Repositioning 3-D Objects

[Craig Moellenhoff (Unregistered Guest)]

I (still) don't own a SB, but I have been considering an idea that would require use of some type of CNC. The object I would like to cut is an asymmetrical 3-D shape that would have to be machined on multiple sides.

It seems that it would be possible to cut the rough stock into a 2-D shape on the SB and machine the top profile. Then I was hoping to cut some type of registered "pockets" in a spoil board that would match the shape of the top and sides of the initially cut object. Then the stock could be moved into each molded pocket, fastened and cut on alternate sides until the complete profile is complete. Since the size of my object is ~6" X 3" X 12", I think this would be feasible.

I'm not sure if my description is clear, but I was curious if anyone has used a similar method for cutting multiple sides of a 3-D object.

I would also have to determine the best way to create the toolpaths for staged cutting from different sides, but that is another challenge.

Any feedback would be appreciated. Thanks.

Craig

[stevem]

Craig, you seem to have a good understanding of the concept.

[Brady Watson]

Craig,
Check out DeskProto (http://www.deskproto.com/)

-Brady

[dvanr]

I don't have an indexer. But, would an indexer be able to do an asymetrical job as Craig describes? At 6" x 3" it might just fit.

[gerald_d]

Craig, your logic is sound.

You can also consider leaving the ends of your blank as rectangular shapes as long as possible. Trim those reference ends way right at the end - maybe in a "cup" jig as you described, maybe in a bandsaw......

[moehink@ieee.org]

I originally thought about the indexer, but then I realized that the SB "borrows" one axis of movement to control the indexer. Since my object is asymmetrical, it would require true 4-axis machining to achieve the correct shape. The SB only allows the router to move in two directions with an indexer (along the indexer axis and the z depth of cut), so it could not rout my object properly. The third axis is used to control the rotation of the indexer.

I did consider leaving extra scrap at the ends to provide stable positioning and reference. This definitely seems like it would be a good approach.

Thanks for your responses. I will probably post in another folder to see if I can find anyone willing to help me work on a prototype. Hopefully I will find someone in my general area (St. Louis, MO) who can spare some time and knowledge. Thanks

Craig

[Brady Watson]

Craig,
I mentioned DeskProto because it not only allows true 4th axis machining & 'interpolated' 4th axis machining, but also supports double-sided milling and 'N-sided' milling. I believe that VisualMill also has an 'N-sided' feature as well in the full version.

It would be helpful if you had some type of picture that you could post on here (you must be registered to post pics) so that we can get an idea as to what you want to accomplish. I have milled many projects using DeskProto's double-sided milling feature (without a 4th axis/indexer) that would be impossible to do using other CAM packages. No matter what strategy (4th axis or 2-sided milling) you choose, you will need some method of holding the workpiece while machining. DeskProto will attach 3D bridges to the material block, which is very helpful. If this is a prototype, then it would be the path that I would pursue...However, if you are looking to produce these pieces on a production basis, then you will have to invest the time and research to create a purpose-build vacuum jig accurate enough to hold the workpiece.

-Brady

[gerald_d]

"The SB only allows the router to move in two directions with an indexer (along the indexer axis and the z depth of cut)..." It can also move across the axis and I think it can do all 4 moves simultaneously (an M4 command?) - but the big headache would be to generate the code.

[bill.young]

Hey Craig,

Gerald is correct for sure...generating 4 and 5 axis toolpaths is a tricky process. You could however use an indexer to rotate the blank but make the cuts as regular 3d files. You could create a toolpath for the top and cut that, then have the indexer turn the part 180 degrees (with no other axis moving), and then cut the file for the other side.

With lots of shapes this technique may not make sense and just complicates things, but if your shape has details near the edges they might not turn out well when just cutting from 2 sides. If that's the case you can create a file that's rotated 90 degrees ( or 52.5 or whatever works best for your shape) and use the indexer to rotate to that angle before cutting that file.

There's an indexer virtual tool in the newest software download and an instruction pdf file in the "Virtual Tools/Indexer" folder in your ShopBot install directory if you're curious on how an Indexer could be setup.

Bill

[craigmoe]

Thanks for clarifying the indexer capabilities. I thought I read somewhere that it could not move horizontally perpendicular to the indexer axis, but I cannot even remember exactly where I got that impression.

In this case the indexer would probably be the easiest way to setup the object for cutting, instead of multiple repositioning, fastening and cutting steps. The cutting could be much more detailed, becuase undercutting would not be a problem. On the other hand, generating the toolpath would be far more complex. Would DeskProto handle this application?

I registered as a user, but I don't have a picture that is ready to post at this time. Imagine a staircase handrail section that follows an arc. It is asymmetrical about any axis and the profile must be machined on all sides. The curvature of the object would also require a fairly large Z-dimension capacity to account for the convex side of the arc being machined (projecting up from the axis). I have also wondered if torque resistance would be a problem when the curvature lies in the X-Y plane. If the cutter is pushing down on a curved arc that projects a few inches from the centerline, would the indexer be able to hold its position? If not, would there be a way to use programmable switches and some type of clamping mechanism to provide additional support?

Sorry for all the questions. I had the pleasure of attending the Jamboree a couple years ago, but have never had the opportunity to experiment with a SB. I frequently peruse the forum and have read quite a bit, but that is no substitute for experience. Thanks again for all the information.

[bill.young]

Hey Craig,

The Windows software can move in 5 axis at the same time and so has many more movement possibilities than the DOS software...in the old days you had to make a splitter to run both X-motors off of one driver chip and even then it was slow and problematic.

You might want to send Grant at ShopBot an email...he's spent a lot of time on these kinds of file-creation issues for the indexer and may have some suggestions.

Bill