R vs I J interpolation code. Which is better?

[honeybunches]

We seem to be having issues getting R codes to work on a Haas VMC and keep getting 303 and 330 alarms. I think we can get I J to work but was curious if that is preferred or if there is any advantage of one or the other? I notice the R coding is shorter and I really like that because the machine we are working with is older and the process or is slow so we are trying about everything to get it to keep up. I figure, the less code, the better.

 

Opinions?

[Allan]

In my experience the IJ formatted arcs are more reliable. In using R the machine calculates the IJ position sometimes arriving at the incorrect answer.

 

Allan

[Jack Neelands™]

IJ...sorry Allan is correct...R looks great on the screen but can be flipped on the opposite side by the machine....yes this has happened before and doesnt look to good on the part either

[john316]

.

 

IJ for the same reason stated above.

 

An R- will cut the long way around the arc but it's still safer and more consistent with the IJ

 

.

[Dave.L]

IJ is far more accurate. Last year we made a change to our control by mistake and it took days to figure out why we have radial features that were off location. We found the error, re-posted with IJ and all errors were gone.

[honeybunches]

WOW, that really makes me wonder why R code is even still in the picture. Answers seem unanimous.

[Justin Beebe at Folsom Tool]

IJK is more accurate in my opinion. Two times I had parts coming off the machine and the radial features were coming out almost out of tolerance. Both times I noticed I was posting R's instead of IJK's and switching back fixed the problem. I beleive the default selection for mpmaster machine def and post is R's. It seems to me that the default should be IJK's since most people prefer IJK's to R's.

[shazam/TPP]

ijk ditto ^^^

[Thad]

quote:

---

WOW, that really makes me wonder why R code is even still in the picture.

---

I think people are over exaggerating the error with using R.

 

Thad

[Dave.L]

Sorry, no exaggeration whatsoever. We saw (yes, with our eyes) the difference.

 

"R" is really mean for adding radii on corners, and that's all.

[pip]

I have been tasked with the duty of teaching others at work how to program at the machine. Only simple items, and I often have to show this example to get it over to them, what can happen when using R.

Check your program thourouly before running.

On 1 mill, we never have a problem, but the older one, can do this with nearly every example.

[Guest]

quote:

---

I think people are over exaggerating the error with using R.

---

Thad, no we're not really, the geometry being cut will matter to some degree but I have seen with my own 2 eyes, "eggs" being cut, that as soon as the post was changed over to IJK, the machine, a "BRAND NEW" Mori NV5000 cut perfect circles.

 

No exaggeration on these peoples part.

[Josh T™]

If you are dead set on using R instead of IJK I would strongly suggest breaking the arcs into quadrants. Sometimes the control interprets the code much like Pip's illustration and the arcs get screwed up... most of the time this has happened to me in a case where the arc is more than 180 degrees. Breaking at quadrants would solve this case and alot of the issues stated in this thread. My guys like seeing the R's and so I've left the posts alone. I agree that IJK can be more accurate but it really depends on personal tastes and application needs.

 

HTH

Josh

[Dave.L]

Josh,

 

I feel strongly about this from experience. When using R values, we did not have obvious arc issues (meaning wrong direction, etc), however, we had 270 degree arc off location by 0.030" or more, so, this is not a taste issue, it's an accuracy issue.

 

Summary, if you want accurate parts, don't use R values.

[Philcott]

The reason R can give you errors is that there is often more than one way to get from point A to point B. Having code such as X* Y* R.5 tells the machine to move to a position by moving on .5 rad. It does not know which .5 rad path to take. Using I J K tells the machine exactly where to go because it knows where the starting point is and the I J K describe the distance from the start point to the center point of the arc. There can be only one solution to the path when using I and J.

[Acro Engineering]

What method is the best to use (Absolute, Delta start to center, or Delta center to start)?

[Superman]

quote:

---

What method is the best to use (Absolute, Delta start to center, or Delta center to start)?

---

Only one will work, the others will either error the machine, or put the arc centre somewhere else and create an incorrect radius.

 

Machine manual will tell you what it must be and there will be no other choice.

[Bryan Johnson]

quote:

---

What method is the best to use (Absolute, Delta start to center, or Delta center to start)?

---

I do not understand why more control manufactures do not support absolute arc centers. Using absolute, would allow the path to always swing about the same center point regardless of start/end point. With Delta or with R value, there is rounding that can give inconsistent results.

I can think of very few controls that support absolute, and Fanuc aint one of em.

[Jim at Gentex]

quote:

---

There can be only one solution to the path when using I and J.

---

Yep.

I totally agree with Phil's statement.

 

I had the same experience as many here have stated with a part we run almost every day. We were getting out-of-tolerance results on some holes using R values. Changing to IJK fixed the problem and we haven't seen an OOT since.

[Rob @ Target Machine]

I have to agree that IJK is a whole lot safer, but there is no reason why R's can't be used if you break into quadrants, whether the machine is looking for an R or an R- for more than 180 Deg.

 

I am having no problems using R's for many different machines.

 

You just have to understand how R calculates to IJK in the control and "Play by the Rules".

 

IMHO

 

Of course...I prefer IJK, but the higher up's prefer R's, in my case, so I'm stuck with them.

[MMT-USA]

It is always best to program using I,J,K

 

R: Used, if desired, in place of I,J,K. R is programmed simply as the distance between the arc center and any point along the path.

· R is a positive value if the arc swings through LESS THAN 180°

· R is a negative value if the arc swings through MORE THAN 180°, but LESS THAN 360°

· If the arc swings through EXACTLY 180° or EXACTLY 360°, R cannot be programmed, Instead use I,J,K

 

 

I,J,K : Signed (+ or -), incremental distance measured from the start point of the arc to the arc center.

· I,J,and K values are measured along X,Y, and Z axes respectively

· I,J, and K values are always incremental, without regard to G90 or G91

· If I,J, or K values are found to be zero, the programmer may omit that I, J, or K word.

[Tony35]

R means roughing or redo

 

i never use R always IJK i had to many problems with out of round issues use IJK and all was good

[Rob @ Target Machine]

quote:

---

If the arc swings through EXACTLY 180° or EXACTLY 360°, R cannot be programmed, Instead use I,J,K

---

180 deg is R

180+ to 360 is R-

 

This is usually the case.

 

If different machines, usually older ones, vary in how R is interpreted, breaking to quadrants fixes the problem just as IJK will.

 

That's why I say "Play by the rules".

 

Depending on the control, of course, R is interpreted in different ways depending on the firmware/parameter settings.

 

When in doubt break to quadrants, and all interpretations will be the same at the control, just as IJK have only one solution.

 

I would, however, choose IJK if I had any say in the matter.

 

The problem is R will do the move no questions asked even if it is wrong, and IJK will cause an alarm if it isn't logical.

 

[ 05-12-2009, 06:03 PM: Message edited by: mastercamguru ]

[Thad]

EDIT: Wrong info.

 

Thad

[MMT-USA]

(EDIT: Wrong info.)

 

This is control specific. Fanuc and Yasnac