Looking for info to get started with injection molding

[Max@Mac]

Hi all,
I'm currently looking into getting a small injection molding machine to make consumable sandblasting fixtures in bulk. They will be for use inside the company, so I'd like to see if I can make the molds myself. The fixtures could be modified to have draft angles very easily, and ABS seems to be the closest analogue to the PLA+ I've been 3d printing the molds in, but each pair of printed parts takes approx 3 hours. Those numbers are especially inconvenient when I'm being asked for 24 pairs of fixtures, and each individual part has a different sandblast fixture. From my 3d model, the sandblast fixtures are in the neighborhood of 50cm^3 to 70 cm^3.

Would a BOY 50E be a good starting mold machine (76.5cm^3 shot size according to website)? Or should I err toward a larger shot size in case I need a larger part made someday?
We do have cnc machines in the factory as well as a tool room, so simple fixtures in aluminum should be feasible.
Would molding simulation software be worth it?
Is UHMWPE an option for injection molding? The sandblast fixtures were originally machined out of that before I switched to 3d printing in PLA+.
Would I need some kind of plastic pellet dryer to feed the machine from? If so would it be best to buy or design my own?
How do you decide how large to design your ejection pins?

I'm aware injection molding is pretty complex, but I'm also looking at this as a learning opportunity, since my company does do injection molding in a different one of our locations, and it would also be an exercise in machining.
Any info you could give me about getting started would be much appreciated.

Thanks,
Max

 

[jhov]

I'm still just learning injection molding myself, so I won't try to answer things I don't know, but I can address a few of your more basic questions.

But first, does your part need to be solid? Injection molding works best with parts with uniform wall thickness. Non-uniform and thick walled parts are subject to defects such as warping, voids and sink marks. Can you redesign the part to be shelled out? This would also reduce the material used and size of the injector required.

Typically you want to use 25-80% of your shot volume for your part for standard plastics (PP, PS, etc) and 25-60% for engineering plastics (PC, PA, ABS, etc). Also, in addition to shot size, the length to diameter ratio of the screw is important to ensure the material is fully mixed and melted uniformly.

You'll need to calculate the clamp force needed based on your part dimensions.
Clamp Force Required = Molding Pressure x Sectional Area x 1.3

The 1.3 is a typical factor of safety. For ABS, a typical molding pressure is between 300 - 500 kg/cm^2.

You will need a drier for ABS.

Making the molds is another can of worms entirely. There are several guys on here with heaps of experience with that so I'll let them chime in. I imagine you'll get a lot of "you're in over your head".. which is probably true (it certainly is for me).. but don't let that discourage you if you have the drive, resources and opportunity to make it happen.

There is surprisingly little useful information about injection molding on the internet. Here are some books that I'd found to be helpful.

Injection Molding Handbook by Donald Rosato
Injection Molds and Molding by Joseph Dym
How to Make Injection Molds by Menges/Mohren

 

[Max@Mac]

Thanks for the response.
I'll need to make some prototype 3d printed parts to see if a thin walled geometry more suited for injection molding will work with the sandblaster and not just disintegrate.
To find sectional area, would I find the largest cross sectional area in the part and use that for the required clamp force?

 

[jhov]

Thanks for the response.
I'll need to make some prototype 3d printed parts to see if a thin walled geometry more suited for injection molding will work with the sandblaster and not just disintegrate.
To find sectional area, would I find the largest cross sectional area in the part and use that for the required clamp force?

Not exactly. You need to find the cross sectional area normal to the platens. What you're calculating is the force required for the clamp to overcome the high injection pressure the plastic will be imparting on the mold trying to open it.

Force = Pressure x Area

So clamping force required to overcome the injection pressure is Pressure inside the mold * Cross sectional area normal to the platens * Factor of safety = Clamping force to keep the mold closed and sealed during injection.

 

[RC Mech]

Projected area is where you get your clamp tonnage. Always go up to the next size machine. If you need 80 tons better to go up to a 120 ton press.

I’m not convinced at all that part needs to be injection molded. What about a plastic extrusion? Extrude and pop holes in it.

3D printing doesn’t scale. Not the right application. Injection molding is only for scale, potentially not the right application. Unless you’re doing 20,000 of these a year I would not attempt to learn mold making from scratch for this simple of a part, when there are better alternatives.

 

[Houdini]

Oh boy

You are probably not going to want to shoot that solid, and I'm sure for what it is, it doesn't need to be.
just a FYI rubber seems to last longer in sandblasting particles seem to bounce instead of tearing down the material,
but I don't know what strength in rigidity you need.

Are you buying a new machine?
If so I am looking at these machine due to cost, I know a owner who has 20 of them.
if not TOYO would be my second choice.

 

[Houdini]

Projected area is where you get your clamp tonnage. Always go up to the next size machine. If you need 80 tons better to go up to a 120 ton press.

I’m not convinced at all that part needs to be injection molded. What about a plastic extrusion? Extrude and pop holes in it.

3D printing doesn’t scale. Not the right application. Injection molding is only for scale, potentially not the right application. Unless you’re doing 20,000 of these a year I would not attempt to learn mold making from scratch for this simple of a part, when there are better alternatives.

+1 on this, with the volume he could macheen those from solid stock at those volumes, cheaper faster, and solid. maybe not @ 20,000, but unless you going to make a couple thousand.

 

[TeachMePlease]

So, call me crazy here... But your company already does injection molding in a different facility... Can you not just talk to the guys there to get some good info? Maybe even make your own molds and have them squeeze you in a run of 200pcs or whatever will last you for a year?

 

[Houdini]

Would a BOY 50E be a good starting mold machine (76.5cm^3 shot size according to website)? Or should I err toward a larger shot size in case I need a larger part made someday?

As mentioned if shot size is close move up to next size machine, don't go way oversized, polymer heated in barrel for too long can degrade.
But larger machine can run a mold with multiple parts instead of one.

We do have cnc machines in the factory as well as a tool room, so simple fixtures in aluminum should be feasible.

Then why dont you machine them from solid?

Would molding simulation software be worth it?

No, and its very expensive, if you don't know what all the technical parameters are its worthless.

Is UHMWPE an option for injection molding?

Yes

Would I need some kind of plastic pellet dryer to feed the machine from? If so would it be best to buy or design my own?

ABS?, Yes, any hydroscopic polymer

How do you decide how large to design your ejection pins?

The larger the better, then it reduces the chance of part pushing issues, there are a few. Some say you can never have too many ejector pins, I disagree.

I'm aware injection molding is pretty complex, but I'm also looking at this as a learning opportunity, since my company does do injection molding in a different one of our locations, and it would also be an exercise in machining.
Any info you could give me about getting started would be much appreciated.

Thanks,
Max

This is a doable task, you can just buy a pre fab Aluminum mold and go to town, if it gets messed up or needs changed, that's what inserts are for.

Hope this helps

 

[plastikdreams]

Oh boy

You are probably not going to want to shoot that solid, and I'm sure for what it is, it doesn't need to be.
just a FYI rubber seems to last longer in sandblasting particles seem to bounce instead of tearing down the material,
but I don't know what strength in rigidity you need.

Are you buying a new machine?
If so I am looking at these machine due to cost, I know a owner who has 20 of them.
if not TOYO would be my second choice.
View attachment 438138

We use toyos for wax injection molding at work, pretty solid machines. Took some time to get them tuned to melt the wax right. We had to have custom screws made. $$$$

 

[pcd]

each individual part has a different sandblast fixture.

If you go the injection molding route that sounds like lots of molds and setup time.

What about casting these mask fixtures from urethane or silicone? Like @Houdini said, rubber stands up better to blasting than hard plastic. A silicone or flexible urethane mask might last a long time.

As a first step you could try making a silicone mold using a drafted version of the 3D printed mask, pull the print part out of the silicone once it's cured, and then pour resin into the cavity. I haven't tried the 3D printed aspect of this, so not sure about getting the printed part smooth enough to release. Vapor smoothing might help if you're set up for that.

A more advanced version could involve machining either an aluminum mold or an aluminum version of the mask to make a silicone mold.

It looks like you're clamping the 2 masks onto the parts with a pair of tie bolts. A metal bar on each outside end would prevent bowing if you move to a rubbery material for the masks themselves.

I'd try a silicone mask first and see how it stands up to sandblasting. If it wears out quickly then move on to urethane. Urethane is much cheaper than silicone, but it's a bit toxic and sensitive to moisture exposure during storage and mixing.

 

[Houdini]

We use toyos for wax injection molding at work, pretty solid machines.

Yeah Japanese machine, muy bueno

Took some time to get them tuned to melt the wax right. We had to have custom screws made. $$$$

Yeah I'll bet standard machine could just burn that shit into gas off Let alone viscosity differences for semi non-newtonian almost.

 

[plastikdreams]

Yeah Japanese machine, muy bueno

Yeah I'll bet standard machine could just burn that shit into gas off Let alone viscosity differences for semi non-newtonian almost.

Yep and it needs to have all the velocity and volume up front especially the larger 3 and 4 blade vanes...or else it solidifies where it shouldn't and doesn't solidify where it shouldn't and I get called out to pick wax out of tiny crevices when the operator pulls a liquid wax filled mess apart lol.

Funny note...when the techs come in to work on the machine they are escorted in and out of the machine which is shrouded in black tarps so they can't see anything.

 

[dstig]

I didn't get a feel for your mix on this: do you need a lot of the same fixture or a few fixtures of many different types? If the first, then injection molding MAY make sense, but if the latter, it never will. 24 is not "a lot," either. You'd be talking about needing hundreds or thousands at a time to justify molding...at the very minimum.

They make much larger scale 3D printers that can handle multiple parts at once to save some time, and they are often far faster on top of that due to better hardware, drives, controls, etc. Stratasys has some SUPER large machines...

 

[Max@Mac]

The internal injection molding we do is for solenoid coil cores, which are off the top of my head about 1" x 0.5" x 1.5", and those molds are made by an outside company for about $25k last I checked. I don't think their machine can fit a mold for a part that is about 8 inches long in the longest direction.
As for the parts, I believe my coworkers would like each part to have a set of 24 pairs of fixtures, and at a glance there are dozens of parts we already have machined fixtures for.
The fixtures are "clamped" together with rubber bands, but most of the gripping is a tight fit on the bores gripping the parts just so that sand doesn't get where we don't want it.
I'm not sure I need a brand new machine, is there any reasonable way to verify a used machine will work other than just seeing it run a part?
Can injection molding do any stiffness of rubber? It seems to me that if I got the flexibility right a thin walled fixture could be intentionally undersized to grip the part by stretching slightly and then not require as tight tolerances as a hard plastic part would.

 

[memphisjed]

I am not seeing the shape need for injection molding. 2 part plastic in a silicone mold with 3d printed master.
The 3d prints need smoothed. I do a few molds for a candle maker and found rough sanding/filing the big stuff then cheap sandable primer then sanding mostly with coarse 3m abrasive brushes on flex shaft goes pretty easy.
You do not need to draft with silicone molds.
48 parts is big difference from 4800 where injection wins.

 

[russmill]

If so I am looking at these machine due to cost, I know a owner who has 20 of them.

Hey I know him too! Great guy, really knows his business

 

[???]

Can injection molding do any stiffness of rubber?

Compression moulds for rubber. Various shore hardness ans silicone can be moulded. Easy to make moulds, basic heated platten press to mould with. Material availability is an issue as its normally sold in tons.

Old school but also adequate for low volume is dough moulding.

 

[???]

Yeah Japanese machine, muy bueno

Yeah I'll bet standard machine could just burn that shit into gas off Let alone viscosity differences for semi non-newtonian almost.

Look at Arburg, great machines. The ability to reposition the entire injection unit vertically is priceless. Lots of companies here run Haitian machines. One machine I wouldn't touch is a Fanuc.

 

[Houdini]

The internal injection molding we do is for solenoid coil cores, which are off the top of my head about 1" x 0.5" x 1.5", and those molds are made by an outside company for about $25k last I checked. I don't think their machine can fit a mold for a part that is about 8 inches long in the longest direction.
As for the parts, I believe my coworkers would like each part to have a set of 24 pairs of fixtures, and at a glance there are dozens of parts we already have machined fixtures for.
The fixtures are "clamped" together with rubber bands, but most of the gripping is a tight fit on the bores gripping the parts just so that sand doesn't get where we don't want it.
I'm not sure I need a brand new machine, is there any reasonable way to verify a used machine will work other than just seeing it run a part?
Can injection molding do any stiffness of rubber? It seems to me that if I got the flexibility right a thin walled fixture could be intentionally undersized to grip the part by stretching slightly and then not require as tight tolerances as a hard plastic part would.

A used injection molding machine is fine, we bought the TOYO ones we used as old as the late 80's for $20-30k but then you will need some way to repair them, the old ones do have issues quite a lot.
Yes you can injection mold some rubbers, soft and harder durometers.
Silicone requires a different type of machine, not a standard machine, different world.
After the number of parts you mention you don't want to injection mold them.
Just have someone machine them.

If you wanted only that many molded, make a silicone mold for making them out of urethane plastic or rubber or a rubber over rigid substrate would probably be better.
These guys are awesome give them a call or email with your situation if you want to make hand pour mold.

Mold Making & Casting Materials | Rubbers, Plastics, Foams & More!

Smooth-On manufactures industry-leading two-component materials such as silicone, urethane rubber, urethane plastic, urethane foam, epoxy resin, epoxy coatings and adhesives.

www.smooth-on.com

Also on the cheap, for shits and giggles, you could machine a MUD set frame, then all your molds would just be MUD set to go in it, that cuts cost.
make you a mold for less than $2k each.
Then give the frame and the MUD sets to a molder to run and cut the cost of the machine out, and the high cost of molds out building them yourself.
Still not cost effective for 24pr. but fun experience.

Here is a look at MUD sets
and a link for prefabbed Aluminum cheap molds
you could buy one Aluminum mold and copy it yourself, dunno if you could make one for less though, maybe slightly if you did all the work.

https://www.dme.net/mud/

PCS Company - Mold Base Standard Aluminum

PCS Standard Aluminum Mold Bases are precision machined from 7075 aluminum.

www.pcs-company.com