Injection molding machine - Can we do it?

I am looking for a way for us to bridge the gap between prototype and production run. Parts could be made from mixing resins and pouring into a mold, but I think there is a need for creating 50 or 100 of a given item for market testing. Molds are expensive and set up fees add up. One can spend a few thousand dollars on a mold just to bring about a saleable product, the viability of which won't be necessarily known until after the products are pushed to market.

I want to explore the feasibility of building a small (2 to 3 oz, perhaps?) injection molding machine capable of perhaps 5 to 6 minute cycle times.

First, I'd like to discuss the possibilty of using some high strength filaments to make a mold. The survivabilty of any injection mold will depend on more factors than I am aware of, but I am sure that injection pressure, injection temperature, and clamping pressure would have to be lower than what is seen in industrial injection molding with steel or even aluminum molds. For reference, I've seen "rules of thumb" that suggest mold clamping pressures of at least 2 tons of clamping force for every square inch of the part's projected area (note this is tons of clamping force on the platens I believe, not a tons per square inch). Injection pressures are also high, as in usually over 1000 psi.

What kind of material could an FDM printer use to have a decent shot of surviving some injections? I am thinking Taulman's Tritan or something similar for the mold (post processed to smooth out the cavity, of course) and a thermoplastic with a low melting temperature and relatively high melt flow rate as the injected plastic: Perhaps LDPE would work here. For what it's worth, I have seen where polyjet printed mold was made an put to use for injection molding (see [www.youtube.com] ).

A number of questions would have to be answered for a project like this to even be approached.

1) (Already asked, sort of): What is the highest temperature, injection pressure, and holding force that an FDM printed mold could withstand? Can injection molding take place reliably within these constraints.

2) If the constraints laid out in the first question could not be satisfied, are 3 axis prosumer CNC machines (modded Shapeoko types for instance) capable of cutting alluminum molds of fairly basic geometries with reasonable accuracy and resolution?

3) What type of delivery system could be used? I think the conventional method used in industy (a multi staged reciprocating screw) is out of the question. I would imagine a ram used on "hand pushed" type machines would be the easiest to implement. (see [www.injectionmolder.net] for reference).

4) What existing hardware and software could be modded to run this? I'm envisioning an arduino + grbl shield running hacked up marlin or grbl to actuate clamping, unclamping, injecting, and temperature control.

5) From a design perspective, I've envisioned using 80 mm alluminum extrusion framing to make a box with cross supports. A stationary platen could be at one end with the moving/clamping platen being driven against it by way of leadscrews. I think the ram could be actuated either by a strong solenoid or a leadscrew attached to the end of a lever. This would need significant force.

I've been thinking about this a couple of days. Before I expend any more time thinking about it, I'd like to get the thoughts of the community. Is there any interest in having something like this? Is it even technically feasible?

Edited 2 time(s). Last edit at 03/12/2015 01:33AM by DRobs86.

FDM printers won't be able to print a usable mold, as in the video you posted it needs to be made with polymers because of the way layers are printed.

Now you can print a mold and then cast it with a polymer resin and use that cast as final mold.

Since you planning on doing small runs, you don't need to follow traditional injection molding settings. You can grind the plastic pellets to a powder form and use the injector as your heating chamber too, you can circulate hot air in a previous stage to raise the temperature of the powder before going into the injector? You can rotate the injector at high speeds while heating the powder to ensure complete melting of the powder?.

You need though an hydraulic system to inject the plastic, no other setting will do the job at the speed and pressure needed (You don't want to be around when a leadscrew fails under pressure) and don't built the frame with aluminum, use thick plate metal welded by a kknowledgeable person, aluminum will stretch under pressure and each time you inject something it will require more and more pressure Until it fails. With a little bit of knowledge and patience you can built your own hydraulic system and will not brake the bank.

To clarify some other points, a simple Arduino uno with basic switching capabilities it's all you need to operate the machine unless you want to make a fully independent system. Your only parameter is temperature, and what you control is the on-off of the heating and hydraulic system and when they start and stop. A basic endstop switch on each end of the hydraulic cylinder stroke will establish length and duration can be control with timers, a solenoid valve will operate the oil supply and direction and a thermistor the temperature. Basic firmware can be modified to accomplish this since most or all of what you need is already there.

If there is a real need for something like that I can't tell, actually 3d printers take that part of the equation out as to small productions refers. So if your intentions are for small runs, why not just 3d print them?

At the end if needed or not will all depend on how many parts you need to make, time frame and mold duration. If a mold only last 10 to 20 runs and you need 300 parts, you will have to calculate the time it takes to print the molds and run the injections vs printing the 300 parts.

Thanks for the replies. Good advice on a hydraulic system. This is doable I'm sure.

Your thought on grinding the pellets is interesting. From what I understand of industrial injection molding, the pellets are usually melted by frictional heat as the screw turns, not by external heaters. Since I don't think a reciprocating screw is "doable" for homemade machine, it may help to preprocess the plastic.

Creating an injection mold from poured molds is interesting, but I have concerns about the heat deflection temps of the mold resins that I've seen. It seems as though the pressure resistance properties would be acceptable, but not susceptibility to heat. I'll have to look more into this.

I do not know if this will work but anyway - Print a hollow prototype. Mix ABS pellets and acetone to a consistency that can be injected or poured in the hollow prototype. Dry out the acetone by sucking it out in a vacuum chamber. Maybe worth a try!

---

You only learn when you change your mind.
[www.deltaprinter.co.za]

Quote

From what I understand of industrial injection molding, the pellets are usually melted by frictional heat as the screw turns, not by external heaters

The ones I have seen use heating elements attached to the screws housing. I was recently at an o ring manufacture and all their injection machines have them (most were old 80's or so models) , maybe new machines use friction heating?

I assume you've seen these? [www.techkits.com]

Would this limited run be sold as an actual product, or just used for consumer testing?

Also, it might be easier to build a CNC machine to make your own molds and a DIY injection molding machine to inject with than try to use an FDM mold.

---

greghoge.com

HUGE 3D PRINTER PARTS SALE!!!

Quote
gmh39
I assume you've seen these? [www.techkits.com]

Would this limited run be sold as an actual product, or just used for consumer testing?

Also, it might be easier to build a CNC machine to make your own molds and a DIY injection molding machine to inject with than try to use an FDM mold.

Thanks for adding to the discussion. I have seen those. While affordable and seemingly easy to operate, they have small shot sizes.I think the larger of the two machines is only capable about 1.25 cubic inches, or about 2/3 of an ounce, and is priced at 1500. There 600 dollars machine is probably great for what it does, but the quarter ounce shot size limits you to small nick nacks. I do appreciate you bringing it up though.

The product would be an "actual" product.... Or more correctly, probably something between a marketing sample and a professionally produced product. The idea is to be able to build some inventory of a given product that initial research shows being potentially viable within spending thousands to tens of thousands on tooling or professionally produced samples.

You may be right about needing a CNC. This in itself is a bit limiting though, in that I think you need rotational axes with subtractive manufacturing to achieve certain geometries possible with a three axis additive method. Maybe I'm wrong, but I think you are going to spend a lot of money on the hardware, control boards, and CAM software to achieve rotational axis cutting. Of course this isn't always a necessity, but sometimes it would be. In practice though, it might be required.

Edited 1 time(s). Last edit at 03/14/2015 01:13AM by DRobs86.

Quote
ggherbaz

Quote

From what I understand of industrial injection molding, the pellets are usually melted by frictional heat as the screw turns, not by external heaters

The ones I have seen use heating elements attached to the screws housing. I was recently at an o ring manufacture and all their injection machines have them (most were old 80's or so models) , maybe new machines use friction heating?

I'm the farthest thing one can be from an expert in this field, but I believe that in an industrial machine that uses a screw for injection, the heaters are just there to help maintain thermal consistency. I'm certain they are helping the plastic melt, however.

Quote
Xabbax
I do not know if this will work but anyway - Print a hollow prototype. Mix ABS pellets and acetone to a consistency that can be injected or poured in the hollow prototype. Dry out the acetone by sucking it out in a vacuum chamber. Maybe worth a try!

Intresting thought. I'm not sure that the acetone would be removed independent of the ABS. I think the vacuum would act upon the solution of abs/acetone, not on the acetone itself. I think by pushing or pulling air around it you help it dry, but with vacuum I think you'd just destroy/deform your pour.

Quote
DRobs86

Quote
Xabbax
I do not know if this will work but anyway - Print a hollow prototype. Mix ABS pellets and acetone to a consistency that can be injected or poured in the hollow prototype. Dry out the acetone by sucking it out in a vacuum chamber. Maybe worth a try!

Intresting thought. I'm not sure that the acetone would be removed independent of the ABS. I think the vacuum would act upon the solution of abs/acetone, not on the acetone itself. I think by pushing or pulling air around it you help it dry, but with vacuum I think you'd just destroy/deform your pour.

I don't think this will work well. Since you don't put 100% plastic in your mold, your structure will not be conserved once you remove the acetone. Maybe you will get a foamy thing, maybe it'll just shrink. I also doubt that you'll be able to recover all the solvent, meaning that you can have drops of saturated acetone trapped in your structure which will, again, cause structural problems.

Mold makers take a lot of factors into account when they make a mold.

you need exit holes, and the molds tend to be super polished, multi-part things for multi-run.

Lost PLA Casting is a decent idea until you see it done (youtube) and then you're like, uhmmm, no.

By the time you're done you'd be thinking, why didnt i just build an SLA/Clip style printer...

The barrier to the SLA printers is of course material cost. That stuff has got to come down in price..