Official Luthiers Forum!
http://mowrystrings.luthiersforum.com/forum/

Visual Mill problem
http://mowrystrings.luthiersforum.com/forum/viewtopic.php?f=10106&t=43167		 Page 1 of 1
Author:  Saul Koll [ Tue Apr 08, 2014 6:20 pm ]
Post subject:  Visual Mill problem
Hello,
I'm having problems with Visual Mill. I will import my Rhino 4 drawing and everything looks good except for surfaces. I'm getting a shift where the surface is not in the right place compared to the normal curves.
Check it out: here is my drawing in Rhino
Image
and here is what it looks like when imported into Visual Mill:
Image

You can see how the tummy cut surface has moved up and over.

Does anyone know why this shift is happening?
As a workaround, I've been drawing a larger surface that is larger than my containment zone and it if shifts a little bit, it will still work-at least for this tummy cut.
Looking to fix this.

Thanks for any help!
Author:  RandK [ Tue Apr 08, 2014 8:05 pm ]
Post subject:  Re: Visual Mill problem
I'd have to guess that it doesn't quite like the quality of that surface as much as Rhino does. What kind of surface (type) is it ? If it is a loft of the two profiles then there are two degenerate points where the isocurves all meet up. I've had lots of issues with cam software puking over Rhino degenerate points which Rhino handles real well but they don't. If it's a different type of surface which is really 4 sided with a trim curve you might get away with simply doing an extendsrf on the body profile edge to make that oversized. I like having that hang over the body edge (~radius) so that the ball mill cleans up the body edge real nice.

A couple other troubleshooting things I often do are to try a different format like export selected to IGES and see what VM thinks about that. If that fails, open that IGES file in Rhino and make sure it matches the original Rhino.

I've had my issue with the tummy cut before and currently I do a loft to an oversized 3-sided solid cutter than I difference out of the body solid to make a clean cut. Otherwise drawing it as a 4-sided oversized surface and then trimming that as you want would probably yield a better surface. Yell if you want a hand or want me to try reading the file in my roo-cam to see what it thinks.
Author:  Sheldon Dingwall [ Tue Apr 08, 2014 11:41 pm ]
Post subject:  Re: Visual Mill problem
Hey Saul, it looks like the surface has changed shape, not just moved. What's it look like in Rhino when you show the control points? If it's a real complex surface, maybe it can be rebuilt.

Have you run this past support at mecsoft?
Author:  Saul Koll [ Wed Apr 09, 2014 12:24 am ]
Post subject:  Re: Visual Mill problem
No Sheldon, I came to my experts first ;-)

I'll try to get a points shot and post that.

Rand, thanks for the info. I'll try some of those ideas. Thanks for the offer, if I can't figure it out, I'll send it over.
I've been making a larger surface and a smaller containment line to deal with this and then clean it up with sanding later. It will be nice to figure out how to do it correctly.

thanks for the help guys, I'll be back...
Author:  Saul Koll [ Wed Apr 09, 2014 4:07 am ]
Post subject:  Re: Visual Mill problem
Ok, looks like I built this one using some sort of solid boolean kind of deal. (can't remember, but the size of the point display makes me think that.)
Hoping you can see what's going on:
Image
I used a big sphere and subtracted it from my body, I think.
If there is a better way, please advise.

thanks!
Author:  RandK [ Wed Apr 09, 2014 9:28 am ]
Post subject:  Re: Visual Mill problem
Differencing a sphere should be a good and clean method. I would want to see what the isocurves and points look like on that surface to see the layout but it still looks like a difference of opinion of what that surface should be. The .3dm file contains the parameters (ingredients) for the surfaces and it's up to VM to read those parameters and recreate the surface just like it is in Rhino. Choosing a different file format like IGES or STEP makes Rhino have to use those standard formats and sometimes the cam programs do a better job because the bugs get caught by more users.

Beyond that I'd just try changing something like the position or depth of the sphere a few thou and see what difference that makes. I'd untrim the edges of the suspect surface to see what those look like. Untrimming the edges of body surrounding the belly cut should bring it back to its original shape etc. Yell if you want a hand. Mec might help but also probably want to upgrade $$ you to the latest versions.
Author:  Saul Koll [ Wed Apr 09, 2014 1:19 pm ]
Post subject:  Re: Visual Mill problem
Thanks Rand, Good suggestions,
I'll play with it.
Yeah, there are newer versions of VM, but I'm trying to put off the upgrade for a bit.

Thanks for the help!
Author:  Sheldon Dingwall [ Wed Apr 09, 2014 3:27 pm ]
Post subject:  Re: Visual Mill problem
Hey Saul,

I don't know if it's a better way but it seems to me a sphere would be harder to sand than a simpler surface.

Try drawing a curve on the back surface that you would like to see as the edge of the tummy cut (red in my quickie drawing), then draw a line (blue) at a downward angle to represent the angle of the tummy cut. Do this in front view.

Attach the line to the end of your guide curve.

Make a copy of the curve and connect it to the other end of the line.

Sweep 2 rails using the two curves as rails and the line as a cross section. You'll have a nice simple tummy surface that's easy to sand.

I drew the line 90 degrees to the centerline when viewed from the top for simplicity sake. On our bodies the line would be angled in top view as well.
Author:  RandK [ Wed Apr 09, 2014 3:32 pm ]
Post subject:  Re: Visual Mill problem
You did show the points and isocurves, that was the large point field representing the sphere. The sphere is a nice simple revolved surface. It looks to me like VM isn't treating the trim curves the same way as Rhino or it's messing up the position of the sphere.

What do you have your absolute tolerance set to in tools>options>doc...>units ? I generally set this to 10x the smallest measurement I care about, so I set it to .0001. It can make a difference in the quality of the surfaces it makes. Changing this setting won't change anything in the file it just affects calculations made after the setting change. So in a test copy of your file you would need to make new versions of the surfaces you want to mess with like your body extrude and sphere etc and then do your differencing. If you are at this tolerance or lower forget about it. Put this near the bottom of the list below downloading a trial version of Rhino 5 to see if that is better. Upgrades eat a lot of time as well as the money.
Author:  Andy Birko [ Fri Apr 11, 2014 1:16 pm ]
Post subject:  Re: Visual Mill problem
I'm sorry that I don't have any advice to add but I will tell you that I've also had some weirdness with Rhino stuff imported into SolidWorks. I got a file with 2D sketches of text and upon importing into SolidWorks, I made an extrusion out of them and it looked fine. However, the tool paths I created with them were all goofy, like the splines were using different control vectors or something (made using VisualMill).

So I tried exporting to Vcarve Pro and the exported dxf had the same goofiness.

I ended up exporting from Rhino for Mac directly as a dxf and those worked fine in both SW and Vcarve.

Just another data point. There's got to be something non-standard about how Rhino handles splines for some reason or other.
Author:  Saul Koll [ Fri Apr 11, 2014 4:56 pm ]
Post subject:  Re: Visual Mill problem
Thanks Andy.

I'm on the road right now and will be back to work on this in a week or so. I'll be playing again with some of these suggestions. Thankfully there are lots of work arounds in Rhino, different strategies to arrive at the same place. Using all of your suggestions and more, I'll be experimenting when I return.
I'll let you know what I come up with,

Cheers,
saul
Author:  RandK [ Sat Apr 12, 2014 9:49 am ]
Post subject:  Re: Visual Mill problem
Saul - my old version of VM had some basic Booleans. You might try not differencing in Rhino and doing that in VM. Sheldon's sweep is another whack at it if that shape is OK. Extrude along a curve is a couple less steps than sweep2. The sphere is a common method of making those cuts. There's a way to do whatever you want and not have to change your design because of the software, it's just a matter of fooling with it enough.

Andy - Rhino to SW is problematic. When SW first supported .3dm as an add-in everything worked fine, including geometry into a part, its gotten worse since then. Their support director told me to use parasolid for solids/surfaces and DXF for geometry but Rhino's splines don't seem to translate well to DXF, or maybe needs special export settings like a new one labeled "CAM" that I haven't fooled with yet. I've had good luck for geometry import by making dummy surfaces in Rhino (text can be made solid or surfaces upon creation) then opening in a part file and using the edges if needed. My CAM software reads Rhino 3dm's perfectly but it prefers to have the tolerance tight. Try saving a model as STEP in SW and then open that file and see the edge gaps and things it complains about. Import/Export can be a patience test even with the same package.
Author:  Sheldon Dingwall [ Sat Apr 12, 2014 7:55 pm ]
Post subject:  Re: Visual Mill problem
Extrude along a curved works great. Thanks for the tip Rand.

Saul, I thought I was having the same problem yesterday. Turned out to be my top view was slightly off axis. The surfaces looked to be misplaced. Re-selecting top view re-aligned things.
Author:  Bob Garrish [ Thu Apr 17, 2014 5:46 pm ]
Post subject:  Re: Visual Mill problem
I use IGES and STEP to go between software. DXF if I need to.
Page 1 of 1 All times are UTC - 5 hours
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/