Why I,J,K over R?

[DavidSV]

I see a lot of code posted on here, and most of it is in I,J,K radius output, and I was wondering why? I have only used R, so I don't know the difference. Can anyone explain why its better, or is it just different?

 

Thanks

[Midwest]

Linky

 

 

Been talked about alot B4

 

 

HTH

 

 

[KCollins]

my control wont accept an R (an old cincinatti that was born before me)

[Craig Madsen from CM3D]

The short reason:

 

Your machine has internal resolution limits. This means your control has to round values to fit your command to its nearest values.

 

draw two points. -1.0,0 and .9999,0

put a arc with radius 1.0000 between the two points. Analyze the arc center.

 

 

If you use R and are near 180 degree your control calculates where a 'center of rotation' is at. Since the machine resolution may be +/- .00004 x 2 x 1.4 (the 1.4 is for rounding error) you will see the control move the center up to +/-.014" from what you (using what YOU KNOW) expect. I've seen the result of this dozens of times. The error doesn't even happen the same on every part if the numbers match up just right.

 

If you use IJK, you are telling the control what to use for the calculation. This forces the control to conform to your idea of where the center of the arc is at - not fit an arc.

 

Yes, you can get away with it most of the time. But you can't get away with it ALL of the time. Just like I can drive my car 100 mph across the Nevada desert and not get a ticket. (Most of the time).

[Mike@Apollo]

I have noticed that when entering/exiting an arc "off-quadrant", an R value will result in (sometimes) severe out-of-round conditions.

 

Doesn't seem to be a problem when using I,J,K method.

[mkd]

i'm thinking a person really needs to have a post ready for either method. i have always thought i,j,k is THE fool proof method. a couple years ago I learned different. I had a joggle/step along a wall that was nearly equal to the cutter radius with a corner radius nearly equal also. the MAPPS II mori control blew thru the wall in a major way, yet the program looked fine in Vericut. I reposted with R values, and the parts came out fine.

[Todd Tracey from Fastech]

Some machine tools can't accept R's and some don't like I,J,K depending on the control is another possibility.

[jimp]

Same here we have 2 Kellenberger cylindrical CNC grinding machines one use 'R' other I,J,K

both produce the same results.

[Jaz]

Theoretically 'R' programs would produce the same parts as 'I,J,K' programs. But because the control has to compute the actual path and it starts with different data in each case differences can occur (due to rounding errors). Check your machine's manufacturer for which method is recommended.

[Bryan Johnson]

It looks to me like the way to go would be for IJK as absolute values. That way, the center point would always be the same regardless of where you start or end the path.

 

I have seen this option on a few controls, but never on Fanuc. I wonder why?

[?Mark]

If control accept both ways then either is good as long as you break the arcs at quadrants (full circle being the only exception where ijk's are the way to go)

I personally like to see R's, as I can tell what's behind the number just looking at it, but most importantly, I'm used to it because of limited memory...even though the memory is not a problem these days

 

imvho