[jrmach]

I got this tool path with adaptive roughing going on.It is going along real nice,almost like Trochordial milling,then towards the end it decides to stop and go to the other end of the part and do a terrible job.All said and done it leaves a mound of cr_ _ at the end.WTF.Anybody care to take a look,I would appreciate it.

[jrmach]

I think I figured it out.It is trying to reduce cutting air.If the tool path proceeded as nicely as it was all the way thru the last few arcs would partially be cutting air.So it thinks it solution should be to rapid to the other end and finish it from there.
Well if this is the case,I would rather cut the air and have a nice clean tool path all the way thru.
If this is the case I would like to see an option to control this.
I would rather see a nice clean tool path,instead of unnecessary rapiding (jerking) the machine around and leaving a mound of cr_ _ ,just to save 2 seconds.
So maybe an option to control this would be in order.
The work around(always a work around)is to extend model.Not a good option.Or I could use zig-zag cutting option.But doing that results in conventional cuts.That is not good at all.The negates most the benefits of having the adaptive roughing strategy .Especially if you are really doing some hogging.
So if I am correct on what is happening here with the tool path,an option to control direction of the trochodial tool path would be best,with a lead out amount.

[bobcad guy]

i took a look jr, got the same results as you, so i ran it without adaptive rough, just the advanced, and it looks good that way. not sure if you have your heart set on adaptive, but atleast its an out. i find many times i have to settle for option 2, it is wierd that the tool path is going along great, and then...... maybe we can get bobcad to put in a " thats not what the f*%@ i meant " or " really, like that makes sense" button. sorry

[BurrMan]

I think what you are seeing is the toolpath respecting the stock defined as you described. It wont just run around and cut air unless you tell it to. You could allow this in the stock setup by defining an offset for the stock. You may then want to calculate a boundry to control it exactly how you need, so it doesnt then run to the outside area of the part..

[BurrMan]

There are other toolpaths that have the lead in/out overlap control.. They cost a bit more but are powerful. They fall under the multi axis stuff.

[bobcad guy]

hey jr, i had a few minutes to relook at that, i found all kinds of double geometry around the patched area, like every line is double. i took them out, and i also changed to zig zag, instead of just zig, and its damn close, might be another errant string of geometry

[jrmach]

hey jr, i had a few minutes to relook at that, i found all kinds of double geometry around the patched area, like every line is double. i took them out, and i also changed to zig zag, instead of just zig, and its damn close, might be another errant string of geometry

Yeah I know about the doubles,that don't affect anything.The patch is just there temporary so when I turn model to machine next side I can turn patch layer off and have at it.I know zig -zag almost does it right,BUT then we be conventional cutting 1/2 the time.Not good.

[jrmach]

There are other toolpaths that have the lead in/out overlap control.. They cost a bit more but are powerful. They fall under the multi axis stuff.

More money

Are you talking about the multi-axis in the 4th and 5th pro ? ?

[jrmach]

I think what you are seeing is the toolpath respecting the stock defined as you described. It wont just run around and cut air unless you tell it to. You could allow this in the stock setup by defining an offset for the stock. You may then want to calculate a boundry to control it exactly how you need, so it doesnt then run to the outside area of the part..

Yeah thought of that,but more work,ugh.
also thought about another layer and just extend solid model,probably the better alternative ? ? ?
But then you still going to have the tool rapid to the other end and do all that funky stuff.At least the actual part will look good.

[BurrMan]

Yeah thought of that,but more work,ugh..

Well, I just typed "2" in the stock dialogue. It didnt seem like too much work.

also thought about another layer and just extend solid model,probably the better alternative ? ? ?
But then you still going to have the tool rapid to the other end and do all that funky stuff.At least the actual part will look good

I suppose a boundry would halt it? But this method sounds like adding work and also is a "fakeout/workaround". I think the stock needs to be there for the path to know to continue on. Seems legit to me.

Are you talking about the multi-axis in the 4th and 5th pro ? ?

Yes. Full control of "extending off of the surface" and such. I think it needs the higher end toolpaths because they have the check and drive surfaces and continuity with the surface normals thats needed to create the extended toolpaths without wrecking your parts.

You've gotta work with what you have. Maybe enter the request to have the extend toolpaths option added at the request form.....

[BurrMan]

You might also look at doing 2d cuts, since the step is so minimal. You could just create open pockets with the adaptive? Anyway, just shooting pidgeons now...

[jrmach]

Burr typing the 2 is just part of it,still have to create boundries.All doable,xtra work Yes,and still have the tool do all that weird stuff although actual part will look better.
More are options are key here.
Yeah this part is going to be cut 2d.Thing is the customer wanted a model drawn up to see how the part would look with the dimensions he gave me.That being said I thought what the heck,I got a solid model,lets play.Truth be told,I don't use 3D tool paths very often.%99 of parts I do are faster in 2D and combination of using the right tools,which include form cutters.
You not shooting pidgeons,that is exactly what I would normally do.2D Adaptive Rough,

[jrmach]

2D Adaptive Rough fairs no better in this situation.I went with using V23,,I got HSP,and it rocks for this particular situation.So quick to program with the desired results.I am not knocking adaptive rough,because it has 3D capabilities where as HSP does not.
So yeah we need more functionality with 2D Adaptive Rough to bring it up par to previous V23.

[bobcad guy]

2D Adaptive Rough fairs no better in this situation.I went with using V23,,I got HSP,and it rocks for this particular situation.So quick to program with the desired results.I am not knocking adaptive rough,because it has 3D capabilities where as HSP does not.
So yeah we need more functionality with 2D Adaptive Rough to bring it up par to previous V23.

well hell, that was the first tool path I was gonna mention, all I did was turn off adaptive roughing the very first time I opened the file, and got the identical toolpath, but I thought ya had your heart set on fixing the issue so I tried to work within your constraints.

[jrmach]

I was trying to fix it.
Now it was showtime today,took the best fix.
By the way HSP is High Speed Pocketing offered as an add on module for V23 and V24.This is still a trochoidal approach to milling slots.
Simply turning off adaptive roughing in V25 is not a good way to mill this.The cutter will be %100 in the cut.

[bobcad guy]

I was trying to fix it.
Now it was showtime today,took the best fix.
By the way HSP is High Speed Pocketing offered as an add on module for V23 and V24.This is still a trochoidal approach to milling slots.
Simply turning off adaptive roughing in V25 is not a good way to mill this.The cutter will be %100 in the cut.

I hear ya jr, I spend more time trying to make the features run properly than I should, but I want it to work right, for the next time I need it. I really like knowing there are others out there, to pick up tips from, and maybe give a few back. I already picked up a quick little trick you were telling someone else in an older post, I didn't know that the rapid plane, and feed plane kinda work as incremental planes. that I can set my z feed plane at -.75, and my rapid at .1, and the z will rapid to -.650, instead of .1, that was helpful, as now I can program the z depths to the right depth, instead of moving top of part, and having to "lie " about the depth. right now i'm trying to find a way to drill and mill with the same tool, without having to fake out the system by changing the tool number later on, and it seems if you drill, then mill, it works, but if you mill first, no way can I get the system to output a g81 or 83. I have tickets into bobcad, and I am restricted to edit by hand after posting, by the boss, he wants the software doing its job. ive seen others mentioning about being able to do this too.

[jrmach]

I hear you.I use manuals tools all the time.V23 it was much easier,V25 you have to jump thru hoops and run circles.I had a whole thread about it,got me nowhere.It seems to be It IS wHAT iT iS.

[aldepoalo]

For some reason when the "pocket" is open on both sides you'll find the crazy looking cut back option that doesn't clean up the best. For the geometry you are working on I would create a 2D open pocket on 1 side. Extend the wire frame boundary on the other side to make sure the tool goes past the edge. Doing this forces the tool to say in the cut and clear everything out.

[jrmach]

Yeah,that's a good option Al.No worries,used HSP ,3 parts done Saturday.

[SBC Cycle]

I get a similar result with the adaptive rough. Having an "open" side and a fillet radius at the bottom of the pocket seems to goof it up. I thought it was because I was using a "square" bottom (normal endmill). I'll try to poke around my files and find one I can upload. It's happened to me about 4-5 times in the last few months.

EDIT: One additional detail - Usually when I get a weird spot like that I have to take another look to make sure it's not plowing full diameter into the part. More often than not, it does somewhere.