Compensation Type set to control or wear?

[ujmujm]

When I do a contour tool path for a mill  for compensation type I've used "wear" for many years and never had any of  the issues that I use to have when I used "control"  and now I work at a shop where the tool room manager tells his people to program with compensation type set to "control" because that's how everyone does it.

Any thoughts

[Guest]

New job.....

Control is more problemas that it's worth...

Do a search, this battle gets fought about every other year here

[Matthew Hajicek - Singularity]

"Control" is for emulating a hand-written code style, and to allow impromptu substitutions of say a 1/2" endmill for a 3/8".   This was fine in the 70's and 80's when a shop didn't have to be very efficient to stay in business and HSM strategies hadn't been invented yet, but it makes it far more difficult if not impossible to write efficient and optimized code.  A toolpath optimized for that 3/8" cutter with a factory edge prep is not optimized for a 1/2" cutter or a .365" regrind; if you want to be efficient and maximize profits you pick the best tool for the job, optimize your program for it, and use that specific cutter when you actually make the parts.  No substitutions, no regrinds (except perhaps for very large and expensive tools), no hand editing on the control, those are all wasteful and error prone.  If you're using a 3/8" cutter to the extent of it's capabilities, you will snap a .365" regrind in the same path.  If you detune your path to allow the regrind, you're wasting time when you run the full size cutter.  Your operator isn't going to follow the code of a dynamic path anyway as the thousands of lines scroll past on the screen, and he sure as heck isn't going to tweak it in any beneficial way.

Subbing a different sized cutter would also invalidate your toolpath verification.

[ujmujm]

I always do a search before I post a topic, but nothing came up, I guess I have poor searching skills, I've been with this company about a year now, but I have  run into this other places I've worked and I've always thought these people were very old school and not willing to change, they still use calipers with vernier scale lol just for kicks I wrote a tool path using "control" and posted it and was surprised at the out come, back in the day when your lead in lead out was critical not any more I guess.

[SlaveCam]

I have always used control and haven't had any problems with it since I learnt that lead in and out MUST be over 50% in Mastercam. 

Regrinds here

[Newbeeee™]

Where I was, we used Control. One reason is because initially I knew no better, and the other was because the guy who I was with had "always done it that way".

After reading here and seeing the great answers as to why we should be using wear, I was convinced.

The problem was for where I was at, there was a lot of progs already written, and to change would have been a nightmare. I could see a lot of expensive scrap where 0 was in the register opposed to 1/2 the cutter diameter.

And that's another stupid bloody English thing - we call diameter offset diameter offset, but use the radial value. Yes I know it's only a parameter to change to diameter, but everyone I know in all the shops I know (UK), run a radial offset...

[Matthew Hajicek - Singularity]

I've found that regrinds have 20% to 50% the life of a fresh tool, so in my size range it loses money.  I can understand using regrinds for $200+ endmills, but for a $40 one?  No way.  I can also understand using a regrind for a finish pass where the size of the tool doesn't really matter much, but in a dynamic path it's going to cost you.  There will be more material left on the part than Mastercam thinks there will, so you'll have to add extra finishing passes to account for it, and if your path is truly optimized for the on-size cutter you'll have to slow down to run the regrind.  If the regrind runs the same path without failing then you could have pushed the on-size cutter harder.

[ujmujm]

Thank you for the excellent feed back as always, I wasn't sure if I was the one being narrow minded, I'm  always opened to suggestions if it adds value, I do feel bad for the guys in the tool room though.  

[Leon82]

Can The Tool Room manager really tell you to do that or is he just giving you his opinion?

[nickbe10]

44 minutes ago, SlaveCam said:

I have always used control and haven't had any problems with it since I learnt that lead in and out MUST be over 50% in Mastercam. 

And this is the other problem with running comp in control. What if your feature requires a shorter lead in because of its geometry but you need a larger diameter endmill than the comp in control requires?

Much less flexible IMHO, nothing more annoying than a cutter gradually comping itself as it does a profile because there was not enough room to buy the initial comp......

This and Centerline vs. Macro rotation calculations on 4 axis will always be a source of debate.....

[Henk]

On 6-10-2017 at 9:45 PM, Leon82 said:

Can The Tool Room manager really tell you to do that or is he just giving you his opinion?

IMHO he SHOULD order this, using both systems will be a nightmare.

[gcode]

On ‎10‎/‎6‎/‎2017 at 1:00 PM, nickbe10 said:

 

This and Centerline vs. Macro rotation calculations on 4 axis will always be a source of debate.....

that is the only sure thing on this subject

We debate it endlessly, and I doubt many minds get changed.. though hopefully

the youngsters joining the industry get steered clear of Control Comp early in their career.

Back in the day before there was Cad/Cam  I used PencilCam  and control comp because

it was easier to calculate the required geometry.

Now I use Wear.

Thankfully I've never worked anywhere that used Control Comp

It can be used successfully but requires knowledgeable and alert operators to avoid disaster.

The more complex your parts the more likely you'll get bit by Control Comp.

One thing to keep in mind...  no CAD/CAM system can know with 100% reliability what any given machine

tool will do when Control Comp is activated ... what you see in backplot/verify is only a best guess.

In simple contours it's probably right. In tight contours and tricky entry/exit motion it's probably not. 

Personally I avoid Control Comp at all costs

[Thad]

On 10/6/2017 at 3:17 PM, Matthew Hajicek™ - Conventus said:

I can also understand using a regrind for a finish pass where the size of the tool doesn't really matter much, but in a dynamic path it's going to cost you.

Correct me if I'm wrong (that probably went without saying ), but dynamic paths don't have wear/control comp settings.

[nickbe10]

1 hour ago, gcode said:

We debate it endlessly, and I doubt many minds get changed.. though hopefully

 

1 hour ago, Henk said:

using both systems will be a nightmare.

And herein lies the rub. If you are already committed to one of these techniques it is incredibly hard work to change, not only have you got the volume of committed programs to consider but you will also be changing the "culture" on the shop floor.

I was involved with converting a shop to centerline programming which had been in business for 30 yrs. The resistance of the operators/machinists was huge. Luckily we had full management backing. For 18 months it was like pulling teeth without anesthetic, when we finally hit a critical mass of programs the floor guys came to realize they actually had just as much control and all they really had to worry about setting was TLOs. I remember asking one of the biggest original nay-sayers if we should go back to rotation macros and the reply was "Why would we want that nightmare back......"

It's tough......

[Matthew Hajicek - Singularity]

3 hours ago, Thad said:

Correct me if I'm wrong (that probably went without saying ), but dynamic paths don't have wear/control comp settings.

Exactly.  So what happens if you use a .460" regrind for a dynamic path that was programmed for a .500"?  In addition to having a significantly weaker tool with uncertain edge prep and reduced chip clearance, it leaves .020" extra stock for your finish pass to try to take.  I'm arguing why not to use regrinds because one of the main arguments for control comp is easier use of regrinds and swapping out different diameter tools without having to change the program.

[Leon82]

We used to have a few older programs with a combination of control and wear. I never ran them but there were a few times the value wasn't imput and parts were scrapped.

[Colin Gilchrist]

2 hours ago, Leon82 said:

We used to have a few older programs with a combination of control and wear. I never ran them but there were a few times the value wasn't imput and parts were scrapped.

Forgetting the correct entry is an easy way to scrap a part. One useful feature of the machine is that is essentially "adds" both CRC columns together when invoking Cutter Compensation. I used this to my advantage recently in fact.

We have a very old legacy job that is programmed using full CRC. So you must enter the actual cutter radius when setting up the job. No problem, we have a tool pre-setter, so it's not a big deal.

The NC program however was created in a CAM system that we no longer have access to. So all we have to work with is the G-code. I found myself needing to make dozens of tweaks in different places. But I didn't want to have to "move" by editing the G-code.

My solution was to use the Macro programming capability of the Control to my advantage. I added lines of logic to write new values to the Tool Offset Wear Registers for both radius, and length. I was able to switch from .002, to .01, to -.03, to .121, all "on the fly". Since the Wear register gets "added" to the CRC field, I could easily move +- from the periphery of the cutter, or tip of the tool. I even had some passes where I would use a negative offset on 'one side' of a cut, then switch the offset to positive, just before the move to the other side. It worked beautifully. I could take a -.0032 cut on the left, then a +.0128 cut on the right. Sooooo much easier than having to use a G52 coordinate shift for each cut on the part. And, the kicker for me was that I could adjust the part without "reprogramming it", since that was rejected by the customer. I was able to show that my modified program matched the original program path in Vericut, so they approved my method of improving the process.

[Leon82]

44 minutes ago, Colin Gilchrist said:

Forgetting the correct entry is an easy way to scrap a part. One useful feature of the machine is that is essentially "adds" both CRC columns together when invoking Cutter Compensation. I used this to my advantage recently in fact.

We have a very old legacy job that is programmed using full CRC. So you must enter the actual cutter radius when setting up the job. No problem, we have a tool pre-setter, so it's not a big deal.

The NC program however was created in a CAM system that we no longer have access to. So all we have to work with is the G-code. I found myself needing to make dozens of tweaks in different places. But I didn't want to have to "move" by editing the G-code.

My solution was to use the Macro programming capability of the Control to my advantage. I added lines of logic to write new values to the Tool Offset Wear Registers for both radius, and length. I was able to switch from .002, to .01, to -.03, to .121, all "on the fly". Since the Wear register gets "added" to the CRC field, I could easily move +- from the periphery of the cutter, or tip of the tool. I even had some passes where I would use a negative offset on 'one side' of a cut, then switch the offset to positive, just before the move to the other side. It worked beautifully. I could take a -.0032 cut on the left, then a +.0128 cut on the right. Sooooo much easier than having to use a G52 coordinate shift for each cut on the part. And, the kicker for me was that I could adjust the part without "reprogramming it", since that was rejected by the customer. I was able to show that my modified program matched the original program path in Vericut, so they approved my method of improving the process.

Must have been a flight safety part. Once the process is approved your stuck with it unless you get it re certified

[Thad]

15 hours ago, Matthew Hajicek™ - Conventus said:

I'm arguing why not to use regrinds because one of the main arguments for control comp is easier use of regrinds and swapping out different diameter tools without having to change the program.

And all of your arguments have to do with dynamic toolpaths, which has absolutely nothing to do with the question asked. He's asking about control or wear comp in a contour toolpath.

[Rstewart]

The thing I love most about wear is that its really hard to get a CRC alarm cause you didn't make a correct lead-in move.

Wear Is the Only comp I'll ever use.

[Matthew Hajicek - Singularity]

On 10/11/2017 at 5:02 AM, Thad said:

And all of your arguments have to do with dynamic toolpaths, which has absolutely nothing to do with the question asked. He's asking about control or wear comp in a contour toolpath.

Are you arguing then that OP should as a rule never use dynamic paths?  Or mix comp types?

Look at it this way; why would someone want to use control comp?  The most common reasons are to make it easier to hand edit code, and to sub out different sized cutters.  If you're going to use 21'st century techniques like dynamic paths, both of those reasons are invalidated.  The only remaining reason is "because we've always done it this way."

[RonC.]

What would be nice if one of you geniuses .. come up with a converter to change previous programmmed

job from control to wear... :-)

[Thad]

13 minutes ago, Matthew Hajicek™ - Conventus said:

Are you arguing then that OP should as a rule never use dynamic paths?  Or mix comp types?

I'm not arguing anything. I'm trying to figure out why you're talking about oranges when the question is about apples.

It's the age old question of control vs wear and all of your talking points are about dynamic toolpaths WHICH DON'T EVEN HAVE THOSE SETTINGS.

That's all I've got. I'm done here.

 

[Matthew Hajicek - Singularity]

Just now, Thad said:

I'm not arguing anything. I'm trying to figure out why you're talking about oranges when the question is about apples.

It's the age old question of control vs wear and all of your talking points are about dynamic toolpaths WHICH DON'T EVEN HAVE THOSE SETTINGS.

That's all I've got. I'm done here.

 

Dude, chill.  We're just talking.

I'll try to spell it out a little plainer.

1. OP was asking if control comp or wear comp is better.

2. I think it's fair to assume that the shop has reason to operate with reasonable efficiency.

3. If a shop wants to run efficiently, they will use dynamic paths for roughing in at least some cases.

4. If you use dynamic paths anywhere in any program, all of the arguments for using control comp go out the window.  Yes I know that comp isn't used in the dynamic paths, but if you have any dynamic paths in your programs your operators need to know that they can't be swapping out tools for different diameters because bad things will happen.

5. The arguments about program optimization apply even without dynamic paths.  Subbing tools of different diameters will invalidate your toolpath verification, lead to gouges and broken tools, and reduce efficiency.

[gcode]

37 minutes ago, Matthew Hajicek™ - Conventus said:

4. If you use dynamic paths anywhere in any program, all of the arguments for using control comp go out the window.  Yes I know that comp isn't used in the dynamic paths, but if you have any dynamic paths in your programs your operators need to know that they can't be swapping out tools for different diameters because bad things will happen.

the same applies to old school pocketing routines

If you program a pocket for an Ø.750 endmill and run a regrind at  Ø.710, your finish wall pass will be .020 thicker than you planned.

This may make the difference between success and failure