PR 235967 Angle dimensions with planes can calculate measured value in the wr

huangyhg

Angle dimensions with planes can calculate measured value in the wrong direction.
<< Wade Burton  --  10/14/05  16:35:35>>
Doug Sjogren has been doing a lot of work lately with the angle dimensions.  We've discussed the steps to predict the measured angles of the angle between dimensions.
He has sent me a part program that shows that when the features are planes, the measured angle can be calculated in the wrong direction easily.   For example, when you have a plane at -1, 0.0005, 0, (x1) and find the angle with the xaxis (x2), it takes the plane normal, crosses it with the xaxis, and because it is a plane, it finds a line on the plane by crossing those together to get a vector mutually perpendicular (cross x1 and x2 to get x3), and then cross x3 back with x1 to find a line on the plane.  However, if the plane's normal is -1, -0.0005, 0, so its j value is negative, it is nearly the same plane, but because of the crossing, instead of getting near -90 degrees, it would give +90 degrees.  So the j direction has an affect on it.
<<END>>


<< Bret Naylor  --  01/24/06  11:47:48>>
Reviewed.  This seems really convoluted to me, and could probably be simplified.  However, at this point, it will have to wait.
<<END>>
<< Changes made by Bret Naylor -- 01/24/06  11:47:58>>
Action:  Bret Naylor to Wade Burton, Status:  REVIEW to CLOSED
<<END>>
<< Don Turcotte  --  01/20/06  11:05:30>>
Reviewed.
Seems complicated but works ok.
<<END>>
<< Changes made by Don Turcotte -- 01/20/06  11:05:30>>
Action:  Don Turcotte to Bret Naylor
<<END>>
<< Changes made by Bret Naylor -- 01/06/06  13:02:27>>
Action:  Bret Naylor to Don Turcotte
<<END>>
<< Wade Burton  --  12/21/05  16:47:18>>
Fixed V3.7 up.  Thanks, Don.
After determining the angle by using ACOS, it wasn't determining if that was positive or negative for the possible options, as the nominal was.
V37\DIMENS\DIM_2D_A.CPP
V37\DIMENS\DIM_3D_A.CPP
V40\DIMENS\DIM_2D_A.CPP
V40\DIMENS\DIM_3D_A.CPP
V41\DIMENS\DIM_2D_A.CPP
V41\DIMENS\DIM_3D_A.CPP
<<END>>
<< Changes made by Wade Burton -- 12/21/05  16:47:32>>
Action:  Wade Burton to Bret Naylor, Status:  OPEN to REVIEW
<<END>>
<< Don Turcotte  --  12/20/05  11:51:55>>
Reviewed.
I created a simple program with four planes.  I created angle dimensions versus X and Y for each plane taking the PC-DMIS calculated defaults for Nominal and Meas.  Several of these report nominal -45 and meas 45.  This is in program PlaneAnges.PRG attached.  I then executed the program and these meas values changed to -135.  This program is attached in PlaneAngles_measured.PRG.  These are both 4.0 programs using legacy dimensions.
<<END>>
<< Changes made by Don Turcotte -- 12/20/05  11:53:33>>
Action:  Don Turcotte to Wade Burton, Status:  REVIEW to OPEN
<<END>>
<< Bret Naylor  --  11/22/05  07:20:28>>
Don, please take the first look at this.
<<END>>
<< Changes made by Bret Naylor -- 11/22/05  07:20:34>>
Action:  Bret Naylor to Don Turcotte
<<END>>
<< Bill Wilcox  --  10/17/05  10:38:20>>
Please review this carefully.
<<END>>
<< Wade Burton  --  10/14/05  23:10:08>>
Fixed.
V37\DIMENS\DIM_2D_A.CPP
V37\DIMENS\DIM_3D_A.CPP
V40\DIMENS\DIM_2D_A.CPP
V40\DIMENS\DIM_3D_A.CPP
V41\DIMENS\DIM_2D_A.CPP
V41\DIMENS\DIM_3D_A.CPP
<<END>>
<< Changes made by Wade Burton -- 10/14/05  23:10:24>>
Action:  Wade Burton to Bret Naylor, Status:  OPEN to REVIEW, Priority:  to Critical
<<END>>
<< Wade Burton  --  10/14/05  16:49:55>>
I've been able to work with this to check the possible options, and take the one that makes the deviation smaller.  I'm not guessing the quadrant, as we had years ago, but just flipping it over and seeing if -95 is closer to the nominal than +85.
But I have to check the 3d angle also before it is completed.
<<END>>
<< Wade Burton  --  10/14/05  16:38:36>>
I've uploaded the file.
<<END>>