Community
HSM Beta Testing (English / Deutsch)
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

New morphed spiral pocket roughing

16 REPLIES 16
Reply
Message 1 of 17
fonsecr
4464 Views, 16 Replies

New morphed spiral pocket roughing

We have added a new beta feature for pocket which allows you to do morphed spiral machining for pockets (see picture pocket_morphed_spiral). Note that the feature is available for both 2D and 3D pocket.

By using this feature you can avoid transition moves between the offset passes for conventional pocket roughing (see picture pocket_normal). For conventional offset roughing the engagement will drop to 0 at the transitions. However, by using this new feature you can keep the tool engaged at all time.

Note that the morphed spiral motion is still controlled by the given stepover and that you will likely get more steps than for the conventional offset roughing. If you need to control the true engagement you have to use the Adaptive Clearing strategy.

You need to enable the setting called "Use morphed spiral machining" (see picture pocket_morphed_spiral_setting). It is disabled by default. As always you need to enable the "secret" beta mode to see this setting.

Let us know if you encounter any bad toolpath quality. Thanks.

René Fonseca
Software Architect

16 REPLIES 16
Message 2 of 17
kb9ydn
in reply to: fonsecr

Oooh neat!  I remember using this strategy in Mastercam before it had the dynamic toolpaths (same as adaptive in HSMW).  So here's a silly question: why would you want to use this strategy over adaptive?  Seems like I end up using dynamic for nearly everything except finishing and drilling.  The only time I ever use conventional pocketing is when I'm using the old 2-axis Bridgeport that has a ridiculously small memory size and only likes to run with conversational code.  ::)  The dynamic toolpaths just create too much code for it to handle.

Will this strategy eventually be available in HSMXpress?

C|
Message 3 of 17
Anonymous
in reply to: fonsecr

Will not NC kod (morphed version) much longer than the normal version?
Maybe try it, it looks interesting ...
Message 4 of 17
fonsecr
in reply to: fonsecr

Yes, Adaptive Clearing will generally beat pocket toolpath. However, pocket does have features not supported for Adaptive Clearing and when using insert cutters the depth of cut will be small in which case pocket might be more useful than Adaptive Clearing.

The feature has to go out of beta before we decide whether to include it in HSMXpress or not.

René Fonseca
Software Architect

Message 5 of 17
Anonymous
in reply to: fonsecr

I'm developing the morphed spiral strategy. I would appreciate all kinds of feed-back. The feature is very new, so it might easily happen that somebody discovers something that I have not seen so far. If you have any comments, please write.

Kind regards,
Mikkel
Message 6 of 17
Rob.Lockwood
in reply to: fonsecr

Mikkel wrote:

I'm developing the morphed spiral strategy. I would appreciate all kinds of feed-back. The feature is very new, so it might easily happen that somebody discovers something that I have not seen so far. If you have any comments, please write.

Kind regards,
Mikkel


Cool concept--

Take a look at the attached part-

It contains two duplicate operations, where the only difference is a slight change in stepover; however with very different results.

My assumption is, that this is where the normal pocket routine loses its ability to generate offset passes.

In practice, this algorithm almost seems like the worst of both worlds. It generates a very large program size, that doesnt have equal tool engagement, and appears to have a significantly longer runtime in virtually every scenario I tested.

Am I missing the upside?  I realize the tool will remain in contact with the workpiece all the time, but I fail to see any real benefit from that.

I havent actually posted any code, so I cant attest to any performance effects.


Rob Lockwood
Maker of all the things.
| Oculus | | Locked Tool | | Instagram |

Message 7 of 17
fonsecr
in reply to: fonsecr

Yes, not useful for this part. You need pockets which are fairly circular like in my example.

The important point is that the tool will stay engaged always which isn't the case for normal pocket and Adaptive Clearing. This can be important for some types of material like ceramics.

René Fonseca
Software Architect

Message 8 of 17
Anonymous
in reply to: fonsecr

I tried 3D pocket clearing with morphed spiral on top surface on cuboid. In some combinations of dimensions of block the path is wrong. After a small change of height (e.g. 25.0 mm) the function produce correct path.

Sorry, I can't attach model now, I forgot it at work 🙂 But it should be repeatable by picture (model origin is on left lower front corner, sketch is on Front plane).
Message 9 of 17
fonsecr
in reply to: fonsecr

I tried a lot of variations but could not reproduce it. When you get the chance just send a feedback with your example and we'll get it fixed.

René Fonseca
Software Architect

Message 10 of 17
Anonymous
in reply to: fonsecr

I will send it ASAP 🙂
Message 11 of 17
Rob.Lockwood
in reply to: fonsecr

René Fonseca wrote:

Yes, not useful for this part. You need pockets which are fairly circular like in my example.

The important point is that the tool will stay engaged always which isn't the case for normal pocket and Adaptive Clearing. This can be important for some types of material like ceramics.


Alright, so its probably not currently very useful for me.

Perhaps feedrate optimization could be implemented such that the incredibly low engagement areas are performed at significantly faster feeds? It seems like a toolpath strategy that could benefit immensely due to a controlled feedrate change- After all, it produces zero linking moves.


Rob Lockwood
Maker of all the things.
| Oculus | | Locked Tool | | Instagram |

Message 12 of 17
Anonymous
in reply to: fonsecr

Mikkel wrote:

I'm developing the morphed spiral strategy. I would appreciate all kinds of feed-back. The feature is very new, so it might easily happen that somebody discovers something that I have not seen so far. If you have any comments, please write.
Kind regards,
Mikkel


As I looked at generated path, there are combination of linear and circular  interpolation and path is not depending on settings of parameter Tolerance. But path looks as general curves, approximated by lines and arcs (am I true?). If so, do you suppose that an parameter affecting approximation will be used? I mean especially transition curves connecting lines and arcs to be possible smoothing acceleration (limiting jerk).
Message 13 of 17
fonsecr
in reply to: fonsecr

The morphed spiral motion should generally be smooth so the CNC can keep the desired feed.

René Fonseca
Software Architect

Message 14 of 17
Anonymous
in reply to: fonsecr

I meant not only smooth path, but path with curvature continuity (e.g. spline, used directly for some CNC systems or approximated by circles), to achieve finite jerk (derivative of acceleration) in X/Y components, because it is way to highest possible machining feed.
Message 15 of 17
Anonymous
in reply to: fonsecr

Nothing change with "Morphed spiral pocket" activated or not.
Message 16 of 17
fonsecr
in reply to: fonsecr

Looks like this happens when the stock contours are used in which case we fallback to normal offset passes.

René Fonseca
Software Architect

Message 17 of 17
fonsecr
in reply to: fonsecr

This feature is now enabled in the 2013 R4 development release.

René Fonseca
Software Architect

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report