CAD/CAM discussion forum > 3D CAD/CAM > Efficient milling and the new variable step Z Level

Efficient milling and the new variable step Z Level

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 1 of 13

 Efficient milling and the new variable step Z Level
06-03-2007 09:29 . am | View his/her posts only
Hi guys,

I wrote a small paper about how a customer part can be milled as efficient as possible. He allowed me to post it here for your attention.

Enjoy,
Dan

Rank: 1

OldForumPost

Newbie

posts: 0

Registered: 2012-1-14

Message 2 of 13

06-03-2007 10:35 . am | View his/her posts only

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 3 of 13

06-03-2007 10:44 . am | View his/her posts only
I know you like these variable step cycles Mike. As a matter of fact this is the reason QM confuses new users having to lace finishing cycles. New users doesn't know the classic variable step lace is dedicated to you.

Rank: 1

Katsuyasu

Newbie

posts: 0

Registered: 2004-3-30

Message 4 of 13

19-03-2007 11:28 . pm | View his/her posts only
Hi Dan and guys.

This is an actual our customers request about Zlevel interpolation tool paths.
I would like to know everybody's opinion about this request. If you have any question,
please let me know.

Thank you,

Yasu

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 5 of 13

20-03-2007 07:03 . am | View his/her posts only
Hi Yasu,

In your file, in the first case you angular limit (AL) is 45-90 and after this you add uniform cuts. Uniform cuts are intended to mill shallow areas or AL = 0-45 that were disabled so pretty much AL defeats the scope of uniform cuts.

In the collapsing Offset2D your rightful comment that a natural bigger cusp occurs when collapsing sharp angles. I always encourage people to Parallel lace instead of Offset2D for finishing. Only in cases that Climb-Conventional (CC) is required Offset2D should be chosen! And in HSM finishing when you have a surface thick = [0.05-0.5]mm and you plan to sand blast it after CC doesn't make much sense so lace can be used. If is a semifinish cycle (micro-cusp variations aren't relevant) and tool engagement gets critical (CC matters) then Offset2D is the choice. So in majority of cases, I believe, isn't a really problem if the optimal paths are chosen.

To conclude:
Rough - Lace (if you can ZigZag) or Offset2D if (is required Climb or Conventional).
Semifinish - ZLevel AL = 45-90, HSM Offset2D AL = 0-45, SurfaceThick = [0.05-0.5] Tolerance = [0.01-0.1], CuspHeight = [0.01-0.1]
Finish - ZLevel AL = 45-90, HSM Lace AL = 0-45, SurfaceThick = [0.05-0.5] Tolerance = [0.01-0.1], CuspHeight = [0.01-0.1]

Other benefits of using lace instead of Offset2D/3D is:
- Lace doesn't propagate all the wobbles in the containment.
- Is 20% shorter NC file. Can get maybe 30% shorter if you align it with X or Y axis and on some machines this can add extra feed speed increase as well.
- In general 20% faster to mill then corresponding Offset2D/3D cycle.

I'm curios about some hands on guys opinion too. Maybe we need to do add some new ways of milling in QM that are missing and I don't know them.

Regards,
Dan

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 6 of 13

20-03-2007 07:39 . am | View his/her posts only
Some screen captures from our latest Optimal Variable Z Level with uniform cuts in version 13. It has more retracts but it aims to achieve equal cusp in a constrained top to bottom + climb conventional manner on complex generic parts in only one operation.

Enjoy

Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 7 of 13

21-03-2007 07:31 . am | View his/her posts only
The optimal variable z operation looks good. When can we get our hands on it?

It would be nice to get some better documentation on operations, maybe even sample files.

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 8 of 13

21-03-2007 07:59 . am | View his/her posts only
In first incarnation of version 13. Sometimes in the summer/autumn.

Rank: 1

Katsuyasu

Newbie

posts: 0

Registered: 2004-3-30

Message 9 of 13

02-04-2007 06:01 . am | View his/her posts only
I also think the optimal variable Z operation looks good. However, I am a little bit worried.
Because these shapes are symmetrical. I doubt whether it keeps Z level constantly to
unsymmetrical shapes.

Yasu

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 10 of 13

02-04-2007 06:40 . am | View his/her posts only
In the latest version it will chop redundant spans out and will keep only the one that have meaningful contribution to the overall cusp this will solve arbitrary asymmetrical parts. I think it looks ok for now. After V13 official we will have user feedback and I believe until the end of the year will be a pretty good milling technique.

Dan

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 11 of 13

02-04-2007 08:29 . am | View his/her posts only
All this pretty obsolete, specialized, variable Z level doesn't make me happy at all: it is long to compute and leads, I believe, often to suboptimal and uneven distributed cusp height when compared with other traditional ways of milling at constant cusp height.

If a toolpath (any toolpath Z Level included) is required to mill to equal cusp the generic Add Uniform Cuts, quite unique to QM as a broad technique, should be the choice. We spent years to perfect this handy feature and generalize it over all none Offset3D cycles giving for any toolpath optimal cusp height property with a flick of a switch and believe me this wasnt easy to do.

I had many discussion with users and distributors about the old, suboptimal, variable step technique that they where used to in other high end CAM systems and why the generic add uniform cuts it is always more efficient, pretty much roughly achieving Offset3D efficiency still keeping the main, predictable, flow of the initial path incontrollable in offset 3D when complex deep pockets are milled. In general my rational reasoning prevailed in many cases and when asked nobody could came with real time cases that are suboptimal milled with Uniform Cuts in only one toolpath.

Unfortunately my subjective, rational reasoning doesn't suffice and The User is King! principle prevails in commercial environments. So this is the reason why we have variable Lace and Z Level in QM that doesn't make me too happy as a purist optimal milling scientist.

I'm still glad that Yasu and George find it useful (so I don't blame me too hard because I waste precious development time ) but I will ask them why is better then constant Z level with Uniform Cuts?

Dan

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 12 of 13

02-04-2007 08:40 . am | View his/her posts only
Quote

User is King


Of course we are

But, in any case. I've shown you how in some instances, variable Z level with cusp height control is very important, and the ONLY way to cut some thin wall parts with good finish. Or cut thin walled electrodes without breaking them.

Rank: 1

ChrisWard2k2

Newbie

posts: 2

Registered: 2011-11-22

Message 13 of 13

03-04-2007 07:08 . am | View his/her posts only
Apparently results in extended tool life (HSS tools)
See also