Unified Curve Parallel Help! MC2024

[tornero32]

I'm using Unified Curve, Parallel toolpath to finish the radius at the bottom of a pocket. My tool is almost the same radius as the corner radius on the part.

The step over is fine until the tool gets to the corner it gives me a very tight step over. Could please show me how to get rid of the tightness.

I tried advance options for surface Quality, Step over calculation. Nothing helps.

The reason I'm using Curve parallel is because it lets me extend outside the curve and clean the corners.

I do not know if it does that because the surface is not parallel to the curve. any help would be appreciated.

 

 

 

 

ath to finish the radius at the 

Part2.MCAM-CONTENT

[Aaron Eberhard]

Tl;dr - Cut Pattern > Style to Guide instead of Parallel, or, Advanced Options for Surface Quality > Method > Exact instead of Approximate.

----------------------------

The reason for this is the collapse of the pattern due to your tool axis control.

If you turn off collision control and set the Tool Axis Control back to "surface", here's what the "raw" Parallel toolpath looks like:

When you turn on Tool Axis Control set to Fixed Angle to Axis (good call, BTW), it keeps all of these passes, but it's now trying to keep the same surface contact point with the new tool position.  Since the toolpath is set to Center/Tip, it looks worse than it would if you could see the contact point, but that's still a really fine stepover for a fillet at the bottom.

The changes I would make:

Change the pattern type to (2) Guides instead of Parallel.  That'll let the toolpath blend between the top and bottom better and you can benefit from Machining Geometry > Calculation Type > Tool-Center Mode (the default), where it's smart enough to calculate it based on the tool center.  It gets rid of the gouging/fishtailing you have on the corner as well.  When your fillet is almost the same size as the tool like it is in this case, I'd recommend tightening up the Cut Tolerance, maybe .0001 instead of the default .001? 

I'd also change your Collision Control to tilt to avoid the walls unless you have a specific reason to ask it to retract if it gets close to the wall?  I think you'd prefer to have it tilt away.  You probably don't need to use the second collision control.

 

[tornero32]

Thank you, Aaron, for your quick response and the advice I really appreciate.

straight forward explanation and it really works better on the corners, the C axis is not rotating like crazy, and the finish is great.  Tool axis control makes a big difference.

I still struggle a lot with tool axis control and collision control. I was using Curve parallel because it gives me the option to extent to the side of the curve.

 

 

 

Curve parallel gives me that option. i was trying to machine the material on the corner from previous operation.

 

Curve guide doesn't give me that option, I guess it has different options, and I will try to check those options.

 

[Aaron Eberhard]

Ah, sorry, I missed that the side extension was essential.

It's definitely not as easy as the surface-based toolpaths to do it with the geodesic engine, and you don't have a Top/Bottom extension, you can only do a blanket extension.  It's not too hard, but you need a bit of background knowledge to really leverage it.  Skip the next paragraph if you don't care about the mechanics of it

One important note is that I'd recommend using this with a single Guide chain (Parallel equivalent), as having two that it's trying to Morph between can get a little weird sometimes. 

Geodesic engines are based on the mesh (they create a mesh in the background from your selected surfaces at whatever tolerance you specify).  The edges of the mesh are considered a constraint that it has to stay within, using whatever pattern you wanted it to fit.  So, you have two problems to solve:

* One is that your mesh (the fillet) doesn't extend at all

* Two is that your containment is based on the original mesh.  

To solve the first problem, use Cut parameters > Machining Geometry Advanced Params > Extend Mesh:

You can preview the toolpath, but nothing will change at this point (other than taking longer to generate!):

But, in the background, the mesh has been extended on all sides.

So, to solve Problem Two, can be solved using Cut Pattern > Containment, and allowing the tool to go outside the containment:

 

Other than that, I think you'd probably have to create the geometry you want to.  

[#Rekd™]

@Aaron Eberhard You have so much knowledge of these toolpaths and Mastercam. On top of that your explanations are world class!!!

Kudos (also envious)!!!!!

[Kyle F]

14 minutes ago, #Rekd™ said:

@Aaron Eberhard You have so much knowledge of these toolpaths and Mastercam. On top of that your explanations are world class!!!

Kudos (also envious)!!!!!

this was me reading his response

LFG

[Aaron Eberhard]

Thanks guys, just trying my best to make the world a better place and everyone who wants to learn a bit more knowledgeable!

[MrFish]

5 hours ago, Aaron Eberhard said:

Ah, sorry, I missed that the side extension was essential.

It's definitely not as easy as the surface-based toolpaths to do it with the geodesic engine, and you don't have a Top/Bottom extension, you can only do a blanket extension.  It's not too hard, but you need a bit of background knowledge to really leverage it.  Skip the next paragraph if you don't care about the mechanics of it

One important note is that I'd recommend using this with a single Guide chain (Parallel equivalent), as having two that it's trying to Morph between can get a little weird sometimes. 

Geodesic engines are based on the mesh (they create a mesh in the background from your selected surfaces at whatever tolerance you specify).  The edges of the mesh are considered a constraint that it has to stay within, using whatever pattern you wanted it to fit.  So, you have two problems to solve:

* One is that your mesh (the fillet) doesn't extend at all

* Two is that your containment is based on the original mesh.  

To solve the first problem, use Cut parameters > Machining Geometry Advanced Params > Extend Mesh:

You can preview the toolpath, but nothing will change at this point (other than taking longer to generate!):

But, in the background, the mesh has been extended on all sides.

So, to solve Problem Two, can be solved using Cut Pattern > Containment, and allowing the tool to go outside the containment:

 

Other than that, I think you'd probably have to create the geometry you want to.  

Hi Aaron

While your doing an excellent job of educating us on the intricacies of how the processing engines actually handle these tool paths , can you explain why the tool path in your picture above now looks a bit janky in the corner compared to the un-extended one ?

[Chally72]

I don't have Aaron's file, but I'd guess that the "jankiness" is the tilt being applied to the tool/holder to avoid the walls as the tool pivots through the corner. Remember that when you're looking at the blue lines, you're looking at the tip position of the tool, and NOT the actual contact point of the tool flute to the surface. The Collision control method may be modifying the tilt, which will pivot the tool about the center of the ball, not about the tip, and it'll swing the blue lines up and over and make them appear uneven.

If you plotted out the actual contact point (or had a way to view it, which you do in Simulator in 2024 and will in Advanced Toolpath Display in 2025), you'd see perfectly crisp stepovers still. And the examples farther up the thread were in some cases looking at the pattern before collision control was applied, meaning the blue tip still represents the true stepover at this stage.

If you've ever calculated a toolpath using an accelerated finishing tool, this disparity between where the blue lines fall and what the tool is actually doing is even greater, since the tool effective radius centerpoint does not fall along the tool centerline, and small changes in tool flute contact can create major disparities in pass to pass tip position visuals. The "gut check" we get by looking at the blue lines for uniformity becomes less and less valuable as a measure of the quality of the path.

[MrFish]

31 minutes ago, Chally72 said:

I don't have Aaron's file, but I'd guess that the "jankiness" is the tilt being applied to the tool/holder to avoid the walls as the tool pivots through the corner. Remember that when you're looking at the blue lines, you're looking at the tip position of the tool, and NOT the actual contact point of the tool flute to the surface. The Collision control method may be modifying the tilt, which will pivot the tool about the center of the ball, not about the tip, and it'll swing the blue lines up and over and make them appear uneven.

If you plotted out the actual contact point (or had a way to view it, which you do in Simulator in 2024 and will in Advanced Toolpath Display in 2025), you'd see perfectly crisp stepovers still. And the examples farther up the thread were in some cases looking at the pattern before collision control was applied, meaning the blue tip still represents the true stepover at this stage.

If you've ever calculated a toolpath using an accelerated finishing tool, this disparity between where the blue lines fall and what the tool is actually doing is even greater, since the tool effective radius centerpoint does not fall along the tool centerline, and small changes in tool flute contact can create major disparities in pass to pass tip position visuals. The "gut check" we get by looking at the blue lines for uniformity becomes less and less valuable as a measure of the quality of the path.

Thanks Dylan, love learning the intricacies of these tool paths, gives me a far better understanding of what/why I am seeing something in my tool path. I'm a bit old school and like to check the look of the "toolpath" for an idea of potential finish (vs verify), especially in 3d surfacing operations.  Advanced Toolpath Display in 2025 sounds like it will have some nice additions for this.

[Aaron Eberhard]

16 hours ago, MrFish said:

Thanks Dylan, love learning the intricacies of these tool paths, gives me a far better understanding of what/why I am seeing something in my tool path. I'm a bit old school and like to check the look of the "toolpath" for an idea of potential finish (vs verify), especially in 3d surfacing operations.  Advanced Toolpath Display in 2025 sounds like it will have some nice additions for this.

Yeah, Dylan is on the right track there, and the closer the radius is to the tool radius, the more jankiness you're going to encounter, as any minor changes to the tilt will cause the center/tip of the tool (where the blue backplot lines are showing) to visually look worse.

I'm very excited about displaying the contact point in backplot, and I've been enjoying it in Simulator.   For the screenshots above, here's what the contact points look like vs. the backplot:

I could easily do an entire 5 day class just on just Unified/Swarf/Pocket.  I've done a few 3 day classes focused on just Unified and digging into this stuff, and it's always so well received.

[crazy^millman]

Vectors really make a difference here. Control your motion with vectors and the herky jerky motion goes away. Smoothing seems to help control this also.

[Kyle F]

22 hours ago, Chally72 said:

If you plotted out the actual contact point (or had a way to view it, which you do in Simulator in 2024 and will in Advanced Toolpath Display in 2025)

I had no idea contact point could be plotted out as aaron just kindly demonstrated, that is so exciting it will be added to ATD in '25. I feel like a kid in a candy store lol

[Chally72]

Here's my favorite little example of this- a simple 3 axis path with a ball tool traversing across a variable fillet where a blend or morph might be used. If I was looking at this like I usually do, I'd see those blue lines and I might as a programmer instinctually say "This path is no good, I have to go fix this or play around with parameters here."

But, if I'm looking at the green in the image below, (actual tool contact point,) I'm getting a very different perception of the path....IE, leave it alone, it was fine to begin with!

 

[byte]