5x pocket problem

[justind]

Hi all,

 

Been a while since I have posted on here, hope everyone is doing well. My latest project is giving me some issues --- I attached a .MCX-6 file with a couple 5x pockets I'm dealing with. The older style toolpaths I usually use aren't getting me good results so I have had to stray into the advanced toolpaths. I'm still not really getting gouge free work out of them. I was wondering if some of the 5x gurus could throw down some example toolpaths on how they would approach this pocket situation. Really trying to use a .50 bullmill .125 cr to finish walls most of the way down - .50 ball is cool for finishing floors and sewing floor to wall together in one pocket. Other pocker has a .125 fillet so little different animal. I used a .75 bullmill .125 cr to rough pockets and parts. I can't post more geomtry, because ITAR and all the rules don't allow. Zero mismatch allowance and +/-.005 on wall thickness to add to fun. I will be cutting this part on a HEAD/HEAD 5x config - +/-25 deg - st A/B setup ifthat matters. If your wanitng to check and see if toolpath is not gouging just write .stl of pocket and do compare. That should give pretty good result. I attached a pic of one of my pocket results. I have reached out to my dealer and they have reached out to mastercam on this. So far no one has a really easy or consistant solution to getting good pockets on this. I'm real curious to see what can be created out of the forum on this. This part made me swtich to X6 because some of the Advanced toolpaths were messed with between releases. So far I haven't got X5 to give me one good toolpath yet, X6 has been the best. With this being said I have converted over to X6 completely and not looking back. Dealing with posts issues as the pop up.

 

thankyou

 

 

JD

5X_POCKETS.MCX-6

[crazy^millman]

Well I see you biggest problem is the size of the endmill. Trying to force that .5 endmill into a model that has a .2499 Radius on the Fillets is not going to work. I understand you are looking for beef here I get that, but the old saying of trying to stuff 5o lbs of carrots into a 5lb bag applies. I would really consider going to a smaller endmill with the same C/R and see what happens from there. Your current approach is going ot be unpredicatable and yeild results I would less that desierable which is what you are currently seeing. You say old school and the new toolpaths, but you do not say what are they are. Give some feed back on your attempts. These are very strightfoward pockets and really think just going to the smaller endmill will solve most of your problems here.

 

Making something you can use is going ot cost you. I have a 2 hr min and taking away from paying customers to spend 30 to 40 minutes making you something up. Sorry just to busy supporting people paying me.

[justind]

Well I would like to think that if these pockets are so straight forward my dealer and mastercam themselves would of sent me good solution so far. Were many days into this - I have been doing other projects waiting on an easy solution and no luck. Now it's just time to get the job done.

 

I can actually drive a .500 tool and do these, just not very easy. The pretty screen shot I attached wiht the posted topic is all done with .500 dia tools. Should actually try driving one of these before saying it's so easy - Even with a .250 dia tool mastercam has issues driving this by the way.

 

By old school I mean basic swarf and curve using chains. Syncing and not. By new, any of the advanced toolpath modules.

 

I am a long time user "V5-to now" and been doing 5 axis a very long time and this is the 1st time I have ever reached out for help on this matter. I sure appreciate the reply from such a 5 axis guru such as your self and sorry if it cost you any money.

[Hardmill]

Yo Ron reread his post, he was just looking for examples or to at least be pointed in the right direction.

He didnt ask you to do it for him dude.

 

and this is the 1st time I have ever reached out for help on this matter. I sure appreciate the reply from such a 5 axis guru such as your self and sorry if it cost you any money.

 

Hopefully someone can chime in and point you in the right direction Justin.

Im not much of a 5 axis guy outside of 3+2(with good help from gcode)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PEACE

[Chris Bloedorn]

Once it roughed out you could use parallel to multiple curves to get a decent toolpath. There are some good sample files in the shared folder....(C:\Users\Public\Documents\shared Mcamx6\mcx\mill\samples\inch\Multiaxis).

 

Then under roughing you can set it to use depth cuts and get a decent toolpath.

 

HTH

[justind]

thankyou guys for the replies.

 

I have actually looked at the samples you have suggested last week and used some of the paths to get my results. Nothing is very pretty so far. Lots of settings on the advanced paths but i'm digging all the options. Some of them settings remind me of when I was programming 5 axis in catia. I like them both. I attached a pic of some of the operations I used to tackle this thing. I did also resort back to all the samples that came with x2 and the x2 multiaxis .pdf that comes in the documentation folder. I literally have tapped into every resource I know and can think of. I will get thru this for sure, being pretty is pretty much out of the question now --- I just toolpath away and vericut heavy to make sure I don't do anyhting to crazy. I wish I could show more of the part, there is another end I will have to use some of the porting toolpaths more then likely. I'm going to have to get better at the advanced paths for sure and know what all the settings can do, there is just so many choices and if you have one wrong or change it i'm seeing it bomb out my whole path. Even the collision settings I have to try multiple choices to get things to go. Quite a part to get my hands dirty in X6 for the fist time. Only true crashing of the software I been getting is when I tried to re create one of the wall surfaces, other then that things have been solid.

 

thanks again

 

JD

[justind]

forgot to hit attach button - ops should show now

[Hardmill]

Good luck

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PEACE

[HT]

Hi JD, did you try flown 5x through a point?

[justind]

Good to hear from you big daddy - I haven't tried any of the toolpaths with controling the tool axis with a point yet, I will have to give that a try. I got one more pocket to get the floor and wall to sew together then hopefully can move past this nitemare. I had to treat each pocket differently with different styles of path. Hopefully the port style side will not give me as much trouble.

 

hope all is going well, have to tell Tim and Charles I say hello also.

[HT]

[crazy^millman]

Sorry if it came across like an arse, but what are the toolpaths you had tried? I see a sample file with no toolpaths for axis controls or anything applied to it. I have always helped people over the years and sorry, but I like helping people who have shown me something to help them with. What I got when I downloaded this file was nothing put into it. You have been doing this for a long time and good for you, but I can only judge what I see here. I would really look into the advanced toolpath and see what you get from there. Did you try curve 5 axis? 5 Axis Parallel or swarf or what? What are the settings you have seen give you the results you are looking for? You say toolpaths did not work what toolpaths did not work and why? You can throw some toolpaths on that part and share and then we can help you tune them from there. I make a living on the side training people to do this so sorry if I am looking the way other may not like, but I did not see a rush of people helping so must be something in what I read in the posting.

 

HTH

[justind]

No biggie, I figured the attached auto diff pic and the fact I was I just said I was curious to see what could be produced out of the forum pretty much setup the situation. I wasn't looking for someone to do my job for me. Sorry if my post didn't come out clear. From what i'm gathering not many have replied because there is no easy solution that can be done quick. We all are busy and got things going on, I understand and respect that. Some have said it will be a good file to save and test future softwares and or releases on. Nobody else has said it was straight forward 5x pockets and I figure if they were we would see some cam files like I have seen on some recent rotary topics pop up pretty quick. Wish you lived closer, I would definately hire ya for some lessons.

 

I posted a pic of all the operations I used to get things done earlier. Take care

[Jrygus]

Here is an example that may help. It is not a 5 axis pocket it is just deep, so I used an advanced 5x parallel to surface path to stitch the walls. You might be able to deconstruct it to your geometry if that is a process you want to try

DEEP 5X POCKET.MCX-6

[crazy^millman]

It is all good and again always trying to help others. I to be honest thought it was more of just work than not doable. I have seen many times where the straight forward approach just does not work. Old School brute force seems to go a long way with solving the issues. Taking the time to draw the toolpath Vectors and then using a chain to drive the toolpath. I have even gone as far as use a 3d toolpath to Keller walls. Take that toolpath export the path as geometry and then use that to drive a curve 5 axis toolpath using my own custom vectors to really control the dance of the tool. Again comes down to how much work you want to put into one pocket. Me personally I have no problem spending 30 to 45 minutes on something someone else might spend 5 minutes on to get what I think it best that runs the machine what I think it the correct way and not what produce just something I think looks okay. I have had to sit down and spend 100 of hours writing programs by hand for Multiaxis and I care as much about how the machine is going to run to produce the part as part of the quality of the part. Do not get me wrong if 5 minutes of programming yield desirable results then great, but if not then I dig in and do what I think is best. I will dig around and see what I have is not ITAR or not shareable.

[justind]

I totally agree and did actually take the time to model in new surfaces, new corner radius surfaces that are .005-.010 bigger, projected nice clean chains that sync with no problems, even created my own vector lines and tried those methods. I sunk way more time in then I was planning on actually when starting this operation on the part. That made me go for the advance paths hoping for some kind of standard or easy way to approach twisted surfaces like I had. The parallel to multiple curves is probally the best and easiest bet for most situations like this from what I have gathered. Now one pocket it just wouldn't work and I went for morph between curves. Now Jrygus posted a file that used parallel to surfaces " thankyou for that file by the way Jrygus", that has some real potential of working also. I used morph between two surfaces too kinda get same thing, but used it more to sew floors to walls. I think me wanting a standard approach every time in 5x was a dream that I need to forget lol. Just need to open up more to the huge selection of setting, especially controling the tool axis in the advanced toolpaths and work on the collision control settings. I really never gave these paths a good chance to now, and honestly kinda kicking myself in the rear for not experimenting sooner. Never any fun when you on a project and have to start the experimenting process.

[Tyler Robertson]

one big thing with collision control is making sure your tool is set up in a standard 'make sense' kind of way. ie: is it your flute, or shaft that is colliding?

 

also, if you over-constrain .. ie: only give it 10 degrees to try and side tilt to avoid collision, and it needs to go 15, it will just ignore your ruleset and try to cut the part, so be careful that you don't limit too much.

 

Make sure your linking parameters are set appropriately too. Sometimes people will see retracts all over the place when they use one strategy and think it won't work, but changing the linking parameters could've had them away to the races. "Follow surface" for gaps is often a good bet because there could be small gaps that you don't notice that when moving 'direct' do some weird things. you still have to watch with follow surface, but I find I get in generla, better results.

 

two criminally overlooked pages (and criminally underwhelming in the help file) in the 'advanced' or 'newschool' multiaxis paths are under Cut Pattern, called "Parameters for Surface Edge Handling" and "Advanced Options for Surface Quality". Surface edge handling shows the toolpath a surface profile over a gap based on the distance. I noticed on some of the rads there are little gaps so you might want to play with this value. Don't make it massively larger or smaller.

Surface-quality might help moreso - if you can make a tighter chaining tolerance it could do a better job of fitting that tool in because it's so close in rad size to the actual rads. also, Adaptive cuts could give you better results on the walls.

[crazy^millman]

Never any fun when you on a project and have to start the experimenting process.

 

Yeah think we all have been there done that and hated it when what should works never does.

[justind]

Tyler , thankyou - That is some good parameter and rules info to go by for sure. I will be saving those to my one of my tech docs for future reference for sure.

 

They just sent me a new revision of the part and a pocket did change and some other areas, so maybe I can try those settings out soon as they say this is the new released data. I am guilty of taking those most over looked settings not serious - think I just left them at default values. I did take some of the mastercam samples and export them then try to use the same settings to drive in my file with no luck when I was doing my early tests. They did get the brain going to finally get someting to drive after changing multiple setting and geomtry. I would have to say these are some picky toolpaths when it comes to settings for sure.

 

Is a .001 chaining tolerance out of the question on this or should I make it tighter? I have always just used .001 in the past.

[Tyler Robertson]

Don't feel too bad about it - most things can be done with default values on parameters like that .. I just get thrown CAD from any system as well as STL files and I have to figure out how to multiaxis cut them... so I've been in the same position too.

 

I would make it tighter - start with 0.0001, you can change the Surface Quality tolerance, but also the total toolpath Tolerance on the Cut Pattern page. That cut pattern tolerance will give you more points - it can help as well in some instances, unless you have a multiaxis machine that balks at too much point data, but usually they tend to run smoother, just slower (I don't have a lot of CNC experience there but the couple I've seen have done that ... robots on the other hand...)

[jbreth]

Here is an example using " Convert to 5 Axis" Works quite well.

5X_POCKETS_JB.MCX-6

[justind]

That is a slick approach jbreth - i like that. - thankyou -- I'm going to see if I can get same result in aircraft cordinates. I seen every thing in sample was brought into local cordinates.

[jbreth]

we do a bunch of complex 5 axis pgm's here and that methods works great! Good luck!

[crazy^millman]

Funny I have seen these threads turn into this type of thing with people sharing ideas and gleaming knowledge from others. This use to eb just about every other thread in this years ago. Nice when you see it.

[MoMo1108]

here is a simple high speed surface to rough and a simpple swarf ... the surfaces wer all broken and that makes the flow lines go everywere and that what it wants to really follow so i fixed the surfaces and did that you see here ... only did one pocket no feeds speed just so you can see the path .... not to sure why you have your wcs so far off in space why not mcx zeros ? back plot looks ok might need a tweek verify was getting funky but i didnt have time to play with it much.. hope this helps

5X_POCKETS.MCX-6