3D Scanners

[ddrueding]

Anyone have one yet? I've ordered a Kinect that I plan to use with one of the many software packages out there. Designing things in SolidWorks is fun, but most projects I have in mind involve an interface to an existing object. If that shape is simple I can break out the digital calipers, but if the shape is more organic that becomes difficult.

The current plan is to scan stuff into SolidWorks, design/modify the part, have a mold for the part 3D printed by one of the online services, and then fabricate the part out of carbon fiber using resin infusion.

(Wow, It sounds really ambitious when I write it down....)

 

[jtr1962]

3D scanning and printing is something I can see myself getting into in a big way once the cost of the hardware drops a bit. I love the idea of being able to either design a part and print it, or scan an existing part, modify it if necessary, and then make as many as I need. I think 3D printing/scanning is going to be bigger than desktop publishing.

 

[ddrueding]

The learning curve is a bit steep considering I'm not reading the manual (3 YouTube videos should be enough for anything shy of brain surgery )

Using my new Kinect ($80) and KScan3D ($200) I was able to create a thorough point cloud (12 million points) of half my couch by combining about 60 "pictures" over the course of about an hour. Not bad for a first shot.

 

[ddrueding]

The printing is not something I'm that interested in doing myself. Creating the model of something you possess from real life is the thing that is really hard to outsource and the CAD work I really enjoy doing. Creating the dimensionally correct and properly finished end result is something that others can do.

 

[jtr1962]

$280 for that kind of capability isn't bad. I would ultimately like to both scan and make items. My larger concern was actually being able to scan things because I know eventually 3D printers will be mainstream. Seeing what you can do with less than $300 worth of hardware sets those worries to rest.

 

[ddrueding]

The big issue I've run into so far is that this method does not work well on small items or reflective items (the Kinect projects an IR grid and uses that data to collect a depth map). My primary goal for this is automotive, so that may be an issue. My car at the moment is just the right amount of dirty to give this a try this afternoon.

 

[Howell]

Matte plastidip?

 

[ddrueding]

Matte plastidip?

I was thinking flour. It also doesn't seem to like matte black; my DSLR and keyboard were giving it fits.

 

[ddrueding]

Making significant progress on generating point clouds. The automatic alignment is good and the manual assisted alignment is awesome. At the moment I'm running into a few issues:

1. My "Home1" machine becomes laggy @ 50,000,000 vertices, and near unresponsive at 75 million. It seems this program only allows decimation after a final mesh has been generated. Grr.

2. Bringing a final mesh into SolidWorks isn't as easy as I'd thought. You can import the solid, but the process of generating a workable native-Solidworks surface is non-trivial. There is a "ScanTo3D" component that I may have to get. There are also 3rd party programs specifically designed for reverse-engineering from 3D scans that look awesome.

 

[Mercutio]

The MakerBot people announced some kind of huge deal with Dell today.

Also:

.

 

[ddrueding]

More importantly, I would download a car, modify it, and then make one of those.

 

[Handruin]

The MakerBot people announced some kind of huge deal with Dell today.

Also:

.

I don't doubt this either. Once those printers are a dime a dozen ($99 with ink cost $4000/ml) you're right that we will probably see this kind of stuff.

 

[ddrueding]

The plastic extruders that are out there now are pretty clearly mostly toys. The ones I'm excited about are the selective laser sintering machines capable of working in Ti and Al.

 

[LunarMist]

What do you use that for?

 

[ddrueding]

One of the commercial applications to the SLS process is making custom-sized Ti bicycle frames as light as many carbon-fiber units but stronger for certain applications (looking straight at JTR on this one). Being able to vary the thickness of the tube walls and enlarge the joints between tubes is a huge advantage.

Small production volume parts are also a strength of this process; I'm considering designing custom handles for the cabinetry in the kitchen just 'cause.

 

[Howell]

"Low volume production of prototype models." It requires a large investment to get the machinery ready for large scale production. This lets you get through prototyping faster and cheaper.