A Polymer Concrete DIY CNC With No Perceptible Budget In Sight

The Jargon File describes a wizard as someone who groks something to a very high degree, or the kind of person that builds a polymer concrete CNC machine with a pneumatic tool changing spindle that they designed by themselves.  It makes you think that maybe Tony Stark COULD build it in a cave with scraps.

It’s a five part video series showing snippets of the build process. The last video gives an overview of the design of the machine. It is all very much in German, so if you speak German and we got anything wrong about the machine or missed anything cool, please fill us in down in the comments.

The machine starts with a 1500 kg polymer concrete pour with some steel stock embedded in it. It is then machined within an inch mm of its life as shown by practically zero deviation over its length when measured against a granite block. The wizard then goes on to make his own spindle, get castings made, and more. We liked his flowery kitchen hotplate, which he used to heat the bearings for an interference fit. It added a certain amount of style.

Unfortunately the videos don’t show the machine running, but we assume this sort of person is happily building arc reactors, power suits, and fighting crime. They probably don’t have time to film “CNC Bearbeitungszentrum im Eigenbau Teil 5”. Videos after the break.

57 thoughts on “A Polymer Concrete DIY CNC With No Perceptible Budget In Sight

  1. At 0.17 in the first video, we can read “Ultrahochfester Beton”, which stands for “High performance concrete”.
    In my understanding, this is regular concrete, not polymer.

    1. Later videos show the use of Mineralguss for the X axis, which looks a lot like epoxy granite to me. Mineralguss means “mineral casting” in English according to Google translate. I suspect that translation doesn’t actually capture what Mineralguss really means.

      Ultrahochfester Beton actually means Ultra High Performance Concrete. UHPC is not the same as a bag of concrete you’d get from Home Depot. Find more here:


      1. What inherent advantages does Ultra High Performance Concrete have over other solutions, aside from being easier to bootstrap? Is it fundamentally “better” than cast iron? Is the few percentage of shrinkage during curing an issue?

        1. It has several major advantages over cast iron. Firstly, it requires limited to no post processing such as stress relieving and hardening.

          Secondly, it is cast in molds that if properly designed and constructed, are both exceptionally accurate, and reusable. Even a material as simple as melamine with a release agent is sufficient to make a reusable polymer concrete mold.

          Thirdly, in properly designed aggregates, it has superior damping properties to cast iron, while on a weight for weight comparison is also of similar strength (going off memory here).

          Fourthly, it is to an extent self leveling. Shrinking is consistent so while it won’t dimensionally match the mold, it will geometrically match the mold. So if your mold is accurate and flat, you can basically bolt hardware like linear rails directly to the casting using threaded inserts cast in to the machine.

          In short – yes, it’s basically the ultimate bootstrapable material for building machine tool bases.

          1. It’s not self-levelling if you move it into your basement. He machines the inlays and tweaks the two blocks after they’re cured. He uses cast iron for the z-carriage.

            However, he DOES say that the “Mineralguss” he uses (in the y- and x-tables, if I recall) has better damping than cast iron.

          2. Waterjet: I meant the shrinkage is consistent in terms of the volumetric magnitude of the shrinkage, not that it’s perfectly linear in all dimensions. As in, if you’re using the same mold over and over, it’ll have the same release properties on curing, and you can design in the shrinkage to the part and accommodate in the mold. But in a way the shrinkage is dimensionally consistent – the slurry will want to stay in bond/contact/vacuum with the mold much more than the atmosphere, and the atmospheric pressure adds to that. So what you’ll see is that the casting will suck in on the exposed face, rather than contract away from the mold faces.

            Many people have found that when cast on a nice flat/consistent mold like well backed melamine, the surfaces have been flat enough to bolt linear rail straight onto the inserts. Those that have felt the desire to, typically rework the casting faces to be flat by throwing the casting on a mill and machining it as you would any material with carbide insert face mills. The carbide tends not to mind the aggregate so much.

            Alternately others use ultra-thin self leveling expoxy cast as a separate layer on top of the as cast base aggregate, like tabletop epoxy, or Moglice.

  2. The American Precision Museum in Windsor, Vermont has a lathe they say was made by one man using only a file. He bought only the raw casting and the lead screw, the rest he bootstrapped from the unit as he built it. This reminds me of that.

      1. I was there about 20 years ago. Wonderful place – my wife wasn’t bored, it’s that interesting. I’d love to go back again, and I highly recommend it to anyone in New England or visiting Vermont. Glad to see it’s still going.

      2. This was not a small unit. As I recall it had an eight foot bed and was used to machine parts for textile mill looms. I don’t see it on the museum’s web page. Mind you my last visit there was almost twenty years ago and it may have been displaced by something else since, as happens in most museums over time.

      3. Wow, that looks like an amazing exhibit! I’ve been meaning to check that place out ever since I heard about it a couple years ago, but it seems they still haven’t reopened following hurricane Irene. :(

        It’s funny, I live just two hours away, on the other side of Vermont, and the first I heard of the place was when my mother found a souvenir brass gear from the museum in a big box of buttons she bought at a thrift shop.

    1. Clearly you’ve never heard of David J. Gingery. He’s MUCH MUCH after that time. But, his methodology is done OLD SCHOOL How he builds a machine shop from scratch is how you imagine someone like Eli Whitney built his first prototype. He has an entire book series that covers everything you need to know. Now, he does assume access to things like screws and scrap metal. So, that may be different. However…using plate glass and marker die and a flat file is how it was done first folks.


        1. Older people are more inclined to actually put the elbow grease into making things like this because we have had the experience of satisfaction from making something that is well done, not just good enough.

          I still have a lot of hand tools in the mechanics workshop, various saws, files a drill even though I have their electric equivalents. It’s much easier to get precision at the lower speeds of hand tools.

          It’s the same in the electronics workshop. I have several tools that I have made.

          1. Well at 63, I’m hardly a youngster, but the point I was making about the Gingery books pertained to my dad’s generation. Far more of those books were sold than shops built. My whole family were skilled tradesmen, and all were DIY because that was just the way it was for them, but there were lots more people that had dreams of big projects that would talk endlessly about it with my dad and my uncles that you just knew would never even start.

          2. Ageism aside, I think maybe free time has something to do with old people being compelled to make things that are truly impractical. Resources are also part of it. Not to mention property – you think people living in cities can follow gingery books?

            Now a days, citizens are told we are worthless eaters, are denied pension, and are told to work until death – and to be glad if we can even find work – never-mind 50% of the country gets welfare and has no incentive to find work.

            That said, nothing but respect for anyone who does these types of projects.

    1. Nearly perfect. Far far better than Portland based concrete, and at least equivalent to cast iron.

      Put it this way, people are building the superstructure for 6 and 7 figure machining centers out of polymer concrete, for tools with anticipated service lives in decades.

      1. My first apprenticeship was the similar. They hand you a block of metal and 4 files and tell you to make a blue printing block and it takes about 3 months of filing.

        At the end of that 3 months one quarter had left so it’s a really good way of sorting the stayers from the quitters early.

        1. A former co-worker who apprenticed as a machinist at Eastman Kodak in the 60s tells a similar story of starting with a file and making tools. As an apprentice, he used simple tools to build a vice that served him will through his entire career and still looks beautiful.

    1. Yep, that is absolutely true. I attended the Technische Fachoberschule in Germany before getting my BSEE in the US. We laid bricks in sand for 3 months. Then we filed metal by hand for three months. Then we made wiring harnesses with strings and knots for 3 months. Everything had to be perfect. 3 days of Praktikum (manual) and 2 days of Lektuere (theoretical) per week. Not sure they do it this way anymore.

  3. Don’t you just love the German language!

    I loved the part where he has used filler /bonding agent to ‘glue’ the z axis mount on and it says “Halten”. In English I would have said “It’s not f(n) going anywhere”.

    Then when he is poring the molten metal cast we waves the camera man back to stand well clear … I was expecting to read “Krematen” as it flashed up in screen.

    1. He wrote, “that’s gotta hold”. Seemed like a weak spot, though. Another one were his y-axis covers. He believes the coolant is just going to obediently flow downhill. Water never behaves! And then just relying on the absolute encoders on his motors. Who cares where the motors are? I want to know where the table is!

      That being said, holy crap this guy is amazing. He’s a tour-de-force of either know-how or balls. I hope he’s making progress.

      1. That’s a pretty standard design for way covers, they will work fine. Though usually they do have scrapers on the leading edges as one of the techs where I work found out about the hard way.

        Most commercial machines rely solely on the motor encoders. Linear scales are a expensive option. They do map the screws though, I am guessing this guy probably has access to an interferometer to do that. With a mapped screw you will be holding sub-thou at a good distance and at that point you need a temp controlled environment.

  4. I’, certainly not the brightest bulb in the string, but not the dimmest either. A bit of time looking to the topic has thinking that UHPC is a polymer. Or at lest contains polymer in it’s mix. The reference to concrete.org says superplasticizers are used in the mix. My conclusion is that there’s nothing wrong with the title of the article. Any concerns about the stability of the product may not be realized in practice While I’m not convinced standard concrete is a good replacement for cast iron in machine tools, but I can see where this concrete mix could be.

  5. I want to design a cnc to 4 axis rout aluminium from about 1 mm thick to what ever I can get away with under 5 – 6 mm.

    In typical hacker fashion I have bought the linear rails without doing any math cos I don’t know the math for this!

    I have 16mm round supported rails.

    In his build he has massive x and y ball screws and a smaller one for z.

    What size ball screw should I use for what I want to do?

    1. Round rails suck for CNC machines. There can be a lot of slop on them and they generally are not made to the tolerances of square. Square THK style rails are the way to go.

  6. Well that is not a wizard. Its a German.
    Give any German engineer or machinist with an interest in machining enough spare time and money and he will most likely end up with something similar.

      1. They said pass the test…so they made it to pass the test :P
        (it’s not physically possible to pass it without a very expensive SCR system, that was still in development for VW at the time)

        1. Thanks for the succinct summing-up but it still has to be said – they didn’t pass the test – they cheated. As did some other companies. Do you like doing business with companies that cheat?

    1. The beauty is that you really do not need one that accurate. — you can use a course cutter and then scrape the precision surfaces. Having access to a large gantry to do the rough machining is a time saver.

  7. I like the machine…. UHPC concrete may save a lot of cash too compared to Epoxy. You need somewhere between 5 and 10% epoxy by volume to make a “mineral casting” or “Epoxy concrete” (the same material) to make component. That can add up on a large casting.
    I like the way he set the machine up using simple jacks to align the column then epoxy grout. Note the way he built in distribution galleries; risers if you are used to sand casting for the grout to properly fill the gap.
    Would it really be that expensive to hand the base, column, and table support casting to a machine shop to mill the steel inserts as he has done?
    Not sure about his tool changer? I used a different approach. The rotary air Picking arm would need to be solid. there is a lot of overhang

    I worked on an epoxy frame mill a while back if anyone is interested.
    The 3D plans are Autocad files Still not built! I have no room. It is on my to do list.
    The following Google search phrase should find it
    epoxy cnc mill McNamara


Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.