Cylindrical interpolation: Thumb rule to determine direction of G2/G3

[angelw]

Hey! I've got a really outlandish idea. Why don't you write a short program in MDI and see which direction it goes?

Except that Cylindrical Interpolation can't be executed in MDI, the same as Multi-repetitive cycles can't.

One only needs to look at the example in the Fanuc Operators Manual to see that the direction of G02/G03 are the same as when programming an X/Y profile in a machining centre.

The drawing I Posted here is the same drawing that I Posted many years ago on this Forum on the same subject and the one that I've used for more than a decade to teach clients how to use Cylindrical Interpolation, particularly with regards to which axis to set in parameters that the imaginary axis is parallel to.

Regards,

Bill

 

[sinha]

The positive direction for A axis shown in your sketch is not as per standard convention.
And, one of your statements about G02/G03 directions is not correct.

 

[angelw]

The positive direction for A axis shown in your sketch is not as per standard convention.
And, one of your statements about G02/G03 directions is not correct.

This sketch was more for explaining to those where I was installing a machine and conducting training, as to how to determine which actual axis the imaginary axis is parallel to, which is a requirement for setting a parameter associated with CI. The direction of rotation of the rotating axis was and is irrelevant regarding the direction of G02/G03 moves. G02 will be CW and G03 CCW i9rrespective of the direction of rotation of the rotating axis. In the HAAS example below, the tool path is Climb Milling and therefore, the corner rads are G03, CCW. If the Tool Path was set to Conventional Mill, the corner rads would be G02, CW. Exactly the same as if the part was flat. Hang on, isn't that what CI does for you, allows you to program the part as if the cylinder were to be unwrapped and laid flat?

Which statement was that regarding G02/G03?

Clearly you're doubting what I'm posting; accordingly look up the subject in a Fanuc Lathe manual and view the part illustration and associated code and you will see that the G02/G03 direction is exactly the same as an X/Y program of a mill.

I've set up many mills and lathes with Cylindrical Interpolation and taught many how to create programs with the CI feature. How many times have you actually used this programming feature?

Following is an example from a HAAS manual. The concept is the same for Fanuc, HAAS, Mitsubishi and others. Study the example part and then the program listing. Show me where the G02/G03s are reversed; I would be truly interested, not to mention amazed.



%
O61071 (G107 CYLINDRICAL MAPPING) ;
(G54 X0 Y0 is in center of rectangular slot) ;
(Z0 is on highest point of cylindrical surface) ;
(T1 is a .625 in. dia endmill) ;
(BEGIN PREPARATION BLOCKS) ;
T1 M06 (Select tool 1) ;
G00 G90 G40 G49 G54 (Safe startup) ;
G28 G91 A0 (Home A axis) ;
G00 G90 G54 X1.5 Y0 (Rapid to 1st position) ;
S5000 M03 (Spindle on CW) ;
G107 A0 Y0 R2. (Cylindrical mapping on) ;
(Move to A0 Y0, Part has radius of 2 inches) ;
G43 H01 Z0.1 (Activate tool offset 1) ;
M08 (Coolant on) ;
(BEGIN CUTTING BLOCKS) ;
G01 Z-0.25 F25. (Feed to depth of cut) ;
G41 D01 X2. Y0.5 (Cutter comp on) ;
G03 X1.5 Y1. R0.5 (CCW cutting move) ;
G01 X-1.5 (Linear cutting move) ;
G03 X-2. Y0.5 R0.5 (CCW cutting move) ;
G01 Y-0.5 (Linear cutting move) ;
G03 X-1.5 Y-1. R0.5 (CCW cutting move) ;
G01 X1.5 (Linear cutting move) ;
G03 X2. Y-0.5 R0.5 (CCW cutting move) ;
G01 Y0. (Linear cutting move) ;
G40 X1.5 (Cutter comp off) ;
(BEGIN COMPLETION BLOCKS) ;
G00 Z0.1 M09 (Rapid retract, Coolant off) ;
G91 G28 A0. (Home A axis) ;
G107 (Cylindrical mapping off) ;
G90 G53 G49 Z0 M05 (Z home, Spindle off) ;
G53 Y0 (Y home) ;
M30 (End program) ;
%

 

[sinha]

Which statement was that regarding G02/G03?

Refer to the previous posts.

 

[angelw]

Refer to the previous posts.

Well, seeing as I've gone to great length to try and explain to you how CI works and that the direction of G02/G03 are in fact not reversed, perhaps you could be more specific and point it out.

 

[sinha]

We may discuss, but what is more important is OP is still confused.

 

[angelw]

We may discuss, but what is more important is OP is still confused.

He and you seem to think that the direction of G02/G03 are revered. I'm telling you that they are not, and I've provided ample proof that they're not. The whole concept of CI is that the cylindrical surface is programmed as if unwrapped, laid flat and programmed as you would an X/Y profile on a mill.

CAM packages are relatively easy to create a Post Processor for to output the correct syntax and Finger CAM is no more difficult than programming a 2.5D profile. If you can't accept that or understand it, well, one can lead a horse to water, but not necessarily make it drink

 

[sinha]

Did you see the figure posted in his first post?

By the way, in case you are not aware, you are interacting with somebody who is PhD in mechanical engineering from a world-class institute.
Better to control your arrogance.

 

[angelw]

Did you see the figure posted in his first post?

By the way, in case you are not aware, you are interacting with somebody who is PhD in mechanical engineering from a world-class institute.
Better to control your arrogance.

Sinha, I think you're all hat and no cattle.

 

[sinha]

I am not questioning your knowledge. I have learnt quite a few things from you, for which I am grateful to you (I have also included some of these things in my books). You are one of the most knowledgeable persons on this forum. A very important member, indeed.
But, nobody is right or wrong all the time.

 

[angelw]

But, nobody is right or wrong all the time.

Absolutely. I have put up evidence that supports my own experience, and directed you to view articles in Fanuc Manuals, that clearly supports what I have Posted, but you still seem to doubt its validity.

As I've pointed out in previous Posts, many other makes of controls have the CI features and the architecture is basically the same. It's not only Fanuc controls I know well, I have clients with just about every brand of control that you can think of, that I support.

I've taught hundreds of people how to use CI, written many Post Processors for CAM software to handle this feature and its no more difficult than Posting an X/Y program for a mill, except for the axes address.

The hardest part to come to grips with and its not all that hard, is that the units for the pseudo X (C) axis of a lathe program are in degrees and therefore, the longest dimension of the unwrapped cylinder is 360 units. When initiating CI, the Radius Value of the cylinder must be specified to make it work.

You simply have to forget the Right Hand Rule related to a rotary axis and simply think of the Cylinder unwrapped, and laid flat. At that point, you can apply the Right Hand Rule as if dealing with a machining centre cutting an X/Y profile; it's no more difficult than that.

I'm also a mechanical engineer, but I specialized in CNC methodology and application, as a business, so I don't bandy the mech engineering qualification around much. Do you know how to tell if a person is a mechanical engineer? They tell you.

 

[sinha]

I have never doubted your CNC knowledge. In fact, you do not need anybody's testimonial. You have proved your worth.

 

[rjwalker1973]

Pulling it in the Z+ direction from the front of the roll, as shown in the picture, will have the roll rotate in a CCW direction looking at the end of the roll that has the direction arrow drawn. The arrow indicates that the roll is rotating in a CCW direction.

When considering whether G02/G03 being reversed or not when using CI, the rotation of the "C" or "A" axis doesn't come into it, for, in terms of programming, the Cylinder is unwrapped and laid flat whereby "C" becomes the pseudo axis "X" and "A" the pseudo axis "Y" or "Z"

For some reason those two little passages helped me make sense of what you said. I mean I had it, but this helped it make even more sense. Thanks Bill.

 

[angelw]

I have never doubted your CNC knowledge. In fact, you do not need anybody's testimonial. You have proved your worth.

I'm not offering a testimonial, nor am I seeking your approval. I know what I know and the likes of Kevin, AKA Vancbiker, others and I, learnt NC and CNC programming when there were no teachers, or Forums; it was basically a matter of swimming or sink. Particularly when running your own business, you have to learn quickly and develop procedures that makes for efficiency in methodology and programming.

 

[sinha]

Yes, quite appreciable!

 

[rjwalker1973]

I'm not offering a testimonial, nor am I seeking your approval. I know what I know and the likes of Kevin, AKA Vancbiker, others and I, learnt NC and CNC programming when there were no teachers, or Forums; it was basically a matter of swimming or sink. Particularly when running your own business, you have to learn quickly and develop procedures that makes for efficiency in methodology and programming.

I had to grab manuals and read them. I got tired of asking questions and getting the , it just won't , answer when I asked if something would work. Luckily, I eventually got into a shop with some guys that was willing to teach A young kid with more questions than answers.

 

[sinha]

Any serious guy has to read the manuals. I also did the same. Quite a few things were not very clear. I never hesitated in asking questions on online forums. These helped me a lot. Also, I had enough time to do experimentation and draw my own conclusions.
Additionally, I also read macro book of Smid, cover to cover.

 

[Jaanyaar]

You simply have to forget the Right Hand Rule related to a rotary axis and simply think of the Cylinder unwrapped, and laid flat. At that point, you can apply the Right Hand Rule as if dealing with a machining centre cutting an X/Y profile; it's no more difficult than that.

Thanks. This way, I can understand your previous image about CI in a vertical machining center.

Now let me ask about the image of my first post. Is my pseudo X and pseudo Y correct?

 

[rjwalker1973]

Additionally, I also read macro book of Smid, cover to cover.

I keep a pdf version of the Smid book with me . That's how I learned what macro programming I know. That and asking questions on PM. Of course PM didn't exist when I started . I think I was about 10 years in before it did, or at least when I found it.

 

[cameraman]

Hardly a week would go by without me having gone to bog and then attempted to wash my hands to then find the true meaning of "two-tenths".

.

There are other ways for "We" fach-idiots to have "Fun" - you are not your Job ... (or jobbies).
~
For some reason I have this vibe that Bill Howard may have retired or is planning to ?
~
It's been a while...