How to make alignment

[DAN_M]

Solved-->

Issue is solved. See comment#15. Thanks for all of your help!

__________________________________________________ __________________________________________________ __________________________________
Original Post-->

PC DMIS 2019 R2 CAD...

I am trying to settle a work "discussion". Can the hive mind please tell me how you would go about setting up the alignment(s) for "ARS"? (Datum -A- is the interrupted surface that is perpendicular to all of the through holes. Everything on part is accessible with a 3X40 probe in A0B0).


Thanks in advance,
Dan

pic for group.PNG

[Matthew D. Hoedeman]

Iterative alignment. Measure a bunch of points on "A", measure a hole (with 3 sample hits) for each of R and S. Bingo-bango-bongo.

[DAN_M]

Matthew D. Hoedeman

Thanks for the input.

Can we take this a bit further?

I can def set up an iterative...Where do you think my XY origin should actually be to report the basic dimensions back to?

[Darrollh]

It looks like Datum R is actually your origin.

[DAN_M]

Darrollh Thats what I am thinking.

I believe I should be Leveled to -A-, rotated iteratively R to S, R=XY origin and A=Z origin. Would you agree?

[Darrollh]

That’s how I would do it.

[Matthew D. Hoedeman]

yep, that's the callout. I'm guessing that 'someone' is trying to get you to rotate about by the correct (theo) angle of the 2 holes. NOPE, I would not do it that way. The only way I would is if the S-Hole were a slot with the axis of the slot the same as the theo angle.

[Darrollh]

I totally agree. I had a print similar to this with a similar call out. When I did what the print asked me to do, the basics were skewed.

[JacobCheverie]

I'm sure you've got it, but I wouldn't bother with an iterative alignment and maybe you can enlighten me to the benefits in this case? A will lock two rotational DOF and one translational. Thus, level let's say Z to A and translate Z to A. Now secondary datum R will lock the remaining two translational DOF so translate X and Y to R. Now all that is left is a trihedron with XY coplanar to A and centered at R that is rotating about the axis of R. Use tertiary clocking datum to rotate let's say X+ to the line formed from R and S.

Why would iterative be better here? Does it matter? This is a good learning experience.

I know your results won't be comparable to your basics (without trig) but the positional result should be the same.

[Matthew D. Hoedeman]

I'm sure you've got it, but I wouldn't bother with an iterative alignment and maybe you can enlighten me to the benefits in this case? A will lock two rotational DOF and one translational. Thus, level let's say Z to A and translate Z to A. Now secondary datum R will lock the remaining two translational DOF so translate X and Y to R. Now all that is left is a trihedron with XY coplanar to A and centered at R that is rotating about the axis of R. Use tertiary clocking datum to rotate let's say X+ to the line formed from R and S.

Why would iterative be better here? Does it matter? This is a good learning experience.

There is a theo angle those 2 holes can be to each other and an actual. The 'standard' 3-2-1 alignment will use that theo angle for a 'rotate-about'. BUT, in the case of 2 holes, it is generally the Y axis (or X axis) offset that needs to be maintained. Easiest way to do that is the iterative alignment. You CAN do a 3-2-1 and still use the single axis offset, but iterative is easier with less code. Also, iterative alignments can be set to re-measure all the alignment features, thus giving a tighter alignment without more code. Your method make the X axis run along the line of the 2 holes, if his X has to run left-to-right (as you look at the picture), you would be rotated out of position.

[JacobCheverie]

In my opinion, a large part of confusion with interpretation stems from the lack of a defined coordinate system on the drawing. Anyway, thanks for clearing that up. I don't use iterative alignments but then again I typically have more primitive parts.

[AndersI]

Creating R - S as an offset line would be my first choice, but I don't see a boxed distance R | S?

[NinjaBadger]

It's unusual, but I disagree with Matt and AndersI for once.

Both offset line and iterative methods do the same thing really, they maintain either (in this case) the x or the y relationship between the two holes, but if the pitch of the holes isn't perfect you get an error.


Level to plane.
Origin on R
Rotate to R-S
Rotate back by the theoretical angle (trig it out) so you remain aligned to the print.


Edit - see below with a nominal 30,40,50 triangle (internal angles of 90°, 36.87°, 53.13°)

PLN1       =FEAT/CONTACT/PLANE/DEFAULT,CARTESIAN,NONE,LEAST_SQR
            THEO/<0,0,0>,<0,0,1>
            ACTL/<0,0,0>,<0,0,1>
            TARG/<0,0,0>,<0,0,1>
            ANGLE VEC=<1,0,0>,RADIAL
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=YES
              NUMHITS=3,NUMRINGS=1
              SPACER=9
              AVOIDANCE MOVE=BOTH,DISTANCE=10
              USE BOUNDARY OFFSET=NO
            SHOW HITS=NO
CIR1       =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
            THEO/<0,0,0>,<0,0,1>,30
            ACTL/<0,0,0>,<0,0,1>,30
            TARG/<0,0,0>,<0,0,1>
            START ANG=90,END ANG=270
            ANGLE VEC=<1,0,0>
            DIRECTION=CCW
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=YES
              NUMHITS=7,DEPTH=5,PITCH=0
              SAMPLE METHOD=SAMPLE_HITS
              SAMPLE HITS=3,SPACER=0
              AVOIDANCE MOVE=BOTH,DISTANCE=10
              FIND HOLE=DISABLED,ONERROR=NO,READ POS=NO
            SHOW HITS=NO
CIR2       =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
            THEO/<40,30,0>,<0,0,1>,30
            ACTL/<40,35,0>,<0,0,1>,30
            TARG/<40,30,0>,<0,0,1>
            START ANG=90,END ANG=270
            ANGLE VEC=<1,0,0>
            DIRECTION=CCW
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=YES
              NUMHITS=7,DEPTH=5,PITCH=0
              SAMPLE METHOD=SAMPLE_HITS
              SAMPLE HITS=3,SPACER=0
              AVOIDANCE MOVE=BOTH,DISTANCE=10
              FIND HOLE=DISABLED,ONERROR=NO,READ POS=NO
            SHOW HITS=NO
LIN1       =FEAT/LINE,CARTESIAN,UNBOUNDED,NO
            THEO/<0,0,0>,<1,0,0>
            ACTL/<0,0,0>,<0.9929178,0.1188031,0>
            CONSTR/LINE,OFFSET
            SURFACE NORMAL = <0,0,1>,MULTI POINT
            ID = CIR1,CIR2,,
            OFFSET = 0,-30
A1         =ALIGNMENT/START,RECALL:STARTUP,LIST=YES
              ALIGNMENT/LEVEL,ZPLUS,PLN1
              ALIGNMENT/TRANS,ZAXIS,PLN1
              ALIGNMENT/TRANS,XAXIS,CIR1
              ALIGNMENT/TRANS,YAXIS,CIR1
              ALIGNMENT/ROTATE,XPLUS,TO,LIN1,ABOUT,ZPLUS
            ALIGNMENT/END
DIM LOC1= LOCATION OF CIRCLE CIR2  UNITS=MM ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH  HALF ANGLE=NO
AX    NOMINAL       +TOL       -TOL       MEAS        DEV     OUTTOL
PA      36.870      0.001      0.001     34.363     -2.507      2.506 <--------
END OF DIMENSION LOC1

[AndersI]

I'm easily convinced - never thought of doing it that way, but it makes sense!

[DAN_M]

NinjaBadger this is what I wanted to do originally and ended up doing. My code is on post15. Thanks for the help!

[NinjaBadger]

Or if your trig isn't very good, do it as Matt or Anders suggested, then do a best-fit rotate only to the tertiary datum.

[Matthew D. Hoedeman]

3-2-1 + theo rotate value will put error into both axis of the "S" hole if the spacing isn't perfect, offset line & iterative will put all error into a single axis of the "S" hole, BUT, iterative will let you measure the alignment multiple times, refining each time through, without adding more alignments or code to the program.

Now, if the print had an angle called out from R to S, then I might (might) go with the 3-2-1 + theo angle rotate, but no angle, offset line or iterative (iterative).

[NinjaBadger]

But doing an iterative you will end up with a deviation (Polar angle) for the tertiary datum - which simply isn't right.

[Matthew D. Hoedeman]

Iterative gives the same devation(s) as an offset line will. It will depend on 'intent'. No angle called out tells me that they want ONE axis only to show devation for the tertiary, as in CAN,MAY,MUST or whatever the saying is.

"A" can control 2 axis of rotation and 1 axis of origin, and it may, so it must.
"R" can control 2 axis of origin, and it may, so it must, it can not control any axis of rotation unless it is a cylinder, but that axis has already been taken)
"S" can control 2 axis of origin, but it may not as they are already taken care of, "S" can control 1 axis of rotation, and it may, so it must.
(re ONE axis of rotation, not split between 2 axis).
As in, it can control either the "X" axis or the "Y" axis (if the level is Z).
As for how you choose to interpret that ONE axis of rotation.....
No basic on the print for either angle or axis offset, so, which IS the way to do it? If they had an angle on there, yeah, rotate to the perfect angle, put error in both axis of "S", if they had the offset value for one of the axis, then offset line.

[NinjaBadger]

We'll have to agree to disagree Matt, it's not rotationaly controlling one axis (x or y) to the feature, it's controlling the rotation AROUND the z axis, much like the plane is controlling the rotation AROUND the x and y axis.

Anyways... I've said my piece.

[DAN_M]

NinjaBadger Matthew D. Hoedeman Darrollh AndersI JacobCheverie
I spoke with the designers & the intent is to control the part R THRU S. That being said, this is what I came up with and will be using:

A5         =ALIGNMENT/START,RECALL:STARTUP,LIST=YES
              ALIGNMENT/LEVEL,ZPLUS,PLN_A
              ALIGNMENT/ROTATE,XPLUS,TO,LIN_RS,ABOUT,ZPLUS
              ALIGNMENT/ROTATE_OFFSET,-28.8614,ABOUT,ZPLUS
              ALIGNMENT/TRANS,XAXIS,CIR_R
              ALIGNMENT/TRANS,YAXIS,CIR_R
              ALIGNMENT/TRANS,ZAXIS,PLN_A
            ALIGNMENT/END
CIR_9_1    =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
            THEO/<5.471,-0.073,0>,<0,0,1>,0.284
            ACTL/<5.4716,-0.0711,0>,<0,0,1>,0.2811
            TARG/<5.471,-0.073,0>,<0,0,1>
            START ANG=0,END ANG=360
            ANGLE VEC=<1,0,0>
            DIRECTION=CCW
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
CIR_9_2    =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
            THEO/<3.836,2.378,0>,<0,0,1>,0.284
            ACTL/<3.8354,2.3799,0>,<0,0,1>,0.2813
            TARG/<3.836,2.378,0>,<0,0,1>
            START ANG=0,END ANG=360
            ANGLE VEC=<1,0,0>
            DIRECTION=CCW
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
            MOVE/POINT,NORMAL,<4,6,8>
            FORMAT/TEXT,OPTIONS,ID,HEADINGS,SYMBOLS, ;MEAS,NOM, , , , ,
DIM REF_ORIGIN= LOCATION OF CIRCLE CIR_R  UNITS=IN ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH  HALF ANGLE=NO
AX       MEAS    NOMINAL
X       0.0000     0.0000 ----#----
Y       0.0000     0.0000 ----#----
D       0.2814     0.2815 ---#-----
END OF DIMENSION REF_ORIGIN
DIM REF_CIR_S= LOCATION OF CIRCLE CIR_S  UNITS=IN ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH  HALF ANGLE=NO
AX       MEAS    NOMINAL
X       7.6438     7.6440 ---#-----
Y       4.2129     4.2130 ---#-----
D       0.2815     0.2815 ----#----
END OF DIMENSION REF_CIR_S
            FORMAT/TEXT,OPTIONS,ID,HEADINGS,SYMBOLS, ;MEAS,NOM,TOL,DEV,OUTTOL, ,
     DIM MIS_9_1__MIS_10_1= POSITION OF CIRCLE CIR_9_1  UNITS=IN ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH  FIT TO DATUMS=OFF  DEV PERPEN CENTERLINE=OFF  DISPLAY=DIAMETER
AX       MEAS    NOMINAL       +TOL       -TOL      BONUS         DEV     OUTTOL
X      5.4716     5.4710                                       0.0006          
Y     -0.0711    -0.0730                                       0.0019          
DF     0.2811     0.2840     0.0050     0.0050     0.0021     -0.0029     0.0000 -#-------
D1                                                        PLANE PLN_A AT RFS          
D2     0.2814     0.2815     0.0005     0.0005     0.0004 CIRCLE CIR_R AT MMC          
D3     0.2815     0.2815     0.0005     0.0005     0.0005 CIRCLE CIR_S AT MMC          
TP     0.0041        MMC     0.0080                0.0024      0.0041     0.0000 ---#-----
END OF DIMENSION MIS_9_1__MIS_10_1
     DIM MIS_9_2__MIS_10_2= POSITION OF CIRCLE CIR_9_2  UNITS=IN ,$
GRAPH=OFF  TEXT=OFF  MULT=10.00  OUTPUT=BOTH  FIT TO DATUMS=OFF  DEV PERPEN CENTERLINE=OFF  DISPLAY=DIAMETER
AX       MEAS    NOMINAL       +TOL       -TOL      BONUS         DEV     OUTTOL
X      3.8354     3.8360                                      -0.0006          
Y      2.3799     2.3780                                       0.0019          
DF     0.2813     0.2840     0.0050     0.0050     0.0023     -0.0027     0.0000 --#------
D1                                                        PLANE PLN_A AT RFS          
D2     0.2814     0.2815     0.0005     0.0005     0.0004 CIRCLE CIR_R AT MMC          
D3     0.2815     0.2815     0.0005     0.0005     0.0005 CIRCLE CIR_S AT MMC          
TP     0.0039        MMC     0.0080                0.0026      0.0039     0.0000 ---#-----
END OF DIMENSION MIS_9_2__MIS_10_2

Above output are real numbers from a few minutes ago, going to have them shift the Y of the two holes & we'll be in business.


Thanks all!