3d rapid prototyping
 

There is a cool service that takes your 3d models and creates a plastic copy of your model. I am working on a tool for a glass light fixture. We are casting it in steel and will use the tool to slump glass into it to make a glass lamp shade. I used Autcad and 3D Studio Max to create the STL file that will be used to create a scale plastic model. The model will be used by a foundry to make a sand mold to cast the steel.
Sent 3d model out to be produced today.

This is the first time I have produced a rapid prototype and it has gotten me thinking about making some other things. Mainly cars and car parts. Just read a great article about the V2 356 and Strahle, inspiring stuff. Anyhow, I am enjoying making 3d prototypes and stuff out of metal.

Sounds like something I would like to learn to do at some point in time in my life....good luck

Stereo lithography? If so, very cool stuff.

I have seen 3 different variations of this process.

1) Metal beads encased in plastic which a laser melts and then the item is placed into a sintering oven burning off the plastic and sintering the metal

2) thin paper which was cut and glued together to make the prototype

3) clear liquid plastic whcih was shot with a laser which made it solidify

All pretty cool processes which can make shapes imposible by other processes.

Speedy:)

lyon,
What part of Wisonsin are you from?

Speedy:)

Stereo lithography is #3. Very cool stuff, but not cheap.

Eventually, according to Popular Science a few months back, the average guy will have a 3D printer in the home for making replacement parts for whatever breaks - like a knob for the washing machine.

I think they will be in Staples / Office Depot in a few years. Already there is a desktop version for around $2500 - the difference in that machine and the ones like you typically use in product design is resolution. The $2500 machine does not match up to the $35,000. machine.

Here's the ultimate 3D "printer" - the Star Trek Replicator:

http://forums.pelicanparts.com/uploa...1188452963.jpg

Best,

Kurt

Yea, but I think the principle behind the replicator was that it used the relationship between matter and energy (Einstein's famous E=mc^2) to create matter from the enormous energy available from the matter/antimatter reaction vessels onboard.

I believe energy-to-matter conversions have been done in particle accelerators but the amounts of energy required even to create tiny quantities are staggering. One of the physicists/scientists on here could probably speak to this a little better.

I've lost track of how many RP parts I've had made in the last 10 yrs, or so. It's an essential step in product prototyping. ...SLA, SLS . ..

As cool of a process that all that is, interestingly, a good number of files that I send out to an RP house are Machined. Obviously a slew of RP processes are available to to these RP vendors, yet with 5-axis machining, the quick part is often machined. ...even parts w/o a single straight edge, or flat surface. Surprising, eh?

Patent was 1993. IIRC.

Edit: nope "Stereolithography (SLA), the first Rapid Prototyping process, was developed by 3D Systems of Valencia, California, USA, founded in 1986."

Maybe another process?

"Selective Laser Sintering (SLS®, registered trademark by DTM™ of Austin, Texas, USA) is a process that was patented in 1989 by Carl Deckard, a University of Texas graduate student."

"Stratasys of Eden Prairie, MN makes Fused Deposition Modeling (FDM) machines. The FDM process was developed by Scott Crump in 1988."

hmmm. What process were you doing, Wayne?

take a picture
 

Going the other way. This has been a very nice addition to our shop.
http://www10.mcadcafe.com/goto.php?encode_url=aHR0cDovL3d3dy53aGl0ZWxpZ2h0c2 Nhbm5pbmcuY29t (find the Atos III in the product section)
[We often use it to create models of organic surfaces. We can make a wood grain texture that really looks like a wood grain because it is a scan of a real chunk of wood. The other cool thing is being able to reverse engineer stuff on the fly when customers do not have or lost their data. http://forums.pelicanparts.com/uploa...1188459168.jpg

The start of a slate textured part.

How is rapid prototyping different from machining a piece of plastic using a CNC machine?

RP starts with nothing and builds up a part level by level.
Machining a part starts with a chunk of whatever and removes material to create the shape.

Wayne. Very cool paper! You were one very bright and well organized cat!

see http://www.efunda.com/processes/rapid_prototyping/intro.cfm

Got it.

So why is adding material better than taking a block of existing material and removing the excess to reveal the shape?

When you add it, layer by layer, you can essentially make it look like anything. Machining is great, but it's pretty hard to shape the inside of whatever you machine without cutting your material in half. With stereolith, you don't have to have access to the inside, because the inside is being "grown" along with the outside surfaces.

The last place I worked had that capability.
We would use a fero arm to scan a part into autocad and e-mail it to Sulzer STURM (a division of the company in Virginia).

They had a 3D printer that made a part out of a corn starch type of plaster which was used in the mold making process. More accurate than a regular casting so less machine work needed to remove excess material.

It would be cool to be able to order and new 911 body produced out of one part. How long do we have to wait? 20-30 years? Imagine the possibilities.

From what I know of stereolith, a big machine is 2'x2'. I don't know how big they get. Pretty tough to build an entire body of it, though it would be cool.

What Matt (onewhippedpuppy) says has some truth to it, but in actuality, any part expected to be mass produced won't have those closed-in geometries. That is, if it can be molded, it can be machined. (or at least the mold can be machined)

The advantages that I see in these layering RP machines are, first and foremost, the easy of set-up. That is, you don't have to be a machinist make a part with one of these machines. Secondly, you don't have to source a big block of material. Those RP machines are like easy-bake ovens ...or a copy machine w/ toner. You have a vat of material. ...the part made uses only the volume of material of the part --not of a block-- so , little is wasted.

The major downside tho, is the materials of RP machines. The parts are often fragile, when made somewhat accurate ... and fairly in-accurate if made from the stronger materials.

The laser sintered metals (SLS) are pretty cool in that you could make, say a wheel that looked just like a Fuchs (in whatever size you wanted.) It would be metal, heavy, have a worse surface finish and weaker than even gravity cast aluminum. ;) ...but at least it's expensive. :cool:

Very true, but it is cool in that it can simulate multiple parts joined together, but created in a single prototype.

This is my first RP, I am using Redeye RPM from Stratasys out of MPLS MN. Nice folks to deal with. I am excited to get my project next week.
I can see so much potential in this process, since I am a sculptor, this will allow me to make some amazing stuff in the future. 3D scanners have also peaked my interest. The ability to scan one of my sculptures and then mass around with a 3d model will be a lot of fun.
I want to try and make a set of H4 repros for my 911, when I get some extra time and cash, maybe produce them in Kevlar or some other lightweight material.

I've had a 3-D scanner for years. Let me strongly suggest you not mess around with digitizing your sculptures. It's a real PITA to get a 3D model of swoopy surfaces that can be 'played with.'

...It's one of those things that sounds cooler than it actually works out to be.

http://www.zcorp.com/products/zscanner700.asp?ID=1
This one claims to be pretty accurate. I am having trouble believing what I see in the demo video, but I would like to try it out.

btw, here's is an example of a 911 motor I digitized a few years ago.

http://forums.pelicanparts.com/uploa...1188493233.jpg

That swoopy engine shroud took the most time.

THose accuracy claims; you've got to ask "over what distance is that accuracy?" Note 5axis' link. They have to have real accuracy for their customers. ...they use a digitizing arm.

wow, nice work man. Did you make any RP parts from that?

nah, too many other (paying) projects. ;)

Island, nicely done. What program?

Here's a few links that show what kind of parts we made and examples of the 3D plaster casting pieces and also the fero arm I used. You can see that trying to machine something like a centrifugal impeller from a solid billet would be nearly impossible or at least very impractical.

http://www.sturm-inc.com/PortalData/16/Resources/brochures/PatternDevelopment_.pdf
http://www.sturm-inc.com/PortalData/16/Resources/brochures/Engineering.pdf
http://www.sturm-inc.com/PortalData/16/Resources/brochures/QualityAssurance.pdf

I have a couple of those plaster 3D printed patterns in my desk, one is a bolt complete with nut and they screw together very well. The other is a ball bearing complete with inner and outer race and balls. It spins just like a real metal bearing. The tolerances they can hold during the printing process are really amazing.

You should check out that Atos system. It is really pretty sweet. From a machining standpoint I would say that it is usually within 0.002

I would say it works out cooler than it sounds. Lots of nerd talk.

http://forums.pelicanparts.com/uploa...1188496453.jpghttp://forums.pelicanparts.com/uploa...1188496871.jpg

I took a look at th ATOS system online http://www.capture3d.com/products-ATOS.html
I like what I see. What is the name of that milling machine? It looks really cool, is that Aluminum?

Last Company I worked for in China could make parts up to about 2 foot by 2 foot. Not sure if there is any limits on the size of the machines. One great advantage is any shape, even hollows can be made. Great stuff for prototypes and one offs.

Yeah, the full scan things are cool, but the problem is all that info, all that massive info, is not easily converted clean surface data. (and then 'solid" data) Some improvements have been made recently ...but still lacking, IMO.

I'm not saying it's useless (or close). Just that for a sculptor, I would suggest waiting for this technology to become more mature before spending too much time fussing with it . .. chasing what you expect it should do for ya. :)

oh, and the above was done with a MicroScribe (arm) various engineering measuring tools, SolidWorks (CAD). engine and parts borrowed from John Walkers Workshop.

I would have guessed CATIA, so I was close.

right on SmileWavy about the massive info, the tiangle files are friken huge. We often mix in regular surface models when making changes to existing items. I have not had any experience with converting them to something more tidy. That is not really a issue in my type of work. I do trust the scans much more than our ferro arm (boat anchor).

This technology is really pretty cool stuff and it is only getting better. (until the chinese rip it off and resell it at $0.25/1.00)

and that is a whole nother issue :D

I am really loving the RP, the first three projects I have worked on have increased my profit margins substantially......

The operative word is "prototype". The stl prototypes, when finally approved, are generally used to make molds for casting, as a model for parts to be machinined, or as a visualization tool for whichever process is most amenable to the number of parts to be produced. To produce a thousand parts using stl would most likely be cost prohibitive compared to alternatives and the media applicable to the process is limited. Unless things have changed recently, that's my understanding.

Great thread guys, I'll have to re-read it when I have more time.SmileWavy