PR 259768 Can not calculate 2D Angle in PCDMIS 4.2 MR1,2

huangyhg

Can not calculate 2D Angle in PCDMIS 4.2 MR1,2
<< Paul Jo  --  02/16/09  00:07:24>>
Result of 2D angle is wrong!
Please see the attached filrs.
Steps to reproduce:
1. Measured two plane.(XZ Plane).
2. Change the workplane Y-.
3. Dimension of the 2D angle.

Results:
1. Result is wrong.
    nominal is 89 degrees but result angle is 91 degrees.
2. Sometimes is OK But sometimes is wrong.(Random).
<<END>>


<< Changes made by Don Turcotte (Field Changes) -- 04/15/09  08:57:54>>
Action:  Don Turcotte to Don Ruggieri, Status:  OPEN to REVIEW
<<END>>
<< Changes made by Don Turcotte (Resolution Data) -- 04/15/09  08:57:42>>
** Bug Fix - New In Current Version -  -  -  **
<<END>>
<< Don Turcotte (Change Sets)  --  04/14/09  14:30:10>>
Files inserted to server
------------------------
V432\DIMENS\DIM_2D_A.CPP
V44B\DIMENS\DIM_2D_A.CPP
<<END>>
<< Changes made by Don Turcotte (Field Changes) -- 04/14/09  14:29:59>>
Action:  Paul Jo to Don Turcotte
<<END>>
<< Don Turcotte (Development Notes)  --  04/14/09  14:29:59>>
Further testing with planes at basic 45 degree angle showed that  previous fixes while correct did not handle all four possible quandrants (45, 135, 225, 315).  I have fixed this in V432 and V44B.
Files inserted to server
------------------------
V432\DIMENS\DIM_2D_A.CPP
V44B\DIMENS\DIM_2D_A.CPP
<<END>>
<< Don Turcotte (Development Notes)  --  04/14/09  10:45:08>>
Conference call with Mike, Ken, Bret.  Ken thinks we should improve angle dialog to update when user changes nominal angle.  Mike mentioned that best user interface would present user with a circle with 4 quadrants and user picks the desired quadrant.  Mike will test fixes to verify they are what customer is asking for.  Fixes have been made only in V432, V44B.  (V43 MR1 is unchanged).
<<END>>
<< Changes made by Sung-jin Cho (Field Changes) -- 04/01/09  12:01:02>>
Status:  RESOLVED to OPEN
<<END>>
<< Don Turcotte  --  03/19/09  13:19:08>>
Based on Don R.'s testing, this is now working as the customer expects.  Fixed in V432 (V43 MR2) and V44B.
<<END>>
<< Changes made by Don Turcotte -- 03/19/09  13:19:30>>
Action:  Don Turcotte to Paul Jo, Status:  REVIEW to RESOLVED
<<END>>
<< Don Ruggieri  --  03/18/09  13:13:58>>
I tried this with last night's v44b.  As far as I can tell so far, it addresses this problem report very wel, and also the requirements of #257415.  It seems to adjust the nominal properly when I use the 0-180 or 0-360 confiduration.  I may try to run a few moer tests today, but this does look good so far.
<<END>>
<< Changes made by Don Ruggieri -- 03/18/09  13:14:00>>
Action:  Don Ruggieri to Don Turcotte
<<END>>
<< Don Turcotte  --  03/18/09  08:36:38>>
Don, could you test this again in light of the new changes?
<<END>>
<< Changes made by Don Turcotte -- 03/18/09  08:36:41>>
Action:  Don Turcotte to Don Ruggieri, Status:  OPEN to REVIEW
<<END>>
<< Don Turcotte  --  03/17/09  17:15:59>>
I have reopened this based on the latest comments from Paul Jo who tested the most recent fixes.
"Dear Donald Turcotte.

I’m sorry being late.

I had tested 2D angle of dimension in PCDMIS4.3_MR2_RC.

But it doesn’t work correctly. (Please see the attached files).

If the nominal is not 90 degree (for example 89.5 or 90.5), the result displays the angle vales closer to the nominal value.

If I can select the vector or set the quadrant about angle of dimension in PCDMIS that this problem will be solved.

My customer used to another ‘Mitutoyo Machine’. He said ‘Mitutoyo program can set the vector in angle of dimension’

‘So result is always same quadrant’. ‘If result value is bigger or smaller than 90 degrees that result value is not closer to the nominal value’.

Thank you. And I’ll wait for your reply.

Best Regards,

Paul.
"
This is consistent with the comments from Don R. below.  This conflicts, however, with the expectation in #257415.  
I have implemented a new algorithm which satisfies the requirements of Paul Jo and Don R. while still meeting the expectations of #257415.  I have also tested against #256828.
I will ask Don R. to retest tommorrow.
Relevant PR's
  256828
  257415
  249541
Latest fix uploaded to V432, V44B.
Files inserted to server
------------------------
V432\DIMENS\DIM_2D_A.CPP
V44B\DIMENS\DIM_2D_A.CPP
<<END>>
<< Changes made by Don Turcotte -- 03/17/09  17:15:59>>
Status:  REVIEW to OPEN
<<END>>
<< Don Ruggieri  --  03/05/09  10:31:33>>
The error that we understand to be identified by the customer was that the same angle calculation on different sides of the 180 degree line was not complementary.  The algorithm simply negated the value of the measured angle.  This is not good.  Last night's modifications cause the measured angle to change to its complement when the negative nominal (90 vs -90)
The question I have though is this.  When the measured value is +89.5 and the nominal is 89, the measured value is interpreted to be short of 90 degrees, and that may be acceptable.
However, with the same angle if you change the nominal to +91, the measured value changes to 91.5, effectively representing a different angle.  This could mean that the angle is being viewed in a different quadrant, or it could mean that the angle has changed erroneously, I am not sure which is better.
What I do know is that this is new (from last night") as it works this way in the most recent v432 but not in an earlier v432 nor in V431.
<<END>>
<< Changes made by Don Ruggieri -- 03/05/09  10:31:43>>
Action:  Don Ruggieri to Don Turcotte
<<END>>
<< Don Turcotte  --  03/04/09  11:50:19>>
Based on the latest info from customer (I have attached their latest program as "test03042009.zip"), I have identified a problem in 2D angle.  Does not find the correct complementary angle.   I have fixed this in DIM_2D_A.CPP.
I also found a problem in 3D angle when reporting in 0-360 mode.  Did not report correct angle.  I have fixed this in DIM_3D_A.CPP.
This is now working correctly using the customer's program.
Don R., please retest this.
Fixed in V432, V44B.
Files inserted to server
------------------------
V432\DIMENS\DIM_2D_A.CPP
V432\DIMENS\DIM_3D_A.CPP
V44B\DIMENS\DIM_2D_A.CPP
V44B\DIMENS\DIM_3D_A.CPP
<<END>>
<< Changes made by Don Turcotte -- 03/04/09  11:52:10>>
Action:  Don Turcotte to Don Ruggieri, Status:  OPEN to REVIEW
<<END>>
<< Changes made by Don Turcotte -- 03/04/09  11:50:19>>
Action:  Paul Jo to Don Turcotte, Status:  RESOLVED to OPEN
<<END>>
<< David Petrizze  --  02/23/09  17:31:08>>
Reviewed.
<<END>>
<< Changes made by David Petrizze -- 02/23/09  17:31:13>>
Action:  David Petrizze to Paul Jo, Status:  REVIEW to RESOLVED
<<END>>
<< Don Turcotte  --  02/24/09  14:03:56>>
Merged into V432.
Files inserted to server
------------------------
V432\DIMENS\DIM_2D_A.CPP
<<END>>
<< Changes made by Don Turcotte -- 02/24/09  14:04:13>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>
<< Changes made by Tim Wernicke -- 02/24/09  11:26:48>>
Action:  David Petrizze to Don Turcotte, Status:  REVIEW to OPEN, Priority:  Critical to Stop Rel.
<<END>>
<< Don Turcotte  --  02/17/09  16:01:33>>
This is related to #257415.  The epsilon to determine which quadrant the nominal is in was changed to 5.0 degrees for #257415.  This is too large for the program attached to this PR.  I have changed the code to try an epsilon of 1.0 degree.  If no match is found, then the quadrant is searched for using an epsilon of 5.0 degrees.  This satisfies both the program attached to this PR and #257415.
If the user creates the 2D angle dimension and enters the nominal based on the desired quadrant, then the measured value will always be reported in that quadrant.  I have tested this by running PC-DMIS "online" with the simulator and an error factor to generate perturbed data.  The measured value in all cases was reported in the same quadrant as the nominal value.
Files inserted to server
------------------------
V44B\DIMENS\DIM_2D_A.CPP
<<END>>
<< Changes made by Don Turcotte -- 02/17/09  16:02:41>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>
<< Changes made by Tim Wernicke -- 02/16/09  10:07:05>>
Action:  Paul Jo to Don Turcotte, Assigned:   to Don Turcotte, Priority:  to Critical
<<END>>
<< Changes made by Paul Jo -- 02/16/09  00:26:11>>
Action:  Tim Wernicke to Paul Jo
<<END>>