Swarf 5-axis

[Travis_Buchanan]

Swarf is not a toolpath I use often. It has been 3 years since I last used it. I am using upper and lower rails, I have my geometery sync'd by enity, and it backplots good as a 5-axis toolpath. The C-axis is only moving +/- 2 degrees off of 270, with all the tilt being on the b axis. I want to lock this to a 4-axis tool path but it goes crazy during backplot leading and lagging the tool down the straight side. Is there a way to stop this?

[John Summers]

I need to see the part. 5-axis goes crazy when the tool axis becomes nearly paralled with ANY rotary axis. It can be a nightmare and really impractical to fight what really is a limitation of the machine tool. They don't have a ball socket in the wrist like we do. Sometimes a wedge under the part will help.

[Travis_Buchanan]

I posted the mcx file to the ftp as swarf-4axis, under the mcx folder. Toolpath 1 is 5-axis, toolpath 2 is same thing locked to 4-axis. This works well with curve 5-axis.

[crazy^millman]

Hey Travis you couldn't Zip it??

 

I will take a look at it as well in about 10 mintues when it finishes downloading.

[CHouser]

Yea I know to zip, I was in a hurry. I am training a guy to take my place here and I am posting from his computer now, cause he is hoggin mine. -- Travis

[crazy^millman]

Well I see a couple things that get my attention real quick. The small sharp corners are going to be a pain. Also have you looked to adding equal number of points to each chain and sync by point verses entity.

 

I know John will call me crazy but I like curve 5 axis on a part like this better than swarf. The simple fact is I can control my tool axis like I want using my tool axis control lines to keep it from getting vertical or almost 0,0 on 4th and 5th axis.

 

Also look at your chains they are not very good either. I had to rework them some what to get them connected.

HTH

[CHouser]

I thought about adding points but I broke my splines so that I get the desired angle, It is the "twitching" of the b axis that I don't like on the 4-axis path down the sides of the part. Those sharp corners dont have to be clean I will leave the radius of the tool. The 5-axis path will work for what I need, I would just like to be able to lock it to 4, without it going crazy.

 

Travis posting from Chris computer.

[crazy^millman]

Why lock it I used curve 5 axis on it with a little bit of tool axis control lines and it looked great. Very smooth angle moves and not the big jumping around you were seeing with swarf. No how it will look on router I would need to think about my wind and unwind a little more for my start point and such but think that is really the easy part.

 

HTH

[Travis_Buchanan]

I really need to get a good grasp on swarf. I have one part that I did with curve 5-axis/lines and I had so many vector lines on it it looked like hair growing on the part. Also one part I did would only let you select x number of vector lines. For the most part I agree with you Ron that curve 5-ax is THE toolpath for 5-ax but swarf has it's place. The problem with curve 5-ax/lines is the tool vector change is linear between vector lines. Alot of parts would need tons of vector lines to accomplish desired movement. Imagine a twisted curved spiral hand rail on a spiral stair case.

 

Travis - back at my machine during lunch.

[crazy^millman]

Well then I would look to points on my chains and sync by points and see if that gives you the desired results.

 

Also have you ever looked at your step maxinum. I will bump this up but turn doen my tolerance to .00001. I have found it will hold my cut good in those drastic areas but allow it freedom to move better on smoother more free flowing areas.

 

HTH

[Travis_Buchanan]

I will give it a shot as soon as I get my computer back, I only have V9 Level II installed on this machine here. Thanks for your help and I will let you know.

[Travis_Buchanan]

OK I have tried sync by point with no luck. So if you ever have this type of problem here is a fix. Do it as a swarf full 5-ax. Backplot with save as geometery then do a curve 5-ax on the backploted geometery using the backplotted vector lines as your vector lines. Lock to 4-axis and you got it. I think the 4-axis button is broke on swarf. Using save as geometery feature in backplot has come in handy for many things in the past as well.

[crazy^millman]

I will use that on curve 5 axis and flowline and zig-zag to give me what I want a surface that is on the surface to dirve a tool. Then use my 3d curve created from save geometry and my surface saves alot of avergation and time.

[Travis_Buchanan]

Another good one is Surface/Finish/Project/Blend then save backploted chain and curve 5-ax that, using the surface as the vectoring source. Great for whittlin' out some pretty solid wood parts

[crazy^millman]

In thinking about this and looking at this part I would then look to 5 axis curve with surface edge as my dirve and my toolaxis lines. Now you have the surface to keep your tool from gouging and you keep control of the tool axis.

[Jim at Gentex]

Hey all,

I use 5 axis swarf quite a bit for trimming polycarbonate lenses used in flight helmets. I basically work with a large polycarbonate lens blank held onto a vacuum fixture which is machined to the inside contour of the lens blank. I've never had much luck using the top and bottom rail method to define the walls because of the syncing issue. I've always just used the vertical wall surface as my drive surface instead of chains, and the surface of the fixture as the tip control surface instead of the lower rail, and I don't check the fanning box. Since I cut all the way through the blank I'm not worried about the orientation of the bottom of the end mill since I only cut with the side. This usually gives me a nice clean toolpath without too much effort and produces a nice clean edge with the proper speed and feed. The 5 axis swarf toolpath may not work for some jobs, but for the application I use it for it's perfect.

 

Like with many other things, I learned through trial and error, so don't be afraid to experiment with the parameters of any toolpath to get the best result.

[Travis_Buchanan]

Ron, Cut is not normal to the surface on the one side. So I don't think surface edge will work.

 

Colin, That I did not know. Thanks for the infro. I actually broke a spline at a point. Creating what I thought were 2 splines. Do the 2 splines act the same as a trimmed spline? I did however manage to get the vector control I wanted.

 

Jim, I tried this method but got alot of lead and lag of the tool (result = collet nut rubbing stock). I will give this another shot tommorow.

[crazy^millman]

Thanks for the tip Colin I did not know this or even use it.