CAD/CAM discussion forum > 3D CAD/CAM > yes cusp, no scallop?

yes cusp, no scallop?

    
  Subscribe Topic

Rank: 1

Nick

Newbie

posts: 0

Registered: 2007-1-2

Message 1 of 19

 yes cusp, no scallop?
04-04-2008 01:45 . pm   |   View his/her posts only
Hello,

I was just wondering ... when running a 3d toolpath that asks for a cusp height ... I suppose cusp is a different term than scallop?

I think I saw one 3d operation that asked for scallop, but I can't swear to it.

Anyway, I wonder if cusp means VX figures the stepover for the tool on a flat surface and then that's the stepover on the toolpath?

I always took scallop height to mean the toolpath stepover would decrease in complex and/or near vertical areas.

Is this why you run the toolpath at a 45 degree angle sometimes?

Is anything in the works for a scallop type of toolpath?

Thanks,

Nick.

Rank: 1

ChrisWard2k2

Newbie

posts: 2

Registered: 2011-11-22

Message 2 of 19

07-04-2008 04:43 . am   |   View his/her posts only
Hello scallopz

Cusp and scallop are essentially the same thing, a relatively small amount of material that remains after an operation. In my mind, a cusp is a small amount of material remaining in XY, for example material between tool passes by a bull/ball nose tool. A scallop is a cusp including "steps" between passes in Z. I don't know if there is a formal definition of either term, nor if a definition can be consistently applied across all of VX's operations. If cusp height or scallop height are asked for in relation to the tool step-over, your input is the maximum height of the material allowed to remain, as a percentage of the tool diameter. A subsequent operation will remove the cusps (or scallops). In the 3x Nurbs operations you will find that VX has an operation that specifically calculates a toolpath to remove scallops left by a previous, referenced operation (i.e. a side cut that leaves "steps").

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 3 of 19

07-04-2008 09:14 . am   |   View his/her posts only
But unfortunately, in QM, it doesn't seem to always work as advertised on a 'sidecut' operation. Take for instance, this part...





Much of the part is perfectly straight walls. But as you can see, the bosses, and one back wall has some decent amount of draft/taper to it.
I'd REALLY like to get an optimized Z level to work on this part, with a .0002" cusp height, while being able to take as much as .06" D.O.C. on the straight walls.
I just can't get it to work well, without many multiple toolpaths.

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 4 of 19

07-04-2008 09:57 . am   |   View his/her posts only
Hi Guys,

I can't see the images

>>But unfortunately, in VXQM, it doesn't seem to always work as advertised on a 'sidecut' operation. Take for instance, this part...

In QM all toolpaths calculates the step according with their primary usage scenario. So:
1. If is a projective path (Lace, Offset2D, Flow2D) the steps are calculated based on a horizontal step because this paths are primary intended to mill horizontal places.
2. If is a Z Level path with constant step then the step is calculated from the vertical step.
3. If is a 3D path (Offset3D, Flow3D, Pencil drive curve, drive curve with step) then the step is genuine calculated variable in 3D.

All this for a very simple reason VXQM tries to mill optimal and by default in "wrongly" selected paths will undermill instead of overmill. QM doesn't know if the not-so-suitable places will not be milled later in an optimal manner so it generates path in the optimal manner is always a good strategy. If the paths are properly selected for the kind of curvature present then the desired cusp should be achieved in an optimal manner.

This is the "advertised" step calculation in VXQM Steve

In latest VXQM's it is possible to set up an advanced generic and optimal path using Z Level path - this is optimal variabil Z Level + equal cusp. This a powerful path for deep, complex molds in only one opperation(!) that guarantees that the toolpath mills:
- always top-bottom;
- per region optimized;
- achieves equal cusp everywhere;
- doesn't generate redundant cuts;

I'm very curious to show me another path generated by any other product that can do something similar. If you guys are interested I can post other screen captures as well or if you'll like just set the ZMin step in ZLevel to = 0.01 and enable AddUniform cuts and you'll achieve similar results.

regards

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 5 of 19

07-04-2008 10:22 . am   |   View his/her posts only
Sorry you couldn't see the screen shots, try this:
HERE

Okay... I couldn't get Z Level + equal cusp to work I know I've seen your screen shots that look good. And I was excited to try and use this toolpath. But unfortunately, I just couldn't get it to work in an efficient manor for this application.

However, I absolutely love the new smooth flow roughing techniques.


The part in question is attached.

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 6 of 19

07-04-2008 11:31 . am   |   View his/her posts only
SmoothFlow is a horrible complex to code piece of code. It easy doubled my white hair and half it overall At beginning I hate it because if you enable climb or conventional it will generate paths that are 10-20% longer and I'm mad with efficiency so this penalty I felt is unbearable and is a to big price to pay for ultimate tool engagement control. But cutting more and more pockets with it people start to send kudos and looks like they enjoy the angular controlled tool engagement and the small overhead doesnt bother them too much. However the zigzag flavor of SmoothFlow is absolutely awesome ultimately optimal + angle engagement control so that's the think I really like and I believe is unique in QM.

At least now are so many ways of laying out a pattern inside a pocket that I believe every each of you can found his way.

Ok I took your part and generate a ZLevel + Add uniform cuts. Because you didn't set up any path or assert to use any tool I chose 1/4 ball that looks reasonable. I used limiting with surfaces (another new feature in the new QM) to limit the part. I believe I know what bothers you: Why on the tapered side isn't denser path? Is because the step of static ZLevel is calculated for vertical walls so the 70 deg inclined isn't milled dense enough. A quick recommendation is set the cusp height smaller with 20% the what u really want and you'll program the path in VX/VXQM in 1/3 of time and you'll spend extra 10% at machine that's kind of a decent deal for this kind of production.

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 7 of 19

08-04-2008 09:31 . am   |   View his/her posts only
Quote

Originally posted by: billator
SmoothFlow is a horrible complex to code piece of code. It easy doubled my white hair and half it overall At beginning I hate it because if you enable climb or conventional it will generate paths that are 10-20% longer and I'm mad with efficiency so this penalty I felt is unbearable and is a to big price to pay for ultimate tool engagement control. But cutting more and more pockets with it people start to send kudos and looks like they enjoy the angular controlled tool engagement and the small overhead doesnt bother them too much. However the zigzag flavor of SmoothFlow is absolutely awesome ultimately optimal + angle engagement control so that's the think I really like and I believe is unique in QM.



See, sometimes us customers know what we want Now gimme a true Z level with cusp height control so I can dump Powermill No, it's not nearly the most efficient way of machining, but sometimes, when it comes to thin wall graphite, it's the ONLY way.

Make it so there is an option to cut the cusp height in one shot, or to do a constant Z, and come back and machine off the cusps. Yeah, I know... Easy for me to say, eh?

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 8 of 19

08-04-2008 02:09 . pm   |   View his/her posts only
Quote

See, sometimes us customers


Yah I agree with you only sometimes


Quote

Now gimme a true Z level with cusp height control so I can dump Powermill


I have a feeling of d

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 9 of 19

08-04-2008 02:09 . pm   |   View his/her posts only
Quote

See, sometimes us customers


Yah I agree with you only sometimes


Quote

Now gimme a true Z level with cusp height control so I can dump Powermill


I have a feeling of d

Rank: 1

cutter

Newbie

posts: 56

Registered: 2011-11-23

Message 10 of 19

09-04-2008 05:51 . pm   |   View his/her posts only
Billator, I was following this thread and tried Z level and got no where with this part. It would do z only and no roughing. Rough smoothflow yeilds an excellent path but it seems to me in the past I could use z level for more than I can now. How would you cut this part?

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 11 of 19

10-04-2008 07:46 . am   |   View his/her posts only
Cutter I can't open your part so please check again. You need to submit first and after this attach it (I don't want to sound rude )

In theory, I should ask our support guy to support you but if isn't that much trouble I will try to lay out some paths for you. I'm representing the development so usually I answer to lack of features / competitive disadvantages kind of issues.

We should not use our forum cluttering it with big files (the forum software doesn't quite like this kind of usage scenarios) and is always better to use support + PCRs if are problems.

We struggle to timely answer to support questions and we have official processes that enforces us to take actions in this regard so a support issue if can't be answered satisfactory will be elevated to PCR level and this will be officially scheduled and tracked. Posts on the forum don't have that "official" enforcement on us. However looks like our support guys are daily scanning and timely answer to your questions on the forum and many of the requests raised on the forum are converted in PCR.

Rank: 1

cutter

Newbie

posts: 56

Registered: 2011-11-23

Message 12 of 19

11-04-2008 10:56 . am   |   View his/her posts only
Billator, this was not a serious issue so dont worry about it. We have a new guy in Cam support and he is still getting up to speed and that is why I thought I would ask here since you were responding to posts. Thanks, Cutter

Rank: 1

Robert

Newbie

posts: 0

Registered: 2011-11-22

Message 13 of 19

15-04-2008 03:46 . pm   |   View his/her posts only
Hi Steve,

Why do you love Smoothflow (I meanthe machining routine not the British beer)



Rank: 1

Nick

Newbie

posts: 0

Registered: 2007-1-2

Message 14 of 19

21-04-2008 02:23 . pm   |   View his/her posts only
Hi Billator,

Sorry I took so long, pushing that ball up the hill, you know.

Maybe I can get this screen shot on here.

It's the top view of a planar (lace) cut with a .005" scallop height.

The toolpaths are squeezed closer together the more steep the area is.



thanks,

Nick.

Rank: 1

Robert

Newbie

posts: 0

Registered: 2011-11-22

Message 15 of 19

21-04-2008 02:50 . pm   |   View his/her posts only
No pics

Rank: 1

ChrisWard2k2

Newbie

posts: 2

Registered: 2011-11-22

Message 16 of 19

21-04-2008 07:07 . pm   |   View his/her posts only
...they are viewable in FireFox

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 17 of 19

05-05-2008 11:50 . am   |   View his/her posts only
scallopz,

We have the lace cycle that does that kind of inefficient and particularized path. Supplementary we developed a generic way to mill efficient at a given cusp height (constant scallopz ) along any kind of features called "Flow3D". This should better address this way of milling.




Rank: 1

Nick

Newbie

posts: 0

Registered: 2007-1-2

Message 18 of 19

09-05-2008 12:26 . pm   |   View his/her posts only
Hi Billator.

I've been looking in the help files about 3D toolpaths and it seems that the bulk of them are first calculated and then

projected onto the cam geometry. What I'm looking is a 3D toolpath stepover that is variable according to the steepness

that will be encountered on that pass, using something like a scallop sample distance to determine the amount the

stepover will be. The tool might step over .027" in one area and step over .005" in another area, not a constant

stepover during the whole toolpath. I haven't found any other toolpaths that will give a good enough finish (unless

the stepover is miniscule and the time is prohibitive).

thanks,

Nick.

PS - I sent this to Ray G. also, maybe he'll have some ideas.

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 19 of 19

16-05-2008 11:30 . am   |   View his/her posts only
After my knowledge QM is the product that has the most paths that mills at constant cusp optimal.

It has many techniques to achieve that:
Genuine Offset3D:
- With or without profiles to mill shallow and simple parts;
- Along a profile drivecurve with steps;
- Along 2 profiles Flow3D;

For deep complex parts ZLevel with variable steps and add uniform cuts. To supplementary control overall flowing top to bottom.

For shallow complex molds projective paths: Offset2D, Lace, Flow2D + add uniform cuts.

All this techniques were designed to achieve the same thing: constant cusp as optimal as possible and still extra constrain the path to flow in an specific way and don't be longer then 5% compared with the best one: Offset3D.

Is a priority In QM to offer optimal milling at constant cusp. My dilemma is how you didn't find paths that achieve constant cusp in QM? Please contact our support for extra-clariffications.
See also
X