3D Printing to injection moulding

[nebbian]

I've seen some posts from some people on here that seem to know about milling and injection moulding, so was hoping for some advice.

Basically I've got a product that is broadly similar to lego, in that it interlocks like Lego does. The pieces are for a game that has been in development for many years, and we're ready to take it to the next level.

I've gotten the interlock to work well using my Delta 3D printer, but at around 30 minutes per block to print, it's painful to print up a full game set. A full set takes almost two weeks to print (daytime printing only).

So we're considering injection moulding. As far as I can tell, the only real way to get this working well is to have a desktop CNC machine to machine the aluminium blanks, and then have some form of desktop injection moulding machine to pop out a set. I would need a CNC machine at home because I can see that I'd need lots of iterations to get the interlock working well. If my 3D printing experiments are anything to go by, then it will be a slow and tedious process to get a nice, reliable lock. This puts commercial CNC machining well out of the question.

So what are my options for:
a) Desktop CNC machine that doesn't cost the earth, but still has good accuracy? Size of milled part would be, at most, 200mm. How many axes would I need? I think only 3? I've never milled before.
b) Desktop injection moulding machine

Would something like the Othermill be accurate enough to mill aluminium injection moulds?

[TheRealRocketBurns]

I have heard good things about these guys:
http://www.bulldogcncrouters.com/

[DavidF]

A 200 mm lego?? I think you may need a fair sized injection molding machine for that, desktop most likely out of the picture. What is the actual volume of the part and its dimensions?

[nebbian]

The max size of a "Lego" piece is 100mm x 15 x 15, and it's mostly hollow so volume would be only, say, 5ml. I thought that the mill would have to be able to face the entire mould though, hence my 200mm figure.

[mhackney]

Don't overlook plastic casting. It's very economical for small (<10000 part) runs, minimal investment in molding and lots of options for castable materials.

[mhackney]

Some additional thoughts...

I think you'll find designing the tolerances for interlocking will be MUCH easier since the dimensional accuracy of machining and injection molding is far better and more consistent than 3D printed parts.

Next, an injection molding machine to produce parts of that size is well beyond the desktop machines. Injection molds for parts that size are not trivial to design and really need to be made in steel and hardened for reasonable production runs. Also, injection molding is not the best technology for hollow parts. Blow molding or casting are both excellent for hollow parts and molds for this technologies are much simpler and cheaper - especially casting molds.

Unless you have an immediate business case to scale up to 10,000+ parts I would suggest exploring casting.

[Captain Starfish]

Or maybe roto-moulding.

But a thin walled 200mm part will need an injection machine about the size of an SUV.

If you can find a local-ish injeciton company, it might be worth sorting out an NDA and talking to them about what's involved in jobbing it out. Brace yourself, though - moulds aren't cheap, setup isn't cheap, it only pays off when you start doing bulk quantity.

[nebbian]

Thank you all.

By "hollow" I think I really mean "thin walled". A lot of this terminology is pretty new to me

I've tried plastic casting before using Alumilite resins, and RTV silicone moulds. I found it messy, unreliable, expensive, and generally not a viable option. At least I can set up a print and know that in 4 hours i'll have 8 pieces. With casting I often had portions missing where the plastic hadn't flowed where it should have. Perhaps if you could inject the plastic under pressure it might help, or use a vacuum chamber or something...

The properties of the plastic are important as well. ABS has the right hardness and flex for the application.

Hmmm. Food for thought.

[Outsider]

http://magicmaker.ca/#revo11

[SEBRET]

http://magicmaker.ca/#revo11

Not that I see anything wrong with their sales pitch, but it just bothers me that they are coming at it from the "3d printing led to this" angle. If I'm not mistaken (which i may very well be), rotocasting has been around much longer than 3d printing. I just hate to see it propped up on printings shoulders like it's not something that can stand on its own. </rant>

[Xenocrates]

If it's rotational casting we're talking about, I'd prefer say, the RotoMaak, which I've seen working in person, and which seems much less needlessly complicated. http://rotomaak.com/

[nebbian]

RE: rotational casting. There really isn't any way that rotational casting would have the pressure needed to make accurate lego pieces, unless I'm missing something. It needs reasonably high pressure I think? Sure rotational casting might work for something like a rubber duck, but I'm talking about interference fits here, and small parts. I was distinctly unimpressed with the quality of the parts produced by that magic maker -- the mould halves were way off. Didn't they even have location pins?

The wall thickness is between 0.8 and 1.2mm for the parts, at least in the 3D printed version. I've also got some slight overhangs to help with the 'snap' as the parts click together. I know that injection moulding doesn't handle overhangs at all so those will have to be left out of the moulded version, unless I can come up with some clever ways of getting the part out of the mould.

What sort of pressure would I be looking at for moulding a basic 4x2 lego brick? I'm talking about a part that has thin 1mm walls, about 2 ml of plastic, measuring 32 x 16 x 10mm.

Would a manual machine like this be capable of doing what I need it to?
http://www.injectionmolder.net/

[stonewater]

well steve wygant http://www.partdaddy.com who makes the rostocks also has a kick ass machine shop and does injection molding as well.


Tom C

[Xenocrates]

I only mentioned the Rotomaak because I didn't much like the roto-caster that was linked, and wanted to plug one I liked better. As far as locating pins, looking at their mold in the background, the answer is not at all. I almost wish white metal was still a viable option (Pewter and suchlike, not lead), as spin casting is a fairly easy and cheap way to get good molding force, reasonable accuracy and surface finish, and low production times.

Lego molds (The professional, official ones) apply between 25 and 150 tons of pressure. That is far beyond the capacity of the machine, even for the low end. I believe it is unlikely that you would be able to match, or even reasonably come within an order of magnitude, the quality of lego bricks. I work routinely with Haas mills, and our usual standard for precision is ±.0005 for most parts, in inches. For comparison, lego block molds are manufactured to ±.0002 MM. That's .00007 imperial. Of course, you don't need the same level of precision, since they make millions of bricks with a single mold. But most affordable CNC's have issues with high precision stuff (A Tormach 770 specs at an incremental deviation of .0006 per ft., enough to swallow the lego mold tolerances in a single move, without even accounting for tool wearing, wobbling, and that's for a nearly 7K$ machine. Something like a bulldog or a shapeko will have serious issues with that sort of precision). I can ask one of the people I know who's family makes injection molding tooling what sort of tolerances they usually use, ETC, but if you want to injection mold reasonably precise parts, the molds need to be quite a bit more precise than the end parts. You could probably ask partdaddy about molding, but as he's no longer taking on new injection molding projects, you can't go to him to make it, unfortunately.

And if these are game pieces that snap together like legos, perhaps consider using actual lego's as mating surfaces (Or megabloks, if you're cheap). It's much easier to design in pockets that you can then solvent bond parts into and mass produce that way, and you gain compatibility with other lego stuff. It does add cost, and assembly/packaging complexity.

I'm also interesting in seeing the actual parts and designs being discussed, as otherwise I'm shooting in the dark a bit.

[nebbian]

Thanks Xenocrates, that's very helpful.

The 3D printed parts are accurate to within about 0.01mm I think. It's hard to measure. All I have to go on is that I can change the extrusion multiplier, and there is a range of about 2% where the parts fit nicely (on a 0.4mm extrusion width). Any more and they stick too hard, any less and they start to fall apart. The nub is around 3mm wide. I don't know if you multiply that 2% extrusion tolerance by 0.4mm or 1.5mm. Also the extrusion multiplier will affect the receptacle wall thickness as well (0.8mm). Anyway that's what I know. I think that a tolerance of 0.01mm on the part would be acceptable.

I've worked very hard on my mini delta to get it accurate and repeatable, and even so I need to print close to the middle of the bed to get good results. I think I'm really pushing the limits of my machine

The receptacle design is also quite tolerant of... well... tolerances. It is as good a design as I can come up with, it really tries to be as tolerant of bad accuracy as possible. Bear in mind that my mini printer uses printed corners...

So if we go with a part accuracy of 0.01mm, is this possible with desktop technology? I'm used to pushing the limits of cheap gear

[U.S. Water Rockets]

I've tried plastic casting before using Alumilite resins, and RTV silicone moulds. I found it messy, unreliable, expensive, and generally not a viable option. At least I can set up a print and know that in 4 hours i'll have 8 pieces. With casting I often had portions missing where the plastic hadn't flowed where it should have. Perhaps if you could inject the plastic under pressure it might help, or use a vacuum chamber or something...

Alumilite casting is an art form in and of itself, and there are really good people out there who know how to avoid the issues you are having. You're getting voids in your castings most likely because you have undercuts and air is trapped in there. You need to carefully plan multiple inlet and air outlet holes for a successful pour, and sometimes a vacuum chamber is helpful. A pressure chamber is nice if you're doing transparent parts like lenses, etc.

But it sounds to me like you might want to try one of the low volume mold companies. Companies like protomold.com can do low cost molds from a soft material like aluminum for low-ish volume production and the prices are fairly low. You can get the same thing for a bunch less if you go offshore and deal with those hassles.

On the other hand, you might be able to get everything you need in short order if you were to start a crowdfunding campaign.

[mhackney]

Injection molding plastics comes with its own set of challenges - especially shrinkage. This has to be accounted for in the mold design AND also the injection parameters and even the plastic source and properties (like color). I spent 4 years injection molding early in my career and it is an art in and of itself. I recommended casting because I also have first had experience molding complex parts via molding and it has way fewer challenges. Shrinkage is almost non-existent with many of today's resins. There are some incredible resins that have properties remarkably similar to ABS. I know several people who have manufactured some very complex cast parts for their products. I have one of the KickStarter RotoMaaks and it is a nice simple machine that can reproducible create thin walled parts to well within the wall thickness tolerance you listed. No presser or vacuum required.

But, we all digress...

You really need to understand and talk to experts (most likely none of us!) by showing them your parts and discuss your requirements for annual parts count, any product requirements (i.e. color, ductility, surface finish, etc) and price. Then call/email a few specialty shops - injection molding and casting - and get some quotes for your parts. They you'll have some idea what professionals charge and, more importantly, recommended process and materials for making these. In today's world you do not need to restrict yourself to local shops. Then, armed with this information you can determine whether outsourcing or developing the manufacturing technology in house makes sense. If it were simple to mold 100s or 1000s of parts, everyone would do it!


I work with a shop in Tx to make a part that they can make much more economically than I can.

[nebbian]

I recommended casting because I also have first had experience molding complex parts via molding and it has way fewer challenges. Shrinkage is almost non-existent with many of today's resins. There are some incredible resins that have properties remarkably similar to ABS. I know several people who have manufactured some very complex cast parts for their products.

Thanks for your input, I really appreciate it.

Can you point me in the direction of where to look for those resins? Also would you recommend RTV moulds, or something harder?

Regarding injection moulding, the kickstarter campaign is definitely the way we've been planning to go, but I'm trying to figure out what the actual costs will be before making the video.

[Xenocrates]

I got a chance to talk with some folks at Reynolds advanced materials, the guys behind smooth on, they will get you an account manager for relatively small orders (a few hundred) and if you can get an engineer interested (which Isn't hard, they like printing) they will handle the account without a sales droid for the first few years while you get set up and stay with it afterwards. Find them at a trade show and chat, maybe get materials samples as well.

[RandomFactor]

You could use your 3d printer as an injection molding machine. No reason this couldn't work for larger parts:
http://www.instructables.com/id/Making- ... 3D-Printe/

[Eric]

You could use your 3d printer as an injection molding machine. No reason this couldn't work for larger parts:
http://www.instructables.com/id/Making- ... 3D-Printe/

That's quite ingenious, but it only works because the part is quite small.

As you scale up, you need higher injection pressure to get the plastic to fill the mold completely before it cools. Which in turn requires a sturdier mold that can handle the pressure. And something stronger than a hand to hold everything together during the process.

[RandomFactor]

Might be interesting to see what you can get away with. Putting a real mold on the heated bed to keep everything flowing, larger nozzle for a faster fill, etc. Just the idea that you could multi purpose the printer is intriguing.

[barry99705]

Might be interesting to see what you can get away with. Putting a real mold on the heated bed to keep everything flowing, larger nozzle for a faster fill, etc. Just the idea that you could multi purpose the printer is intriguing.

Unless you can get your bed temp to the same temp as the extruder the plastic will still freeze up in the mold.