Has anyone got any experience with using rapid prototyping / 3D printing for creating their own customs?

-What software do you use?

-What file format do you export into?

-What rapid prototyping method do you use? sintering / CNC / whatever?

-How much tolerance do you put into plugholes and tabs and whatever else on the parts?

-How did you come up with your transformation mechanics in the first place? (Assuming you've designed your own complete TF)

-What's the best way to get the individual components produced ... 40-odd separate pieces or one-two super-complex parts with all the itty bits on a sprue?

 

All I know is it's VERY expensive.

 

I've used a 3D printer...I'm a mechanical engineer. I have access to a 3 million dollar, unlimited geometry sintering system. But it's 3 million dollar.

The products I proto end up quite strong, becuase the starch used can be impregnated with ABS/nylon/polypropelene/certain rubbers to change its mechanical properties...I can rapid prototype a shock boot or a hing, something felxible or rigid.

You need a 3D model, and usually a step file. I use ProE, Inventor 9/10/11, and Ansys Workbench 10.

Rapid prototyping really isn't at the "do it yourself" stage yet....maybe in a few years.

 

Thanks AutobotEngineer - that's a fantastic start.
Knowing there are polymers that can be added to sintering starch is a huge help ... was starting to wonder if the finished compound (described on one site as 'plasterish') would be strong enough for producing a TF toy.


When you say DIY rapid prototyping is still a few years away, what do you see as the reason Joe Public isn't ready? Is there something in the process that isn't quite refined enough yet ... Joe Public's not sensitive to the engineering principals and constraints involved ... or Joe Public's just a plain idiot?


Wanting to know about unexpected engineering constraints is probably the leading reason why I'm posting these questions.
I mean, I've done a bunch of reading on different production processes, and learnt about some physical constraints (eg, sloping [2degrees] internal walls and flow lines for injection moulding, >0.7mm wall thickness for sintering etc etc etc), but it's always going to be the enemy you don't see what's gonna kill ya.

 

The other question I really should ask...

What are the limitations of a 'garden-variety' sintering system, compared to an unlimited sintering system?

 

I think the reason it's not ready for Joe Public is because the technology is still out of the feasable price range that Joe Public can afford. I would LOVE to do some 3D TFs of my own design and just be able to print them out. Last time I looked one of those machines up online, it was like $24,000. And that was the low end model.

 

I NEED one of those.

Maybe you could start with a machine that does 2D cut-out type stuff, just to start...if you're resourceful, you could probably build up a 3D form from 2D pieces.

 

The thing to do would be to get a group of people and rent/lease one. Some of the companies have try out times for a limited amount of time... during which you could churn out a bunch of parts (or as many parts as time allows)

I was looking to find one to make new fins for my lawn darts.

 

There's a reason they're for prototyping, folks... These things simply aren't capable of even the lowest level of production. While a small group might be able to purchase time on one, you would, at best, have a master to use for casting, but that's about it. There can be alot of trial and error involved if you're making something complex. If you have a definite pattern ready you'd be far better off having something made at a machine shop, especially if you have the expertise in CAD/CAM packages. At least then you're left with a solid master, and if the machine shop screws up it's their dime, not yours.

Anyway, my company just pitched a rapid prototyper. It was twelve years old, and virtually worthless. It was capable of running the demo "model," which was a little sailboat, but that was about it. These things were in their infancy when the company bought it, and they certainly paid the price. Whoever ran the thing certainly did, as the two-part material it used was later found to be highly dangerous and pulled from the market. In any event, it was incapable of using any other material, and was therefore a massive doorstop.

 

The thing that keeps them out of Joe Publics hands is really just cost. They're pretty complex devices, and require that they be fed healthy cad models. If you tell it to build an assembly that has errors or conflicts...it'll give you a pile of trash.

One of the things that keeps them expensive is the lack of demand. Sure, we all want one...but not that badly. Everyone needs to print 2D stuff, but in reality, nobody has much use for a 3D printer. The stuff you can make doesn't do much...it often times wouldn't survive molding if you wanted to cast copies of your prototype, and it certainly wouldn't last if you actually tried to make a homebrew transformer with it. Big companies buy 3D printers becuase for very complex parts, its cheaper than machining, but not by much. You can't do a lot with the end product...just look at it and make desicions regarding where to go wi th the design.

As far as the difference between limited and unlimited starch sintering printers, a limited printer needs to build your part with internal supports. For example, if you wanted to make a arced structure, like a model bridge, it would look like your cad model, but full of little beams and columns. The starch needs the support as it is laid down. You cut them out the extra parts by hand when it's done. These printers use what is mor eor less an HP inkjet head spraying a CA glue onto the starch, so the resolution is only ok. An unlimited printer takes a large tub full of starch, and builds the part inside of the starch. Any arches or hollow parts that need support are supported by the starch as the part is built, so the finished prototype has no extra supports...just pull it out and shake off all the extra starch. The resolution is higher on these printers, and they can really do nearly any shape you want. Fun stuff.

 

... It's probably worth pointing out that I ain't planning on buying the machine, just going thru one of the online Rapid Prototyping mobs to get stuff printed.

AE, thanks for the clarification on limited v unlimited sintering. I had read about unlimited sintering (but simply called sintering) on one website - it had been the only website to talk about the use of unactivated material to support the objects being printed. It's good to know about the difference.
I reckon unlimited sintering would be the way to go if the choice was available!  

 

What software do you use?
solidworks, zprint, insight


-What file format do you export into?
.stl


-What rapid prototyping method do you use? sintering / CNC / whatever?
starch, fdm (sintering)


-How much tolerance do you put into plugholes and tabs and whatever else on the parts?
depends on the kind of machine you have.

-How did you come up with your transformation mechanics in the first place? (Assuming you've designed your own complete TF)
reference of existing 'bots


-What's the best way to get the individual components produced ... 40-odd separate pieces or one-two super-complex parts with all the itty bits on a sprue?
if you are printing in 3-d you want to keep the cubic inches as little as possible. it can print ball joints, etc.... so you don't have to put anything together, keep the HEIGHT low and the WIDTH as large as the machine you are working with.


cubic inch cost is $11 for abs plastic $5 for plaster, the finishing on both types of prints on the machines here are VERY bad and needs a lot work before they can be casted/painted. the warpath above is a good example of the grainy finish the parts come out with.

 

I am also a mechanical engineer and also have experience with rapid prototyping. The machine I worked on in grad school was the Stratasys FDM 3000. Basically, it took your 3D design, which needs to be in *.stl format, chops it up into layers, then forms the layers on the machine by tracing the contours and filling them in. It does this by using two different plastics, one being a support plastic that can be chipped or dissolved away after production, and the actual model plastic. Both of which are ABS. The cost of the machine with all the parts when Rutgers first got it back in 2002 was $100k. So it is cheaper to buy it but still not chump change. Nowadays, I'm not sure what it goes for. Check out the website for more machines like what I described.

www.stratasys.com

To answer your questions:

-What software do you use?

Insight...whatever the version is and any program that can create *.stl files.

-What file format do you export into?

*.stl

-What rapid prototyping method do you use? sintering / CNC / whatever?

Stratasys FDM 3000

-How much tolerance do you put into plugholes and tabs and whatever else on the parts?

Generally none. The plastic is so robust that I can just sand it to make it fit.

-How did you come up with your transformation mechanics in the first place? (Assuming you've designed your own complete TF)

Sorry, I never did a TF on this.

-What's the best way to get the individual components produced ... 40-odd separate pieces or one-two super-complex parts with all the itty bits on a sprue?

In my experience, just the pieces are necessary, the injection molding enginner then figures out how to pattern them and where all the sprues and runners go for optimal efficiency. But for rapid prototyping, I don't need to worry about sprues and such, just give me the *.stl files and I could go to work.

 

Interesting topic. I've been wondering about this technology in regard to making TF parts or customs. I read somewhere a few years ago that 3D printers for the general public are in the works, though I'd imagine they'd still have quite a way to go from what I've read in this thread.

Thanks to the engineers here for sharing the info! Very interesting.  

 

I agree. There's gotta be some sort of a business opportunity for someone to pick up where Transrepro finished up with this emerging technology.
How cool would it be to design and build a decent looking face for G1 Scorponok??

 

Okay, from everything I've read and seen here (thanx WeeWoo for the Warpath pic - it's very cool), am swinging away from using sintering / FDM because of the bubbly surface texture it produces.


Does anyone have any experience with stereolithography?

 

very informative topic.

hey weewoo, care to elaborate on that piece?

 

Yes, I have experience with stereolithograpy and I can tell you that it is very expensive. The way it works is that you have an Ultraviolet radiation-sensitive polymer that you shouldn't touch with your bare skin while it is in its liquid form due to harmful properties. This polymer sits in a vat and a little elevator that moves up and down sits in the vat. Just like the FDM, you first need a *.stl file and the right program...I believe it's still Insight. And just like the FDM, the program will chop up your 3D drawing into many layers.

These layers are then traced out in the UV sensitive resin with a UV laser. Once a layer is created (the laser hardens the polymer just enough to retain its shape), the elevator moves down and the laser traces the next layer. Once the part(s) is(are) made, then you must take them out and very carefully remove all of the support structure, which is a messy, pain-in-the-ass since you need a chisel and a sink full of isopropyl alcohol. And it is extremely likely that if you're too rough, you'll break your part. We're not done yet though, we then have to let the part(s) bake in a UV oven to fully harden the polymer, then you have your part.

The great thing about stereolitharaphy or SLA for short is that it is incredibly accurate. Afterall, it is traced with a laser. The robotics people at Rutgers loved this machine for all of the tight tolerances required for their robots. The biggest problem is cost. SLA is very, very expensive. The resin, last I checked, is about $300 to $400 per gallon. The machine itself is around at least $200,000. The laser also burns out fairly quickly and I think that cost $10,000 or so to replace it. The other problem is that since you're tracing the parts out with a laser, all of the layers a super-thin so you'll need thousands (literally) of layers to make even a small part and that takes time. If you have a blackout, your part is gone and you have a mess to clean up and you'll have to start all over again. In my opinion, SLA just ain't worth it if you're making TFs.

Here's a website that RU dealt with when acquiring their SLA machines: http://www.3dsystems.com/

On a side note, I made parts for guys studying the fluid dynamics of flying missiles with the FDM, and usually what they did is after the parts are made, they would cover the parts with a coating so that they're nice and smooth. What that coating is I couldn't tell you.

 

Stereolithography can do some really impressive shapes. We made a Rook (the chess peice) with an internal spiral staircase, unsopported. It was clear plastic with brick detailing on the outside. Neat machines.