What's new
What's new

Circular Interpolation

TimH

Hot Rolled
Joined
Oct 12, 2004
Location
Upstate NY
Can I use G03 or G02 to do an arc defining it with degrees & radius not an x/y end point. Thanks Tim
 
Holy Cow Tim! I actually looked in the manual just in case I was missing something!
The short answer is no, you can't.
The long one is that radius and angle is not enough information for the controller. It needs start point , end point and either the radius or location of arc center.
Now, interpolating from that, your start point is given by where your machine is from the previous block. Your end point is needed for the motion in next block, and now all you need is the radius or the center point. Which makes me think that you're trying to calculate the points by reeeaaaallllllyyyy longhand?
Degrees can only be defined in bolthole patterns as start angle, incremental angle and # of holes, but then it still needs to know the center point and radius.
 
hmmm


what is you used a G70 for a bolt circle...
but instead of drilling, you put a tool down and leave it down...

Like a G81 with an R of Z-.125 for instance...
Start up the spindle, then dry run the program...
So the rapids are controllable with the jog dial...

Tell the machine to drill from Z-.125 to Z-.125 with an R of Z-.125 to the depth never changes...

Then you can tell the machine that you want to start a bolt circe at X,Y drill a hole every 1 degree or something like that...

Guess that would not exactly be a radius, more like very small straight lines...

But that way you could control the degrees...

43 holes 1 degree apart = 43 degree ark...

OOhhh... sounds fun... wonder if its even possible...
 
Doug

Radius and angle has an infinite number of solutions, so you still need more info. Endpoint still not enough, there is still alway at least 2 solutions. A simple R is also not enough, still 2 solutions. You need a signed R and endpoint, in which case angle is mute.

Solar

Pretty inventive thinking.... Except you'll be slam-milling between hole locations doggy-style.
 
Radius and angle has an infinite number of solutions, so you still need more info. Endpoint still not enough,
If you have the start position, direction of rotation and circle center location (i,j) then there is only one solution.
You can use G12 or G13 to define a circular move using a polar coordinate system.
 
hansdie

You're correct in that case, except Tim wanted only angle and radius without center < I and K > location or end point.

Tim

Yes, G12 and G13 are two awsome cycles, specially with full cutter comp. Doesn't apply here since you want angle and radius, but for interpolating a circle they are indispensible.
Still wondering why you'd want the angle as a definition though?
 
Yes, G12 and G13 are two awsome cycles, specially with full cutter comp. Doesn't apply here since you want angle and radius, but for interpolating a circle they are indispensible.
Still wondering why you'd want the angle as a definition though?
I have to program multiple "arc" passes. I like to work from the center of my work piece versus a corner. On my old Heidenhain you could do coordinates either linear or polar. It was pretty handy. I'm suprised it can't be done with G codes.
 
to seymoure...

just wondering...

in my post i said to put the machine in dry run mode so (No rapids) all speeds controlled by jog setting... (on my machine F.2-F157)

would that work ?
 
Tim


The machine still neds to know where the center point is, Heidenhain or not!
As for center point programming, we all do it. Center point, off-part 0 point, whatever.
I still think you're a little unsure how to use the work coords. Feel free to post a question.
 
Solar

The doggy-style remark referred to the way machines make rapid, i.e. non-interpolated motions. It's commonly called dog-leg rapid motion, whereby each axis travels at full acceleration to it's required destination without attempting to syncronize with the other axis. The motion between hole locations are rapid motions, so even with rapids turned way low you still get linear motions that looks like 45 deg for the beginning and straight at the end in the axis that has a longer distance to go.
 
I see...
Oh well...

In my controller manual, it has a G71 (BC via Ark)

I could do...
-------------
G0 G54 G90 X.0 Y.0
Z-.125
G81 Z-.125 R-.125 Fx.xxx
G71 I3. K1. L53

This would make 53 holes 1 degree apart...
3 inch radius... for 53 degrees...

But as you said there would be small little straight lines... not a true ark...

Oh well...
 
The machine still neds to know where the center point is, Heidenhain or not!
As for center point programming, we all do it. Center point, off-part 0 point, whatever.
I still think you're a little unsure how to use the work coords. Feel free to post a question.
On the Heidenhain I just defined CC ( circle center) then program Circular Path then define Polar Radius & Polar Angle.

You're right SeymoreDunmore I'm not very comfortable with the coordinate sytem....Yet. I'm sloooooowly getting the hang of it. I programed the Heidenhain (conversational)for the last 20 some years so this g-code stuff is all new to me
 
Tim

With the Circle center, you gave the I and K equivalent in G-code. Now all you need to do is use the endpoints instead of the polar angle and you're there.
 
My book says G12 & G13 are for circular pocket milling.
Well my book and my standard (DIN 66025) say it's for circular path in polar coordinate system. Circular pocket milling would be G77 and G78 according to my book.

Still wondering why you'd want the angle as a definition though?
It is useful for doing circle moves greater than 360 degrees, programming off a print where features are dimensioned in angles with given center points, that way you don't have to do any calulations. You can also progam lines with just angle and distance. So why plot or trig it out when you can enter in the coordinates directly...less work.
 
Hansdie

The G12/13 is a Yasnac style circular pocket milling canned cycle. The DIN equivalent maybe different, but it's a Yasnac holdover.

As for the angle and distance, you still ain't gonna get away without calculating the circle center. Distance and angle is not enough info!
 








 
Back
Top