DIY Injection Mold Design for the Home Shop

3D printing is great for prototyping, and not bad for limited runs of parts. Unfortunately though it really doesn’t scale well beyond a few pieces, so when you’re ready for the mass market you will need to think about injection molding your parts. But something like that has to be farmed out, right? Maybe not, if you know a thing or two about designing your own injection molds.

The video below comes from [Dave Hakkens] by way of his Precious Plastic project, whose mission it is to put the means of plastic recycling into the hands of individuals, rather than relying on municipal programs.  We’ve covered their work before, and it looks like they’ve come quite a way to realizing that dream. This tutorial by [Dave]’s colleague [Jerry] covers the basic elements of injection mold design, starting with 3D modeling in Solidworks. [Jerry] points out the limitations of a DIY injection molding effort, including how the thickness of parts relates to injection pressure. Also important are features like gentle curves to reduce machining effort, leaving proper draft angles on sprues, and designing the part to ease release from the mold. [Jerry] and [Dave] farmed out the machining of this mold, but there’s no reason a fairly complex mold couldn’t be produced by the home gamer.

When you’re done learning about mold design, you’ll be itching to build your own injection mold machine. Precious Plastic’s tutorial looks dead simple, but this machine looks a little more capable. And why CNC your molds when you can just 3D print them?

Thanks to [Hassi] for the heads-up on this one.

28 thoughts on “DIY Injection Mold Design for the Home Shop

  1. Uhh. There are plenty of reasons why a fairly complex mold couldn’t be produced by the home gamer. They would have to already have high end CNC equipment and knowledge of how to use it just to start with, even if you set aside the mold design software part of thing and started with finished CAD drawings.

    Setting aside the fact that you have to design a proper mold as well as deal with all of the other nuances of injection molding plastic, hand-waving this and just saying it can be done with a 3D printer is not just inaccurate but unprofessional.

    That all said, I would really like it if injection molding were to become easier to do and much less expensive. But unlike a mill, CNC molding doesn’t really scale down in the same way because the pressures involved don’t suddenly change when you scale things down, necessitating proper, high quality equipment that doesn’t really drop much in price unless you were to cut out basically everything and make it a few thousand dollars of nearly all manual work. At least the controls and such are getting easier to deal with and less expensive.

    1. Your typical home 3D printer certainly leaves a lot to be desired in terms of finish but get your designs printed on a commercial sls machine and the results are rather spectacular. I know of a few parts now been made this way in low volumes for the VW bus restoration scene.

      1. I would agree with you there. But SLS is still a slow and somewhat expensive process. It’s nice for custom quantity of maybe one to perhaps a few dozen parts. Maybe a few more. But it’s certainly not cheap nor quick and it is never going to compete with injection molding if you need greater quantities. It can certainly make sense if the parts are simply no longer available though and the market is somewhat limited in size.

    2. Although these guys referral service is interesting, there is always something to be learned by talking directly with tool and die makers.

      1. Molten plastic is nasty stuff, at minimum both gloves and a full face shield should be worn (burns look similar to napalm victims).

      2. A Sprue can be quite large most of the time if using a rubber washer seal… we often used a 1/4″ port for higher flow-rates and lower pressures.

      3. Heating the mold to just below the plastics melting temperature is highly recommended.
      Even pro shops will usually toss the initial few cycles as the machine warms up to operating temperatures.
      The surface texturing and 2mm thick limit for the article’s cold process means these guys need to do a bit more research.

      4. Epoxy composite injection molds require no machining, and are good for 800+ pieces.
      Aluminum is usually good for 8000+pcs, and EDM carbon steel can handle over 50000+pcs.

      5. The lip around the edge reduces overflow Flash, but allows air to escape as the cavity fills.
      These guys get a B- for the video ad, and a F- for suggesting it is easy to improvise these designs.

      The cost of a mold rarely makes sense for small batch parts, and designs normally include the maximum number of part copies possible specific to a factory machine’s per-injection-cycle (the time limiting factor).

      1. One time i needed a mould for inyecting Kraton, and the guy at the workshop told me that it was possible to make them with Epoxy, but i could’t find much info on the process. Can you give me advice or info about how to desing and make them?
        It is better if the mould has metal charge like steel or aluminum?
        Are the moulds casted from epoxy resin?

        The only advice the guy could tolme is to make a steel frame to avoid the resin to be crushed by the pressure of the injector

        1. For test runs, the mold halve generic cavity are usually a truncated chamfered pyramid shape to prevent the epoxy dies jamming in the cavity or cracking under pressure changes. For small <3" molds, the generic designs are usually a 1/2" to 2" thick walled structural aluminum box (expect $30 for the metal, and $200 for a machinist to rough it out on a mill).

    3. You dont need a CNC machine to make molds. You can make reasonable molds with hand tools using some old techniques. Scraping and Shaping is one of the oldest techniques and was done well before there were CNC machines. Admittedly it takes a lot more time but you can make a really nice multi part mold this way.

      You can also use the mold for injecting liquid resin in to.

      Another option for making molds is using the electro erosion technique, you can make some prety complex shapes this way with some really intricate designs. the process can give just as much detail as etching PCB’s

  2. Just FYI, Autodesk Moldflow Adviser Premium 2017 costs around $1,750 USD PER YEAR for a year long subscription. Are there even any open source or lower priced mold simulation products out there? This would just help get you part of the CNC design simulation done, setting aside all of the prior CAD work necessary before that. Again, extremely skeptical that somehow injection mold making has suddenly become a quick, cheap or easy thing.

    3D printing has made one off designs cheaper and that’s been a huge change. Smaller CNC machines (including lasers) have helped make the shorter run or makerspace types of idea evolution easier. But injection molding is something I really have not seen become easier or cheaper to do much at all, despite well over a decade of work by a great number of people. It’s something that needs to work right, be cheap and the downside of it needing to produce five figures of something to break even but once that’s done, the price per unit is almost trivial. It’s like ill informed people saying they would 3D print things like disposable cutlery or cups because they fail to understand any of the details about how 3D printing or injection molding actually work.

    That all said, if there are any examples that people know about regarding injection molding becoming cheaper or more home friendly and want to share, please do so! I would really like for that to change, I just haven’t seen much progress on that front at all. Despite the somewhat poorly researched or hyper inflated promises by the author. Yes, you can sort of 3D print some injection molds. On $100,000 printers. The molds might last for a dozen or two runs and the molds cost at least several hundred dollars and take 20+ hours to print. It works great as a final test before committing to a five figure mold run but it’s hardly a replacement. Sorry, but 3D printing is not exactly a viable alternative to CNC machining a proper mold.

    1. You can have stuff machined in China, even to reasonable tolerances and finish while being reasonably priced… the mold looks like it’s from aluminum (judging not only by the look, but also by the way they handle the completed mold, steel would not be so light to to move around like that), so the machining shouldn’t be all that expensive.

      Still probably at least a couple of hundred $$$ and shipping + the wait.

        1. Not personally, but I was witness to a China-made, mid-sized (slightly below 1t assembled) tool steel injection mold for a modern plastic injection machine… while several times cheaper, it also had to be sent several times back, because it had defects/missing features/didn’t work, so ultimately it cost the company slightly more then if they used a domestic mold maker…depends who you choose obviously, you (don’t) get what you (don’t) pay for…

          The small mold pictured in the video is…well…small… and made of aluminium, so there should be a lot more shops to choose from, possibly even “western” ones, if you’re willing to spend around $500 for it…

    2. When I was a young pup, I made a HO slot car chassis, similar to a Aurora car by making a mold from aluminum plates that I machined on a Bridgeport mill in my high school shop. The design was made on paper. This was before CNC existed, or CAD programs for that matter. I had lots of threaded holes to force the mold sections apart and I had to cut flash off in places. The shop had a simple injection mold machine that could shoot Nylon. I could get the mold to fill more reliably by heating it in an oven before placing in the mold machine. No, the mold was not as nice as the one in the video, but the final part looked better than what I see 3D printed today. I only say this to encourage people. Even if you don’t have a mill, a mold can be machined with a drill press and files. The accuracy is whatever you find your hands capable of doing.

    3. Looking at the site, I would recomend looking at their shredder design, very DIY-friendly… the screw extruder looks reasonable as well, provided they’ll be able to supply the screw at the advertised €250 a piece…

      As for the molds, there is a cheaper, more DIY-able version

      1. Maybe somebody will start selling somewhat lower cost base molds that can support custom UV printed molds for people who want to do a few dozen test runs? It’s a somewhat more time consuming in the beginning hybrid possibility that has some merit for some applications but you will still need a full sized injection mold plus mold setup time at the very least.

  3. As far as I am concerned the second anything is modelled in solidworks it is not homegamer stuff. Pisses me off when people give tutorials using software that costs 4000usd per seat about things that if you have a licence for you can do in your sleep. What’s the point.

    1. Students have access to the finest tools available in industry, Autodesks fusion 360 is free for home users and small startups and could quite possible become way better then Solidworks in just a few years.
      Open Source Tools are getting better and better.
      So, just wait a little longer :)

  4. If anyone is interested in getting injection molding in China…

    A few years ago we took out first steps to have a 2-part wall mount type enclosure (3.5″ x 5″ x 1.25″) injection molded in China and the results were amazing. The mold cost (using NAK80 Steel) was about $4000 USD… and the parts cost was about $0.50 each for 2000 pieces… This translated to an amortized cost of about $3 USD per unit… over the initial 2000 units… which was cheap given that it was a custom design and looked perfect! The total time from design submital to actually having parts was 6 weeks (including Ocean shipping).

    1. A few questions for you then:
      – what happened to the mold? (do they keep it or ship it to you or destroy it?)
      – how did the initial run work, they sent prototypes or someone had to travel to China?
      – what material did you use?
      – who designed/”debugged” the mold? (certain features are hard/impossible to or significantly increase the cost, so the design of the part tends to be optimised)

      1. I’ve gone through this process, so I can answer:

        -It depends on your agreement with the company. If you specify that the mold is for export, it will cost more. If you specify that it will stay there, then they’ll give you a cheaper price because they’re making up the rest by making your parts for you. Usually they’ll store your mold for you when not in use up to a certain amount of time, and then they’ll destroy it if you don’t order any more parts. If you had it made for export, they’ll ship it to you, but if you didn’t, then they have no reason to give it back.
        -The mold quote usually involves two sets of samples, which they will send to you by air; first the unfinished version, to make sure all the dimensions are correct. Once that is accepted, then they’ll put the surface finish on and send you a second sample.
        -You’ll have to specify the material from the beginning because they account for plastic shrinkage when they cut the mold (so the mold will actually be a couple % larger than the final part dimensions). Common materials are ABS, PP, PS, acrylic, there’s lots. You’ll also have to specify the color and any additives required (like UV stabilizers or flame retardant).
        -After the quote and before they cut the mold, their engineers will look at the part and start designing the mold. They’ll send you a document with lots of pictures and arrows that say “need radius here” “2 degree draft here”, etc. You’ll make the changes to the mold and negotiate with them. They’ll also tell you where the parting line is, where the gate is, the ejector pins, etc. You can negotiate this as well, but their recommendations are usually best. There have been times where I had to have a snap feature onto another part but I couldn’t prototype properly to figure out the dimensions for best snap, so I sent them the thing it had to snap on to. They cut the mold appropriately, then removed metal until it snapped properly. That avoided a bunch of shipping for experimentation.

  5. Amazing work .. about as good as Social Impact technology gets. They have simple, well thought out machines and methods, and a corresponding business model that can work in countries with fairly modest infrastructure.

    1. Though the business model is what I think is the last bit to get this project a break trough. Ther is little to no money in it as the normal people will continue to buy bowels made for cents. It’s like ‚third world shops‘ where only wealthy and idealistic people are willing to buy their stuff.
      What I think is needed is something that would even make the big guys interested in recycled plastic.
      But if they continue to optimize their processes and machines and get the word out, they could enable a huge number of people to be the first in that rising industry of recycled from plastic waste products.

  6. hmmmm… everybody seems to be forgetting one thing. I’m sure that for most people around here, at least for me, there is the following problem:

    THINGS ARE NEVER RIGHT THE FIRST TIME!! Therefore you must be willing to go through multiple iterations of your project, be prepared to accept this OR outsource the entire project to professionals and let them reduce the risks, in the end the results will be more professional. Which is the look you are looking for when going through all the trouble of injection molding. Though I could be wrong (I have that a lot).

  7. Basic injection molding is not that tough to do. Melt plastic and force it into a mold under pressure. Of course, your results are going to be limited without access to high end machines, tooling, and software, but the idea that you can’t do it without a five to six figure investment is false. Large, complex molds will probably always be out of reach for the home shop. I’ve seen guys make small parts with hand made molds and manual presses though. There are frequently some imperfections, but that doesn’t mean no one should try it. Just like everything else, you learn from your mistakes and try again. This is Hackaday after all.

  8. Sigh. This is so fundamentally stupid that I don’t even know where to begin.

    From a safety point, let’s check where we are.
    1) Molten plastic degasses and lungs are prone to not liking those gasses.
    2) Cold mold, hot plastic, and manual force; what could possibly go wrong?
    3) Clamp tonnage. If you’ve never seen a mold blow out consider yourself lucky.

    From a design perspective, let’s note only the big stuff.
    1) Where’s the parting line? You need to vent mold gasses, but there’s no mention of this at all. A highly pressurized chamber is a recipe for disaster.
    2) Awful materials. Maybe this is the point, but the products in question appear to have more splay than reasonable, and they seem to be rather brittle. Reclaimed material can be quite strong, but only when controlled.
    3) That gate. You’ve got a cosmetic piece, and put the gate in the center of the largest uninterrupted surface. The better design would be two fan gates on the sides or bottom/top, with design work being done to make sure flow fronts properly knit (assuming that tonnage couldn’t be applied to use a single gate). Less gate trim, better aesthetics, and an order of magnitude more difficult than a single flow calculation.
    4) Lack of flexibility. Shot volume is determined by the preloaded material. Cost is prohibitively high on a gigantic finely machined block. Time to creation is just nuts, namely mold design relies upon relatively wide finish tolerances and conditions.

    Why should anyone give a crap about what I say? This is an exercise in technology, right? All of DYI is bespoke, and thus better than buying enmasse.

    Let’s do the math. The initial block has to be degassed (H2), so that there aren’t any voids during machining. The blocks have to be machined to very tight tolerances. Once both of these things are done, you’ve still got to contend with a limited number of shots based upon mold wear. Let’s be very friendly, and suggest a few hundred dollars for a block of good aluminum, a thousand for machining time and finishing, with an overall cost of $1500 once the press apparatus is constructed. If you sell each case at $3, and a profit of $2, that means without labor you break even at 750 cases. Factor in finishing labor, and you might break even at 1000 cases sold. I’m being generous on the estimation of a lot here.

    Now, let’s do some other math. A silicon mold costs a few dollars. It’s a few hours of effort to create a positive, that the silicon negative can be used with. I continue to make the silicon negative from the positive, creating a small initial investment of labor. The same exact product can sell for $3, with a $1 profit. Rounding up, maybe 50-100 parts made this way would break even on production monetary investments.

    As others have said, injection molding is stupid. It’s a technology meant to have a large initial investment, with payback from consistent quality and large production volumes. This entire article glorifies using a tractor trailer to drive an impaired person throughout an airport. The right tool is an electric cart, but because the tractor trailer can still do the job it’s an…interesting proof of concept.

    Seriously, please don’t do this. The entire exercise is stupid. It demonstrates an interesting capability, but not a thoughtful one. Injection molding is awesome, for high volume parts. Using it like this is fundamentally not understanding the technology, Please, please, stop doing this. Seriously, “because I can” is not a justification. If you are doing enough volume to justify injection molding you shouldn’t be running reground random bits (seriously, moisture readings matter). Either choose a process that fits the scope of production, or admit this is insane.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s