PR 226993 Strange results changing the reference lenght on angularity

huangyhg

Strange results changing the reference lenght on angularity
<< Alessandro Zanetti  --  09/14/04  15:23:09>>

Steps to reproduce:
Open the attached prog.
Results:
Look at the angularity results when reference lenght changes.
Expected Results:
Angularity should be about the same, as happens with parallelism and perpendicularity.
<<END>>

<< Changes made by Yanhua Huang (Field Changes) -- 05/19/09  16:20:07>>
Action:  Wade Burton to Yanhua Huang
<<END>>
<< Changes made by Neil Kay -- 07/23/08  09:57:30>>
Action:  David Petrizze to Wade Burton
<<END>>
<< Changes made by Don Turcotte -- 10/18/06  11:19:28>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>
<< Don Turcotte  --  12/22/04  11:52:44>>
I have modified the legacy angularity dialog to default the theo reference angle as the user picks features.
******************************
Wed Dec 22 11:51:44 2004
******************************
Files inserted to server
------------------------
V40\MENU\DIMANGUL.CPP
<<END>>
<< Bret Naylor  --  12/17/04  08:12:25>>
From the meeting today.  In 4.0, we will automatically calculate this nominal when the user selects two features - both in the legacy dimension and the FCF.
<<END>>
<< Changes made by Bret Naylor -- 12/17/04  08:12:51>>
Action:  David Petrizze to Don Turcotte, Status:  REVIEW to OPEN
<<END>>
<< Bret Naylor  --  12/15/04  12:14:03>>
Reviewed.  Dave,  let's discuss this in the next GD&T meeting.
<<END>>
<< Changes made by Bret Naylor -- 12/15/04  12:14:29>>
Action:  Bret Naylor to David Petrizze
<<END>>
<< Wade Burton  --  12/09/04  12:36:06>>
Reviewed.  What Don has done here looks fine to me.
As for how this should be working with the FCF's, I have attached a .jpg file here that shows how the standard expects the Angularity FCF to look, and the nominal angle is not part of the FCF, but a separate label.  I'm not sure how to do this in our framework, but instead maybe allow the user to put it on the first line of the FCF.  I think we should probably discuss this in our GD&T meeting soon, and I suspect that will be a separate task from this.
<<END>>
<< Changes made by Wade Burton -- 12/09/04  12:36:18>>
Action:  Wade Burton to Bret Naylor
<<END>>
<< Don Turcotte  --  12/09/04  10:51:08>>
Implemented error message for legacy angularity dimension.  This issue needs to be addressed in the Feature control frame when creating angularity dimensions.  The FCF should default the Reference (Nominal) angle.  (With the current build, I cannot create Datum Definition commands to check this.)
******************************
Thu Dec 09 10:48:22 2004
******************************
Files inserted to server
------------------------
V40\INCLUDE\DIMANGUL.H
V40\INCLUDE\resource.h
V40\MENU\DIMANGUL.CPP
V40\RESOURCE\LANGS\ENGLISH\RESOURCE.RC
<<END>>
<< Changes made by Don Turcotte -- 12/09/04  10:51:48>>
Action:  Don Turcotte to Wade Burton, Status:  OPEN to REVIEW
<<END>>
<< Don Turcotte  --  12/08/04  17:45:29>>
Began work on this to add warning message and do FindTheoAngle...
<<END>>
<< David Petrizze  --  10/07/04  14:31:57>>
Don, go ahead and make this enhancement in the V40 project.
<<END>>
<< Changes made by David Petrizze -- 10/07/04  14:32:20>>
Status:  MOREINFO to OPEN
<<END>>
<< Changes made by David Petrizze -- 10/07/04  14:32:11>>
Action:  David Petrizze to Don Turcotte
<<END>>
<< Changes made by Bret Naylor -- 10/04/04  10:55:14>>
Status:  OPEN to MOREINFO
<<END>>
<< Don Turcotte  --  09/30/04  09:17:22>>
Dave,
Do we want to make this change to V3.7 or should this be changed to an enhancement?
<<END>>
<< Changes made by Don Turcotte -- 09/30/04  09:17:35>>
Action:  Don Turcotte to David Petrizze
<<END>>
<< Alessandro Zanetti  --  09/30/04  09:40:37>>
Don, i agree totally about the fact that leaving the default instead of specifying a theo value or using the Find Theo Angle is a user error.
What i'm saying is that, in case of user's errors, we should prompt the user with an error message as, for example, " a theoretical Angle of 0 deg. means a parallelism " or " please use the Find Theo Angle option ", or better switching automatically to this option showing " PC-DMIS will now switch automatically to the Find Theo Angle option because of you selected a theoretical Angle of 0 deg. that means a parallelism ".
But the user must not receive strange results without an explanation, otherwise he will call for support identifying this behavior as a bug.
Alessandro
<<END>>
<< Changes made by Alessandro Zanetti -- 09/30/04  09:40:53>>
Action:  Alessandro Zanetti to Don Turcotte
<<END>>
<< Changes made by Don Turcotte -- 09/29/04  11:55:02>>
Action:  Don Turcotte to Alessandro Zanetti
<<END>>
<< Don Turcotte  --  09/29/04  11:53:44>>
Alessandro,
When the user creates an angularity dimension, the user must set the Reference Angle to a value that makes sense for the features selected.  If the user is not sure what the THEO angle should be, the user can select Find Theo Angle to help.
<<END>>
<< Alessandro Zanetti  --  09/16/04  12:04:52>>
Ciao Don, you're right. I have not well described where the problem is.
If you look at the last three angularities i inserted in the prog, i mean ang3,ang4,ang5, you will see that the theoretical reference angle is not specified and so is 0 since this is the default in the dialog.
The problem is that, when the user does not specify a theoretical reference angle and does not use the Find button, PC-DMIS calculates the angularity giving strange results ( that are changing if the reference lenght changes ) instead of giving the user an error message showing the error ( a theoretical angle set to 0 means basically a parallelism, and however it is clearly an error ).
Hope this helps.
<<END>>
<< Changes made by Alessandro Zanetti -- 09/16/04  12:05:11>>
Action:  Alessandro Zanetti to Don Turcotte, Status:  MOREINFO to OPEN
<<END>>
<< Don Turcotte  --  09/15/04  16:35:16>>
I'm not clear on what the problem is here.  Since the THEO and MEAS values for LIN2 and LIN4 are the same, when you set the Reference Angle to the calculated nominal of 134.867 then changing the reference length will have no effect on the deviation.  If you set the nominal angle to 135.0  which is probably what the drawing (or cad model) calls for, then increasing the reference length will increase the deviation.  This is the expected behavior.
Executing this program online with the actual part should also demonstate the same behavior.
<<END>>
<< Changes made by Don Turcotte -- 09/15/04  16:35:42>>
Action:  Don Turcotte to Alessandro Zanetti, Status:  OPEN to MOREINFO
<<END>>
<< Changes made by Tim Wernicke -- 09/15/04  10:51:18>>
Priority:  to Medium
<<END>>