PDA

View Full Version : 4th Axis/Indexer/Rotary Laser Scanning Now Available



Brady Watson
06-07-2007, 07:55 PM
In addition to standard 3-Axis Hi-Def laser scanning, we now offering Hi-Def 4th axis rotary scanning. This is an update for the masses who have e-mailed us anxiously waiting for '3D Indexer Food'. Things like cabriole legs, totem poles, acorn & pineapple finals are ideally suited for this service, or any model conducive to being placed between centers, just like your ShopBot indexer accessory. In addition, a vertical-type rotary scanner suitable for statuary type work where the material is too brittle or delicate to be placed between centers, is currently under development.

There are a few ways to machine true 4th axis 3D models including ArtCAM, VisualMill with rotary module and DeskProto. Cut3D can also be used to do rotary axis work by using the indexer as a 'flipper' and creating either a 2-sided or 4-sided machining strategy, then manually tell the machine to rotate either 90 or 180 degrees in between successive toolpaths.

Here are a few examples created on the 4th axis scanner this week, along with shots of the original models. 'Blue rendering' courteousy of Vectric Cut3D, models courtesy of my customers:


3399
3400

3401
3402

3403
3404

If you have the need to reproduce 3D components in your business using the ShopBot indexer and lack the ability to get your 3D content into the computer, feel free to contact me clicking on my name on the left side of the screen. Be sure to send a photo of the item you wish to have scanned. We can provide you with a suitable 3D file format that is compatable with your current 3D toolpathing program with rotary support.

Thanks!
-Brady

myxpykalix
06-07-2007, 10:24 PM
Very cool. So the idea is, you can reproduce a part and furnish a model, correct? What if we don't have the appropriate software to create the cutting file, is that something you can do?

Brady Watson
06-07-2007, 11:08 PM
Down the road I will be offering toolpathing services on 4th axis type models. For the time being, I want to learn all I can about the machining process as it is very different from standard 3D milling. I won't let toolpaths go out the door unless I am confident that they will perform exactly as planned. So for the time being, no.

rustnrot
06-08-2007, 04:53 PM
Brady, this is great news. Inasfar as the pineapple is concerned, it appears it is not possible to machine it exactly like the original without a 5 axis machine? Whether it is machineable or not, does beg the question, can you "clean up" models so they are machineable in 4 axes (i.e. no undercuts)?

stevem
06-08-2007, 05:20 PM
Tom, the CAM software for a 3 or 4 axis machines will ignore the undercuts and calculate the tool path as if there are no undercuts.

Brady Watson
06-08-2007, 05:32 PM
Tom,
Excellent observation ~ you are correct. Models that have undercuts that shield the laser beam will be wrapped only with the data collected by the laser. In most situations, this is not an issue for 2 reasons: 1) Those who own a 4th axis won't be able to cut the undercuts either and 2) Most of customers goals are use the CNC to get the part about 90% & then touch up the rest by hand.

There are times however, when the model needs to be disassembled and scanned in multiple pieces. The Statue of Liberty for instance would need to be broken up into at least 3 separate components to get the entire model. Consequently it would ALSO have to be machined in at least 3 separate parts if your project called for that level of detail, as illustrated below using a sample 3DS from the web:


3405
3406

3407
3408

The method in which the model is broken apart into components is tricky, since some models may be irreplaceable or not conducive to disassembly. In this case, the model needs to be molded in rubber and then casts can be made from the original mold and then carefully broken up into parts for digitizing. At this point, as you can imagine, it is very time consuming and as a result, costly. So, yes this sort of thing can be done if absolutely necessary. The same basic technique would hold true for 4-legged animals and similar models where the laser is shielded by a front or back leg. The good news is that most of the money making parts that people would want to do on their indexer are very straight forward with no concern for undercuts.

-B

rustnrot
06-09-2007, 01:04 PM
Thanks for the insight, guys. Ok, another question, Brady you mentioned using a Shopbot 4th axis as a "flipper" if one is using Cut3D. Since I do not have a Shopbot 4th axis, but do have Cut3D, I can see building a very simple 0-90-180-270 rotatable manual indexer (also for machinists there appear to be several of these on the market that could possibly be used as-is or modified).

That said, Cut3D should be able to provide alot of parts cutting using this simple indexer. What, in your opinion, would be the advantage of a fully programmable 4th axis vs. Cut3D and a poor man's indexer?

Brady Watson
06-09-2007, 03:12 PM
Tom,
Properly sized material, along with carefully placed 'stop blocks' on the spoilboard is perfect for doing 2-sided or 4-sided milling using Cut3D. You can alternatively use a 'manual indexer' as you describe to place the part between centers & cut, although it is unecessary as each of the 4 toolpaths are zeroed to the top of the block at each turn. A toolpath setup for the Indexer would be zeroed to the center of the rotary axis.

To answer your question in regards to the advantages of rotary over a 4-sided part, there are certain designs that cannot be completely machined by simply doing a 4-sided machining strategy. Let's say for instance that you had a column with grape vines and leaves wrapped so that it would spiral around it, all the way down the column. Each toolpath is setup to run the very tip of the ball tool into the relief. As the toolpath approaches the outer edges of the column on that particular side, it is no longer contacting the center of the ball tool, but rather the side of the tool, resulting in what some call the 'sewing machine effect'. The tool does not intersect the relief dead on and the result tends to skew the finished result.

I'll see if I can provide a graphical illustration of this.

-B

Brady Watson
06-13-2007, 02:07 PM
For those interested, I have added some additional photo examples to the website:

http://www.ibild.com/3DServices.htm

Thanks!
-B