2D Dynamic Question - X9

[barrybaker]

Material is 15-5PH 

Heat Treated to H1025

 

I am questioning that chunk of stock that the path removes at the completion of each boss.

 

Looks like the cutter might "hate" it,  takes that in two cuts.

 

It would be ideal if both sides were more balanced as a continuous flow.

 

I'd appreciate a second set of eyes on this.

 

Barry in St Louis

 

 

For the Forum.mcx-9

 

[David Colin]

Yes, this is how dynamic toolpaths manage that and i don't like that too... i usually try to avoid that increasing 'min toolpath radius' and add another toolpath to eat stock left.

 

BTW, in dynamic toolpaths i don't like the way they crunch sharp corner on first cut. In your example, stepover is 0.035 but first cut is cutting 0.1 on corners.

[crazy^millman]

I would start with a stock model and not use 2D HST since it does not use rest material. I would create geometry to make a peel mill toolpath between each part. I would then use a 2nd Stock model that uses the Peel mill toolpaths to remove the stock cut. (Note not Dynamic but regular peel. There is a drop down on this toolpath most people over look.)I would then use a Opti Rough as the Stock model as my Ref Material and life should be good.

 

Yes on your part not a good way to get around it. Where as much as the software gets better the more it just can't see everything. Art is as much what we do as the science of what we do.

 

Also you chip load per tooth is to small. I would kick it up to .006 per tooth with that small of a step over.

[barrybaker]

Thanks a bunch to both of you. I am well on my way now. Kinda long way around the block...

 

David, how were you able to pull off that measurement snip. You must be using the classic backplot / verify? I can never really get the new one to measure well.

 

Ron, I'll try that .006 / flute, but sounds a little aggressive. You knew it was 15-5 at H1025, right?

 

Thank again guys,

Barry

[crazy^millman]

Thanks a bunch to both of you. I am well on my way now. Kinda long way around the block...

 

David, how were you able to pull off that measurement snip. You must be using the classic backplot / verify? I can never really get the new one to measure well.

 

Ron, I'll try that .006 / flute, but sounds a little aggressive. You knew it was 15-5 at H1025, right?

 

Thank again guys,

Barry

 

Yes I am not called the crazy^millman for no reason. Run it till it breaks and back it off 10%.

[David Colin]

You can't get a direct measurement... I had to draw stock wireframe then 'save toolpath as geometry' in classic backplot then offset chain with tool radius value finally i could make a dimension.

 

BTW, investigating about this dimension (0.106 on my pic), even if stepover condition is not fulfilled, cutting arc angle looks pretty constant.

[DPARKS]

This worried me too because the thin wall condition and roll around the sharp corners will not work if it were Titanium.  15-5 is not much easier but different with the reflection of heat like titanium. this is a great configuration for an approach I used for Ti parts where the thin wall that gets machined away just can't be allowed. I would use the dynamic tool path to machine the square shape down to a rectanglular shape with full radii on both ends equal to an offset amount of <some distance? from the boss radii of the pieces you are cutting. then simply helix in between each part boss to final depth staying off the inner sides of each boss about the same as the stock left from the dynamic did before. then helix each boss. material will not get thin and flex away from tool get hot and work harden causing premature wear if it does not grab during break through and break all together. 

 

I use optirough and stock model more than anything else, I like 2d dynamic a lot too. both sure help on material like this still have to get creative in some situations where other material allow more freedom than do 15-5.

[barrybaker]

DParks,

That's a great approach. I hadn't thought far enough ahead to be worried about that thin web...

 

Let me try that. Care to share a helix parameter recommendation?

 

Barry in St Louis

[DPARKS]

Well I would look into what is recommended for the tool you are using.  I do like the 5 flute variable helix end mills coated TiAlN or the like made for dynamic mill type machining - some might be better for 15-5 helical type cutting.  We run 15-5 that is HT to around 40Rc <it is pretty touch> and Titanium.  I have been running IMCO M525 for both. Helix I would probably run 450-500SFM and helix around .015 to .030 for helixing a bore unless the hole diameter is less than 20% of tool.

 

Some times my feeds and speeds are way aggressive but other times very conservative a lot depends on setup and I build my processes to be robust to maintain consistency and little intervention as possible

[ahaslam]

I know this post is mostly dead but I did want to mention something about the corners that you see in high speed tool paths. HST is not calculated based on actual step over. Not in a direct sense. If it was, it would look like the standard contour with multi step enabled. What you are seeing around the corners is exactly what you want. Mastercam HST is based off engagement angle of the tool (as is all HST style tool paths) and compensates around corners accordingly. Think of a standard 90 deg corner around the outside of a part. If you turn off the roll around corner function in a standard contour, the end mill will go past the engagement point and then re engage after changing direction. Turn on roll around and you see it maintain engagement around the corner. But the actual angle of engagement of the tool is not maintained, only the step over. In HST, you tell the cam what your ideal width of cut is and then the cam calculates an engagement angle for that step over based on a straight cut and your tool DIA. Something to remember is that when you change direction, your point of nominal engagement changes with it. There for, a change in step over is not equal to/is not the same as a change in engagement angle. It's not as noticeable but you will see the same function on internal corners. The entire concept of HST paths is to quite literally cut corners buy maintaining a functional engagement angle regardless of actual step over. With all this said, if you filter your tool path too heavily, it will take that opportunity to hug the corner and will indeed over shoot the desired engagement angle. Any way, i'll leave you with some reference material so you know I'm not full of crap (about HST  )   

 

How to calculate TEA(tool engagement angle)

 

Why it works

 

Situations it works well in (when to use it)

 

Enjoy!

[Matthew Hajicek - Singularity]

Ahaslam,  I think what they're objecting to is something that I run into once in a while.  For some of my jobs I'm cutting 17-4 H900 with a 3/8" 7-flute, and cut around 1/2" deep with .020" stepover.  This works great except when it leaves a peninsula, which can be 1/2" tall and .020" thick, and then it tries to roll around the end of it before cutting it away.  This peninsula will flex and slap around, sometimes chipping the cutter.

[Goldorak]

Ahaslam,  I think what they're objecting to is something that I run into once in a while.  For some of my jobs I'm cutting 17-4 H900 with a 3/8" 7-flute, and cut around 1/2" deep with .020" stepover. This works great except when it leaves a peninsula, which can be 1/2" tall and .020" thick, and then it tries to roll around the end of it before cutting it away.  This peninsula will flex and slap around, sometimes chipping the cutter.

 

Optirough always handle this problem or any 2d complex shape better than 2d HST

[Matthew Hajicek - Singularity]

Optirough with step-up is what I'm using on my parts.

[Goldorak]

this is the result of what I've called  "the skin bug"

the thin skin wrapped around the flutes and when it tries to cut into the next wall at 5000 rpm and 200ipm the end mill simply explode.... 

 

[Matthew Hajicek - Singularity]

I like that little disk...

 

I bet there's a PHD in analyzing that fracture pattern.