Art
Wait a second what do you mean it dose not do radius in the above pictured pawn and the 50 example they have several radius what gives.

Dennis

There is no radius or arc move there, all your seeing is the end points of the stright line cuts, they form an arc i flooked at from end point to endpoint, but if you zoom in on any toolpath youll see there iare no arcs, the toolpath is always comp[rised only of straight lines..

(maybe we're talkign of two different things..)

Art

art
look at the profile on page 85 called 50 comment the cut line follows an in and out curve on the profile it's not just a straight line especially the green cut line, am the only one actually cutting metal ?

Dennis

Hi:

 That curve in the cur is created only by sucessive straght cuts. The green lien is not a cut line, its just a guide line for me to use in my calculations.
LazyTurn has never put out arcs. There is no code in it to do so as yet, the green lien is not meant to be a cut line, its all just straight cut lines adding up to the end shape..

Art

Art,

Still with you, just doing to manny things lately.
RICH
 

Hi Art
Yes i agree with what you are saying back and forth straight line cuts Chip and i worked on the file last night and got Mach to load the g-code correctly don't know what was going on but Mach would only cut about 4 pass and never make the step cuts for the curve or arc what i was trying to do was to rough out the part in LT and then make a finish pass in LC more steps then i would like but it gets the job done and that's the important thing is there a rule of thumb for the minimum cutter tip size maybe this has something to do with it thanks for the help.

Dennis

Dennis:

  the only real rule is that the tip fit. LazyTurn is being written as a solid's comparator. Its the only way I could figure out a path to completion,
so it creates solids from the profiles and tools, then trys to fit the tool onto the profile in scanlines. It has to compute intersections and collisions
to do the path. Its all very complex and some bugs remain in the calculations, but its getting tighter all the time, failures can be caused by the code
or the tool size, and even the pass depth, you'll notice a filure at one pass depth may be corrected by simply modiying the depth by a small amount.
even a change from .1 to .1001 can fix the trouble.

  Ill be switching types of algorithsm soon to try to get the last of the intersection bugs, ( so ironically it may get more buggy while thats tracked down.   )

  Until I dream up a few more checking algorithsm to fix up head on collisions with intersectional point duplicated in the tool and profile, this error
will continue, so when a profile fails, try a slightly different depth. If a profile says " Self interseting profile found" try chaning the tool to a slighly smaller
tip radius or slightly larger, typically it will fix the trouble.

Thx
Art

I still have never been able to get anything close to a tool path with LazyTurn.  I keep trying and sometimes can get close.  I have attached a DXF and a screen shot of my latest try.  The red circle on the left is where Lazyturn looks like it parts off the part even though the DXF does not have a 90 degree angle there.  The red circle on the right is where is looks like it put the tool path.

Vince

Hi Vince,

Try the following;
1. increase your stock diameter - there's not much to rough out in your screen shot.
2. In the rough profile dialogue box set pullout clearance, stock clearance and depth per pass to 0.1, 0.01 and 0.1
3. Try a 35Deg diamond tool with a tip radius of 0.01, and an inscribed circle of 0.2.

I appreciate that these settings might not be what you want to use, but should get you a result as below.

The line in the red circle is I believe part of the dxf profile that LT is working towards. A parting off operation is something for the future.  LT just does rouging cuts at the moment which consist of straight cuts parallel to the work axis.
Rgds
Paul

Vince,
You should be able to get the roughing passes out of LazyTurn and finish pass out Lazycam Turn.
That is a whole lot easier than hand coding something like your recent post in another topic.
That's what DennisF and Chip are up to. It should work and just need to a carefull how you append the finish
gcode into the code file from LazyTurn. You could just use Lazycam Turn but it sometimes times generates a
code line which cuts thru the profile ( so make sure you dry run it and watch for that cut through). The fix is usualy nothinig
more than changing the order of an X or Z move from a combination move. See the LazycamTurn manual as i talk about that problem.
Why it's bombing out in Mach for  Dennis  I don't know as it shouldn't.  

Just a reminder that there are a lot of versions of Lazyturn and some are created for Art's purposes of checking the program.

Again, still easier than handcoding, piecing together from wizards, and cheaper than spending a lot of money
for a lathe program.  

RICH