Jump to content

Welcome to eMastercam

Register now to participate in the forums, access the download area, buy Mastercam training materials, post processors and more. This message will be removed once you have signed in.

Use your display name or email address to sign in:

Rant: Dynamic toolpath slivers


JB7280
 Share

Recommended Posts

Sometimes I think that Mastercam is in cahoots with endmill manufacturers to leave slivers and towers on dynamic toolpaths just to try to break tools.  There's some evil guy behind the scenes that changes the code before every new MC release, so that MC intentionally leaves a big sliver in the middle of every dynamic from outside toolpath.  

 

I know every software has its highs and lows, but I don't remember ever working around this problem in the other CAM packages I've used.  

 

End of my rant.  Thanks for listening, enjoy the rest of your day.

  • Like 1
Link to comment
Share on other sites

This is a problem inherent in every package that uses dynamic-style motion, and it's a tricky one to address. At some point, you are asking the path to break with its directive of constant cutter engagement to deal with the end result of the toolpath echo near the center of the part. The first problem to address is- what constitutes a sliver? A certain length vs width ratio? Acute corner angles under a certain degree mark? A certain burr height the tool might push? Should the tool engagement be considered in determining a 'sliver' to avoid? The answer is really- all of those along with a few other variables.

Roughing of remainders that are fine at 0.5" of flute depth become tool killers at 3" depth. Fine stepovers with certain corners become tool killers if the stepover is increased a bit. Aluminum vs titanium changes the interplay of everything we just mentioned that might be a variable. How do we get this right for every tool and scenario and come out with that ideal "Image" that I have as a programmer of what is safe but still fast for my part shape? 

So yes, it's tricky on a toolpath generation side, but what can we do right now that can help this? There are a few things-

  • Use the Corner Pre-Treatment available in 2D dynamic to effect changes in the echo at the center of the part. Especially with big stepovers, trying to do a dynamic path on a shape with sharp corners will result in the toolpath folding in on itself towards the center of the part and creating potentially problematic slivers, arrowheads, or double-backs. By using corner pre-treatment, we're taking a tiny bit of material off the corners at the outset that has a huge positive impact in how the toolpath echoes to center.
  • As a user, recognize when we have large discrepancies in length to width ratio in the shapes that we're cutting, and try to 'head off' the sliver issue with programming changes. Usually I accomplish this through a few different ways, like purposely splitting up the cut area to produce better center results, or purposely put in an avoidance island with a wireframe circle or oval that the toolpath can work down to, and then use a second dynamic to safely and evenly clean up that avoidance "island"

In any of these cases, what we're doing is manually controlling when and how we violate our dynamic principles in order to affect the end shape where the slivers are being produced.

 

Regarding that corner pretreatment, check out this path where we're power milling (65% stepover) a relatively square profile. Even here we start to see some folding of the echo towards the center in order to maintain constant engagement. Especially with such high cutter engagement, anywhere I see an acute angle start to form is going to be a suboptimal corner for my cutter to take.

769042228_PowerMill.thumb.jpg.ba1621f1e323ae4da234fcfd4b5d657b.jpg

 

Now let's look at the same toolpath with corner pretreatment turned on:

421563453_CornerPretreatment.thumb.jpg.48a5ba5c0355e5f1f265f37040b8c72e.jpg

 

We stay pretty clean all the way to the center and my tool would be much happier for the last 50% of the path even with very high engagement.

 

I absolutely understand your frustrations and hope this helps with what you can do in the here and now to effect some behavioral changes.

  • Thanks 1
  • Like 3
Link to comment
Share on other sites

Thanks for the response!!   I didn't intend for that to sound like a "MaStErCaM SuCkS!>" post.  It wasn't.  One of Mastercam's flaws just happens to be something I have to work around often.  I hadn't thought about using the corner pre-treatment to combat this.  I usually end up dynamic milling from outside, to more square shaped avoidance region, then peel milling that square shaped area.  

 

I don't remember how Esprit handled this situation, but I don't recall running into it as much.  Either way, I'm sure there's far more to the problem than meets the eye.  I just needed to rant a bit this monday morning!  😂

  • Like 1
Link to comment
Share on other sites
32 minutes ago, Chally72 said:

This is a problem inherent in every package that uses dynamic-style motion

 

18 minutes ago, JB7280 said:

I don't remember how Esprit handled this situation, but I don't recall running into it as much

Esprit does handle it better so technically not inherent in every package. Just depends on how you attack the stock initially and what you prioritize in the toolpath. You can make Mastercam behave like Esprits with geometry manipulation so its not like we're missing anything in Mastercam, the user just needs to be aware of the problem and know the solution before breaking a bunch of endmills. 

It would be nice to have a toggle in Dynamic to have it use an alternative strategy to reduce the risk of slivers. This could be either an initial starting strategy (recognizing geometry that leads to slivers and changing the initial cut pattern) or an ending strategy where that last several cuts only happen on one side of the sliver. Seems like easy solutions to me but I've never generated a toolpath algorithm so who knows.

And just to be clear, I'm not saying overall that Esprits Profit Milling is better than Mastercam Dynamic. It just handles this scenario better, and better is subjective. Better being no sliver, but there is a cost and that is typically increased cycle time. Overall, I think you will find Dynamic Mill faster on similarly shaped parts due to it wanting to stay in the cut as long as possible but this is what leads to slivers on certain parts. 

 

Link to comment
Share on other sites
17 minutes ago, #Rekd™ said:

Dylan,

Is there anyway to get the software to generate a warning that there might be a sliver that could damage the tool?

 

That'd be tough to pin down for the same reasons it's tough to recognize at a toolpath calc level- and if we can recognize a sliver, why not then go out and fix it? But definitely, we want to recognize this before the user has to take it to Verify and see something that is immediately going to be a problem on the machine.

And yep, referring to just the technology side, the slivers themselves are the problem that anyone doing this style motion needs to overcome. If the software weights or prioritizes the echo shape over staying in-cut, then the risk of slivers is greatly reduced, but the motion is much farther away from true "dynamic" or constant-engagement motion. Since the latter is impossible for a user to "Program in", but the former can be "programmed out", I think starting with Dynamic motion and then tweaking the path to adjust behavior gives the best chance of satisfying all aspects of a cut. But yes, that doesn't assuage me much when I'm sitting there trying to path around that sliver that just snapped my tool!

Link to comment
Share on other sites
3 minutes ago, Chally72 said:

That'd be tough to pin down for the same reasons it's tough to recognize at a toolpath calc level- and if we can recognize a sliver, why not then go out and fix it?

Possibly then it can be recognized at the verify level if it is hard to do at the calculation level? I know this is likely stretching the limits of non-kinematic awareness. 

Link to comment
Share on other sites
2 hours ago, Chally72 said:

That'd be tough to pin down for the same reasons it's tough to recognize at a toolpath calc level- and if we can recognize a sliver, why not then go out and fix it? But definitely, we want to recognize this before the user has to take it to Verify and see something that is immediately going to be a problem on the machine.

And yep, referring to just the technology side, the slivers themselves are the problem that anyone doing this style motion needs to overcome. If the software weights or prioritizes the echo shape over staying in-cut, then the risk of slivers is greatly reduced, but the motion is much farther away from true "dynamic" or constant-engagement motion. Since the latter is impossible for a user to "Program in", but the former can be "programmed out", I think starting with Dynamic motion and then tweaking the path to adjust behavior gives the best chance of satisfying all aspects of a cut. But yes, that doesn't assuage me much when I'm sitting there trying to path around that sliver that just snapped my tool!

Well it can be controlled from the programming stand point and slivers always eliminated when cutting from the outside. We have to get into thinking we will always get them verses rolling the dice thinking we may or may not get them. Make islands where we are doing this work and the middle slivers will disappear 100% of the time.

Throw an island using 60% of the tool on that square shape. Cut to the island then ramp off the island with a highfeed cutter. Trust me will be efficient and cost effective all day long. Better yet try to cut with a highfeed cutter using 60-70% step over and then compare that on hard metals to a full flute cut where you have to use much smaller step over. Compare the cost of the inserts to the cost of the solid carbide endmill. Compare the time difference and the run time and green to green time over 100, 200 or even 1000 parts and I think everyone would see where the true winner is. Not found anyone willing to do it, but seen enough over thee last few years to make me think about what is going to be best approach to get away from the nagging issue of slivers.

The problem is the cutting motion from inside to outside slivers. The algorithm doesn't look all the way through volumetric thinness of a section throughout the cut to evaluate the cut itself. The math is written in such a way to take one shape at the start and one shape at the end and get from point A to point B. Material hardness and characteristics are not taken into account with any of these process. If they were then Material that created these events would use different method to do the toolpaths. I can sit down and draw the correct motion in hard materials to not create slivers, but still machine them dynamically. I think it would take 100 to 1000 more amount of time, but if my little old brain can see the process as many others can then it is a limitation of the input data into the algorithm. The thinness of any cross section is the real enemy here and how to fold or bend the material with the cutting process is the work. When coming into a shape just little changes to the boundary can create no silvers. Been a while since I have had to chase slivers down in hard metals, but just little changes make all the difference to reduce the silvers.

The last part of the process is tool wear stepping up when stepping down. I have made 5 step downs with a tool and never stepped up. I make sure I am getting as much full flute engagement when in hard metals before ever stepping up. I am also using different methods to take care of the step. I will do the Step Downs never doing the step up and make a stock model. I then will step down using a highfeed cutter meant to take that kind of abuse. My Solid Carbide endmill is doing the heavy lifting of getting the majority of the meat off wearing the tool evenly like I want it to be worn and the highfeed cutter is wearing that small cut area it is mean to be used at and tools are lasting 10-20 times longer doing what each is meant to do. Stepping up with a Solid Carbide in hard metal starts the wear process using smaller and smaller parts of the flute. I stumbled onto this by accident, but once I stopped and rethought the process it made total sense. I have asked for this ability in OPTI-Rough to step all the way down with full cuts then step back up. I since adjusted that thought process so do full cuts make a stock model and then work from there and like the results in hard metals and not thin shapes.

Thin shaped parts then use the traditional step down full cut and then step up. That is where just plain old experience and adjusting for each part, material, machine and other things has to be part of the programming.

Sorry no easy button process and still learning and and adapting my process to what each new challenge presents.

  • Like 8
Link to comment
Share on other sites
1 hour ago, crazy^millman said:

Well it can be controlled from the programming stand point and slivers always eliminated when cutting from the outside. We have to get into thinking we will always get them verses rolling the dice thinking we may or may not get them. Make islands where we are doing this work and the middle slivers will disappear 100% of the time.

Throw an island using 60% of the tool on that square shape. Cut to the island then ramp off the island with a highfeed cutter. Trust me will be efficient and cost effective all day long. Better yet try to cut with a highfeed cutter using 60-70% step over and then compare that on hard metals to a full flute cut where you have to use much smaller step over. Compare the cost of the inserts to the cost of the solid carbide endmill. Compare the time difference and the run time and green to green time over 100, 200 or even 1000 parts and I think everyone would see where the true winner is. Not found anyone willing to do it, but seen enough over thee last few years to make me think about what is going to be best approach to get away from the nagging issue of slivers.

The problem is the cutting motion from inside to outside slivers. The algorithm doesn't look all the way through volumetric thinness of a section throughout the cut to evaluate the cut itself. The math is written in such a way to take one shape at the start and one shape at the end and get from point A to point B. Material hardness and characteristics are not taken into account with any of these process. If they were then Material that created these events would use different method to do the toolpaths. I can sit down and draw the correct motion in hard materials to not create slivers, but still machine them dynamically. I think it would take 100 to 1000 more amount of time, but if my little old brain can see the process as many others can then it is a limitation of the input data into the algorithm. The thinness of any cross section is the real enemy here and how to fold or bend the material with the cutting process is the work. When coming into a shape just little changes to the boundary can create no silvers. Been a while since I have had to chase slivers down in hard metals, but just little changes make all the difference to reduce the silvers.

The last part of the process is tool wear stepping up when stepping down. I have made 5 step downs with a tool and never stepped up. I make sure I am getting as much full flute engagement when in hard metals before every stepping up. I am also using different methods to take care of the step. I will do the Step Downs never doing the step up and make a stock model. I then will step down using a highfeed cutter meant to take that kind of abuse. My Solid Carbide endmill is doing the heavy lifting of getting the majority of the meat off wearing the tool evenly like I want it to be worn and the highfeed cutter is wearing that small cut area it is mean to be used at and tools are lasting 10-20 times longer doing what each is meant to do. Stepping up with a Solid Carbide in hard metal starts the wear process using smaller and smaller parts of the flute. I stumbled onto this by accident, but once I stopped and rethought the process it made total sense. I have asked for this ability in OPTI-Rough to step all the way down with full cuts then step back up. I since adjusted that thought process so do full cuts make a stock model and then work from there and like the results in hard metals and not thin shapes.

Thin shaped parts then use the traditional step down full cut and then step up. That is where just plain old experience and adjusting for each part, material, machine and other things has to be part of the programming.

Sorry no easy button process and still learning and and adapting my process to what each new challenge presents.

We really need a "Love" reaction for Posts like this. ❤️ For when expressing our thoughts with a "Like" is simply not enough.

  • Like 1
Link to comment
Share on other sites
3 hours ago, Chally72 said:

That'd be tough to pin down for the same reasons it's tough to recognize at a toolpath calc level- and if we can recognize a sliver, why not then go out and fix it? But definitely, we want to recognize this before the user has to take it to Verify and see something that is immediately going to be a problem on the machine.

And yep, referring to just the technology side, the slivers themselves are the problem that anyone doing this style motion needs to overcome. If the software weights or prioritizes the echo shape over staying in-cut, then the risk of slivers is greatly reduced, but the motion is much farther away from true "dynamic" or constant-engagement motion. Since the latter is impossible for a user to "Program in", but the former can be "programmed out", I think starting with Dynamic motion and then tweaking the path to adjust behavior gives the best chance of satisfying all aspects of a cut. But yes, that doesn't assuage me much when I'm sitting there trying to path around that sliver that just snapped my tool!

Something about your explanation made me think, what about some sort of a slider, or a percentage, where you can choose to prioritize staying in-cut vs the echo shape, this way, I would think it would still be the end user making the determination, rather than the software. 

 

I think this topic has brought out almost all the eMC "big dogs"!!  Only missing a couple!!

  • Like 1
Link to comment
Share on other sites
14 minutes ago, JB7280 said:

Something about your explanation made me think, what about some sort of a slider, or a percentage, where you can choose to prioritize staying in-cut vs the echo shape, this way, I would think it would still be the end user making the determination, rather than the software. 

 

I think this topic has brought out almost all the eMC "big dogs"!!  Only missing a couple!!

While sliders in the toolpath side of life are generally avoided for a number of reasons, it's a good thought assuming that you have a toolpath that you can make react in a proper manner to the changes, and I can tell you that a lot of the ideas here are very similar to things that have been put up for discussion on this topic. This isn't my area though, and I'm just commenting from a user perspective on what we use and how we teach to get around this.

Link to comment
Share on other sites
15 hours ago, AHarrison1 said:

uhm... I think he just called you all fat...😮

Haha, I'm trying hard to follow in their footsteps!!

15 hours ago, Chally72 said:

While sliders in the toolpath side of life are generally avoided for a number of reasons, it's a good thought assuming that you have a toolpath that you can make react in a proper manner to the changes, and I can tell you that a lot of the ideas here are very similar to things that have been put up for discussion on this topic. This isn't my area though, and I'm just commenting from a user perspective on what we use and how we teach to get around this.

I'm very far from a software developer, so I'm sure there's plenty of reasons against it.

Link to comment
Share on other sites
7 hours ago, cncappsjames said:

I call those tuition payments.

You can't replace real life experience with any amount of training.....I feel like I am lacking in certain areas after reading posts here daily! 

Big thumbs up to all those that share their experience and knowledge here freely!!!

  • Like 5
Link to comment
Share on other sites
14 hours ago, cncappsjames said:

When doing dynamic paths and I anticipate slivers that will be problematic.  Typically I'll do what @crazy^millman does mostly... i figure out where the problem areas will be with verify,, create a boundary,, then return to that area with a ramping toolpath. Only took grenading a 1" solid carbide endmill to learn that lesson. 😬

I call those tuition payments. A statement I stole from Pete Roberts CEO of Origin Maine when he spoke of mistakes made in business calling them tuition payments.

These situations are not unique to Mastercam contrary to popular belief. Part topology, approach, stepover, etc... all play a part in the Ultimate outcome and success. 

Or in others   words  You took The CRASH COURSE   Like everyone else here has done from time to time . IE  School Of Hard Knocks, Then Handling School   so you can   practice what you learned in the School Of Hard Knocks .  :)  Good Stuff Guys !!! Thanks :gathering:

  • Like 2
Link to comment
Share on other sites
On 12/13/2021 at 8:20 AM, JB7280 said:

Sometimes I think that Mastercam is in cahoots with endmill manufacturers to leave slivers and towers on dynamic toolpaths just to try to break tools.  There's some evil guy behind the scenes that changes the code before every new MC release, so that MC intentionally leaves a big sliver in the middle of every dynamic from outside toolpath.  

 

I know every software has its highs and lows, but I don't remember ever working around this problem in the other CAM packages I've used.  

 

End of my rant.  Thanks for listening, enjoy the rest of your day.

I haven't read the entire thread so this might be a repeat, but one of the things I will do in ill-conditioned geometry is draw a circle or square in the middle and use that as avoidance, then go back with a second dynamic toolpath (dynamic face) to remove that circle or square.  It avoids creating the sliver and makes for very reliable programs.  We are running titanium parts overnight on our cell so this approach has been a must at times.

 

  • Like 3
Link to comment
Share on other sites
10 hours ago, Bob W. said:

I haven't read the entire thread so this might be a repeat, but one of the things I will do in ill-conditioned geometry is draw a circle or square in the middle and use that as avoidance, then go back with a second dynamic toolpath (dynamic face) to remove that circle or square.  It avoids creating the sliver and makes for very reliable programs.  We are running titanium parts overnight on our cell so this approach has been a must at times.

 

I was lucky with 90% aluminium parts 😀 but this is the process I have used for stainless parts.

Create a strong enough island (5mm wide section) and then I would ramp that with a 2D path back and forth on centreline using the bottom of the cutter. 

Edit: As a mcam solution, I always wondered if the software could look at the cusp deciding to leave the "solid island of suitable strength", then top down machine that off in a straight 2D centreline path by automatically splitting up the depth into say 4 (user definable to not overload the cutter with a S + F override box too, as this would need changing as non dynamic).

In this day and age, it should be possible to do LLLMMMAAAOOOO (how I hate that saying :lol:)

  • Like 1
Link to comment
Share on other sites
31 minutes ago, Chally72 said:

This is a pleasing pattern to the eye, but is not maintaining a constant cutter engagement throughout the path.

I didn't tweak it, just a quick example of what 5 axis paths can do for you. Adjust and play with settings and lie a little on contour selection, that will change the complete path results but you will not get slivers, (until you clean your table)

Link to comment
Share on other sites
13 hours ago, Bob W. said:

I haven't read the entire thread so this might be a repeat, but one of the things I will do in ill-conditioned geometry is draw a circle or square in the middle and use that as avoidance, then go back with a second dynamic toolpath (dynamic face) to remove that circle or square.  It avoids creating the sliver and makes for very reliable programs.  We are running titanium parts overnight on our cell so this approach has been a must at times.

 

That's basically what I do, so if nothing else, it's reassuring to see that my solution is the same as others.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.

Join us!

eMastercam - your online source for all things Mastercam.

Together, we are the strongest Mastercam community on the web with over 56,000 members, and our online store offers a wide selection of training materials for all applications and skill levels.

Follow us

×
×
  • Create New...