compound angles

[terry]

I have a question on how PC-DMIS handles rotations on compound angle faces. I have a part that has a face that has a compound angle of 14.4 deg and 2.1 deg. When I translate to the face and rotate -14.4 about x minus and then rotate 2.1 about z plus. Take four vectored points on the surface, construct a plane, recall my ABC and look at the angles of the face. The result I am getting confuses me because the second rotation nominal is reported as 2.168 not 2.1. So just for giggles I repeat the process but this time I rotate the 2.1 first and then the 14.4 and sure enough the 2.1 nominal stays but the 14.4 is now 14.409.
I have talked to tech support and sent a copy of my program but they have not been able to give me an answer that makes sense.
First example
TRANSLATE_AND_ROTATE=ALIGNMENT/START,RECALL:abc_dcc, LIST= YES
ALIGNMENT/TRANS_OFFSET,YAXIS,-113.9
ALIGNMENT/TRANS_OFFSET,XAXIS,35.8
ALIGNMENT/TRANS_OFFSET,ZAXIS,-142.6
ALIGNMENT/ROTATE_OFFSET,-14.4,ABOUT,XMINUS
ALIGNMENT/ROTATE_OFFSET,2.1,ABOUT,ZPLUS
ALIGNMENT/END
First Result.
RECALL/ALIGNMENT,INTERNAL,abc_dcc
WORKPLANE/XPLUS
DIM ANGL1= 2D ANGLE (TRUE) FROM PLANE D_FACE TO ZAXIS ,$
GRAPH=OFF TEXT=OFF MULT=10.00 OUTPUT=BOTH
AX NOMINAL +TOL -TOL MEAS DEV OUTTOL
A 14.400 0.010 0.010 14.979 0.579 0.569 -------->
WORKPLANE/ZPLUS
DIM ANGL3= 2D ANGLE (TRUE) FROM PLANE D_FACE TO XAXIS ,$
GRAPH=OFF TEXT=OFF MULT=10.00 OUTPUT=BOTH
AX NOMINAL +TOL -TOL MEAS DEV OUTTOL
A -2.168 0.010 0.010 -2.016 0.152 0.142 -------->

Second example
TRANSLATE_AND_ROTATE2=ALIGNMENT/START,RECALL:abc_dcc, LIST= YES
ALIGNMENT/TRANS_OFFSET,YAXIS,-113.9
ALIGNMENT/TRANS_OFFSET,XAXIS,35.8
ALIGNMENT/TRANS_OFFSET,ZAXIS,-142.6
ALIGNMENT/ROTATE_OFFSET,2.1,ABOUT,ZPLUS
ALIGNMENT/ROTATE_OFFSET,-14.4,ABOUT,XMINUS
ALIGNMENT/END
Second Result
DIM ANGL4= 2D ANGLE (TRUE) FROM PLANE D_FACE_2 TO ZAXIS ,$
GRAPH=OFF TEXT=OFF MULT=10.00 OUTPUT=BOTH
AX NOMINAL +TOL -TOL MEAS DEV OUTTOL
A 14.409 0.010 0.010 14.980 0.571 0.561 -------->
WORKPLANE/ZPLUS
DIM ANGL5= 2D ANGLE (TRUE) FROM PLANE D_FACE_2 TO XAXIS ,$
GRAPH=OFF TEXT=OFF MULT=10.00 OUTPUT=BOTH
AX NOMINAL +TOL -TOL MEAS DEV OUTTOL
A -2.100 0.010 0.010 -2.018 0.082 0.072 -------->
WORKPLANE/YPLUS

Does any one know why the nominal values change on the second rotation?

Thanks
Terry

[Mike N.]

its a COMPOUND angle and you are trying to view it as a 2D angle. You will have to rotate to the first angle to report the second separately.

[Matthew D. Hoedeman]

Well, the second one changes because of what you have done with the first angle. Mike is right, you are looking at it in a 2-D space when it is a 3-D angle. It just doesn't work out right. Pcdmis is giving you the correct value for what you have done. Now, the best way to get out of this is to do the dreaded iterative alignment. As long as you have the nominal XYZIJK for enough point to align the part, this would be the way to go. Then, no matter WHAT the 3-D angle is of the surface, it will align it correctly and you can go from there. I have struggled with this back in the DOS days, before the ITERATIVE alignment was available in Pcdmis and it takes a cad system to create the surface on and to calculate the actual angles needed to make the rotations come out right.

[JIM HARRIS]

There is a differance between a compound angle and a angle rotated out of another.What you are trying to ask it for is a compound angle.you would need to ask what the first angle is then rotate that angle then ask what the second angle is.Compound angles are rarely used.If you use a sine plate to check a compound angle there is a forumla used to set your sine plates at.

[terry]

I figured it was because it of two rotations but I figured the first rotation would have been the one effected because the second rotation would have moved it about that axis.
I wasn’t really concerned about what the two angles measured just checking what was happening was what I thought was happening. The main problem that answer leaves me with is I do not have a cad model for this part and I need to check the profile of this face and the location of the holes in it. Am I getting an accurate check for profile on a compound angled surface by making two rotations to get the nominal vector of the face?

Regards,
Terry

[RussL]

FYI
If you do the trig:
2.167 = ArcTan((Sin 2.1)/(Cos 14.4))
14.410 = ArcSin((Sin 14.4)/(Cos 2.1))

[Matthew D. Hoedeman]

Dude, you do too much math, man.

OH C R A P!, I wasted my 1000th post on something way too mundane!

[Mike N.]

you should rotate per the main view first then rotate to the section view second. At those nominal angles you will be able to measure a plane or points and dimension the profile of the face and measure the holes (assuming they are dimensioned from the point of rotation)

[bw_bob]

Dude, you do too much math, man.

OH C R A P!, I wasted my 1000th post on something way too mundane!

CONGRATS on #1000

[Shelley Liebman]

How tight is the tolerance of the profile, and positions of the holes in the surface? Also, when you say that the 14.4 angle is 14.409, are you talking degrees decimal or degrees minutes seconds?
At degrees decimal, that would be pretty small unless you are talking tight tolerances.......
Another question, you say that you have done the rotations to the plane, have you leveled to that plane? If you leveled to that plane, wouldn't it negate out the errors in rotation when dimensioning profile? I ask cause I honestly don't know, but I am faced similar problems and would love to know the answer myself.......and I am just brainstorming here trying to help.....

[terry]

Decimal degrees. The point was not that nominal was only .009 off but I did not understand why the nominal was off at all. Now that I see RUSSL “rain man” trig formula I understand why I get the answer I got. I have not ever double checked the angle on a compound rotation to seen this before. I wouldn’t have double checked it at all if the surface hadn’t been off so far. No, I did not level to the plane or you are right it would negate the ability to check the individual points for profile.
Regards
Terry