4axis output trouble

[marker4x4]

Hi everyone,

I'm out of ideas on this one:

View attachment 3242


It's a simple, circular 4-axis Trajectory cut.
My CL file shows CIRCLE statement (red font below), but GPost outputs point-to-point (blue font):


1>PARTNO / 413A3164S07-BD-7
2>MACHIN/,10
2>PLABEL/OPTION,39,TO,1
2>PPRINT ** TOOL CHANGE # 1 = TOOL NUMBER T 5
1 Pro/NC-GPOST Mill UNCX01 6.0.WN00 P-20.0 MACHIN/UNCX01,10 DATE: 3/ 9/2007 PAGE 1
HAAS HS-1RP
/ 413A3164S07-BD-7 (INCH)
INPUT CLREC N4G3X44Y44Z44B33I34J34K34F34S5T2H2D2M2P41Q23R34
2 2 $ / 413A3164S07-BD-7&a mp;n bsp;
2 2 %$
2 2 O01
2 2 G90 G94 G17 G54 G40 G49
2 2 G28 G91 Z0.
2 2 G90 A0.
2 2 (** TOOL CHANGE # 1 = TOOL NUMBER T 5)$
2>SEQNO/OFF
2>PPRINT 03/09/07-12:29:06
2 2 (03/09/07-12:29:06)$
8>UNITS/INCH
10>PPRINT / 3/8 EM
10 10 ( / 3/8 EM)$
12>SEQNO/ON
12>LOADTL/5,OSETNO,5
12 12 N0010 T05 M06$
12>SEQNO/OFF
14>CAMERA/.422329,0,.906443,0,0,1,0,0,-.906443,0,.422329, 0
15>MULTAX/ON
16>SPINDL/RPM,5555,CLW
16 16 S05555 M03$
17>COOLNT/ON
18>RAPID
19>GOTO /2.847223,.788805,12.1897,.9064,0,.4223
19 19 G000 G094 X-9.8469 Y.7888 A65.019$
19 19 G043 Z7.7288 H05 M08$
20>RAPID
21>GOTO /1.481727,.788805,11.55349,.9064,0,.4223
21 21 Z6.2224$
22>FEDRAT/222,IPM
23>GOTO /1.391083,.788805,11.511257,.9064,0,.4223
23 23 G001 X-9.8468 Z6.1224 F222.0000$
25>ARCDAT/1.44411,1.566231,11.397445,-.9064,0,-.4223,.787 499
26>GOTO /1.497137,2.343656,11.283632,.9064,0,.4223 <br style="color: rgb(0, 0, 255);"> 26 26 X-9.9244 Y.8054$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.0000 Y.8297$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.0727 Y.8615$<br style="color: rgb(0, 0, 255);">1<br style="color: rgb(0, 0, 255);"> 26 26 X-10.1418 Y.9004$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.2067 Y.9461$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.2666 Y.9981$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.3210 Y1.0559$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.3693 Y1.1188$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4111 Y1.1863$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4458 Y1.2576$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4732 Y1.3321$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4929 Y1.4090$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.5048 Y1.4874$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.5088 Y1.5667$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.5047 Y1.6459$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4928 Y1.7243$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4729 Y1.8012$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4455 Y1.8756$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.4107 Y1.9469$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.3689 Y2.0144$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.3205 Y2.0772$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.2660 Y2.1349$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.2060 Y2.1869$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.1411 Y2.2325$<br style="color: rgb(0, 0, 255);"> 26 26 X-10.0719 Y2.2714$<br style="color: rgb(0, 0, 255);"> 26 26 X-9.9992 Y2.3031$<br style="color: rgb(0, 0, 255);"> 26 26 X-9.9236 Y2.3273$<br style="color: rgb(0, 0, 255);"> 26 26 X-9.8460 Y2.3438$<br style="color: rgb(0, 0, 255);"> 26 26 X-9.7671 Y2.3524$<br style="color: rgb(0, 0, 255);"> 26 26 X-9.6878 Y2.3530$<br style="color: rgb(0, 0, 255);"> 26 26 X-9.6088 Y2.3457$<br style="color: rgb(0, 0, 255);"> 26 26 X-9.5957 Y2.3437$<br style="color: rgb(0, 0, 255);"> 27>GOTO /3.441827,2.343656,12.1897,.9064,0,.4223 <br style="color: rgb(0, 0, 255);"> 27 27 Z8.2678$
28>COOLNT/OFF
28 28 M09$
29>SPINDL/OFF
31>POSTN/OUT,20,5,13,6
31 31 T05 M06$
31>POSTN/OUT,20,0
31 31 T00$
31>POSTN/OUT,13,30
31 31 M30$
31>FINI
31 31 %$


I'm sure there's something simple that I just can't see. Anyone?

Thanks!!


Edited by: marker4x4

 

[Don.Rhea]

Mark,


If you look at line 26 it is a 6 parameter GOTO. It is GOTO/X,Y,Z,I,J,K


This tells me you have MULTAX/ON.


Gpost will not do circular interpolation with MULTAX/ON and 6 parameter


GOTO.


Is the part rotated/positioned so the centerline of the circle is parallel


to the Z axis? If not how would you use G02/G03 in an oblique plane?


Don

 

[Don.Rhea]

Mark,


More thoughts. I am assuming your running a HAAS VMC with just the 4th axis rotary table, not the trunion mounted rotary table, and the rotary table has it's centerline parallel to the X axis.


If this is the case the tool axis vector must lie in the YZ plane and therefore the "I" parameter of the GOTO must be zero. This is not your case. There should probably be a "trap" in the post that errors on a "I" other than zero.


The centerline of any circle must lie in the YZ plane. Is this another case of ProE giving you a bad cl file, or is the machine modeled wrong? I am not a ProE guy, just a GPOST GURU.


Don


E-mail: [email protected]

 

[marker4x4]

Don,


thishas beenposted for HMC (horizontal mill) with the axis of rotation parallel to Y Axis. The semi-circular toolpath picturedlies on the XY plane (after rotation) and has its own CSYS that's 90deg. to it, so everything lines up nicely.


This is actually the first case of Gpost breaking down the CIRCLE statement into P-2-P that I've seen... we do quite a bit of parts like that and they always code G2's / G3's as expected. Weird.

 

[Don.Rhea]

Mark,


It looks like I have to re-orient my head from vertical to horizontal. The cl data makes more sense now.


If you would E-mail me the .NCL, UNCX01.F10 & UNCX01.P10 files I'll see if I can get you an answer.


Don

 

[marker4x4]

'K, I've found the culprit.... can't believe it but it's true:

ProE didn't like the angle (65.0183deg from my origin CSYS) of the standing leg that I was trying to machine.

So, I've created my angled CSYS based on sketched geometry rather than the actual model.
This CSYS can be set on any angle from my origin, so all I needed to do was to set it at 65.0193deg and voila! G2's and G3's all the way...

BTW, setting it anywhere within .001deg from 65.0183 will cause ProE code point-to-point... is that odd or what.

Does someone has any explanation to this behavior?

Cheers,

 

[Don.Rhea]

Mark,


This looks like another flaw in ProE that you will never get fixed.


I had a fellow try to machine the outside of a frustrum of a pyramid with


.125 radius cylinder in the 4 corners. The result was terrible. I managed to do


it with APT.


Don

 

[chf65]

Hi Mark,


My best guess is one of 2 issues:


1- Try checking config option MFG_IJK_NUM_DIGITS. This should not be changed from the default (10). By choping digits off the vector, you end up with a marginally different angle/position (exagerated as you get further from the rotary axis pivot location)


2- Make sure that it is not an issue that the vector output by pro/E is indeed as computed by GPost. To do so, we need to see the actual GOTO as stated inside the NCl file. The GPost listing file is not good for this, since it chops the decimal points, so we cannot accurately calculate it. This is more likely than pro/E output, since multax output in pro/E is calculated in double precision and simply output (no transforms are taking place). GPost must do a lot of manipulation of the data, and potentially something may have happened.


Please copy-paste the relevant section from the NCL file so we can help debug it.


Charles

 

[marker4x4]

...
1- Try checking config option MFG_IJK_NUM_DIGITS. This should not be changed from the default (10). .....

Charles, you are DA MAN

That was it, one problem's gone, thanks again.

 

[chf65]

Hi Mark,


It is crazy how many times I see this. Also, MFG_XYZ_NUM_DIGITS although not as severe a problem, should not be changed.


I just do not understand where it came from. Some people tell me they had it this way for 10+ years.


Charles