
Found problems with the 90 degree planes in that although calculated right depending on feature selection order, I could not tell which angle I really was getting.  I mean, I could tell because I knew what I forced, but I could not tell what rule gave me 89 versus 91.
The angle between 2D lines w/ nominal 135 gave only the deviation from 45 value.  Under TRUE/COMPLEMENT these calculated correctly before.
<<END>>
<< Guillermo Nevado  --  04/27/06  21:53:01>>
Tim:
I am not sure I agree with this not being an issue.  I reviewed both PRs  238599 and the one it references 235967 which Bret Naylor (and I would agree) ends with (and I quote):
<< Tim Wernicke  --  04/26/06  22:31:20>>
Guillermo, the True/Complement addition was really a workaround to the real problem.  It should be working correctly now.
<<END>>
<< Changes made by Tim Wernicke -- 04/26/06  22:30:32>>
Action:  Tim Wernicke to Guillermo Nevado, Status:  OPEN to RESOLVED
<<END>>
<< Paola Pallo  --  04/26/06  10:07:00>>
In 238599 you can find a detailed description from Wade.
<<END>>
<< Changes made by Paola Pallo -- 04/26/06  10:07:18>>
Action:  Paola Pallo to Tim Wernicke
<<END>>
<< Tim Wernicke  --  04/25/06  13:47:22>>
Paola, do you remember which PR it was that this was changed for?
<<END>>
<< Changes made by Tim Wernicke -- 04/25/06  13:47:23>>
Action:  Tim Wernicke to Paola Pallo, Assigned:   to Paola Pallo, Priority:  to High
<<END>>