Casting A Lathe Out Of Concrete

Look up ‘concrete lathe’ and you’ll quickly find yourself reading the works of [David Gingery]. His series of books on building a machine shop from scrap begin with a charcoal foundry, and quickly move to creating a metal lathe out of concrete. Before [Gingery]’s lathe, around the time of World War I, many factories created gigantic machine tools out of concrete. It’s an old idea, but you’ll be hard pressed to find anyone with a shop featuring concrete machine tools. Cheap lathes are plentiful on Craigslist, after all.

Building a metal lathe from concrete is more of a challenge. This challenge was recently taken up by [Curt Filipowski] in a five part YouTube series that resulted in a real, working lathe made out of concrete, scrap, and a lot of bolts.

The concrete lathe begins with a form, and for this [Curt] cut out all the parts on a CNC router. Creating the form isn’t quite as simple as you would think – the concrete form included several bolts that would alow [Curt] to bolt bearings, ways made out of gas pipe, and angle iron. This form was filled with concrete in [Curt]’s kitchen, and after a nice long cure, the lathe was moved up to the upstairs shop. That’s a five hundred pound block moved up a flight of stairs by a single person.

The rest of the build deals with the cast concrete carriage which rides along the polished gas pipe ways, a tool post holder milled out of a block of aluminum, and finally making some chips. While it’s not the most practical lathe – the carriage moves along the ways by turning a wheel underneath the tailstock – it does demonstrate a concrete lathe is possible.

75 thoughts on “Casting A Lathe Out Of Concrete

    1. What I found out (to my utter surprise) when moving my newly acquired 175kg (385 pounds) welder, if it has caster wheels and they have the right spacing, you can relatively easily move it on stairs (or in and out of a van), as you are always lifting only 1/2 the weight, which is perfectly doable even for a atrophied IT tech and his equally “strong” friend :D

      The wheel spacing is critical, both wheels have to be able to travel the whole length of a stair, otherwise you have to lift the entire thing, which sucks even on an inclined plane ;-)

  1. The general consensus about this design has been that the rigidity of the concrete would be negated by the flexibility of the ways. Also, concrete is not as dimensionally stable as many people assume. The important question is, what kind of tolerances can this thing hold?

    1. As close as one would need if this is what you had to work with. It’s the repeatability that would be the real bugger if you were using it for production, one off, you could compensate.

    2. Concrete has a tensile strength that is in the region of cast iron. It takes about 4 weeks after pouring to reach that strength. If you use rebar and higher quality concrete, you get high strength (and relatively low weight) and good dampening properties. The problem with the project here are the rails (black pipe) the should be hardened to minimize wear.

  2. As previously mentioned, it’s well known that concrete can be a suitable base for machine tools.

    …for a time. Most of the purpose of such tools is precision, which, after the concrete shifts (it’s not stable over time) they won’t have much of. Regrettable that epoxy granite is so much more expensive.

    Also, I couldn’t help but notice…no compound?

      1. Not nearly as much as you would think. Even very wet concrete with additives to improve the flow has trouble in a complicated mold.

        I improvised by bolting a bench grinder (with an intentionally out-of-balance wheel) to the mold, but even that left a little to be desired.

  3. The spindle bearings are not pre-loaded and you can see the chuck lift when a cut is made on the brass. You need to use spindle bearings or angular contact bearings.

    Also square the rod and angle and weld it together. That or buy linear rod bearings on ebay.

    1. I saw that and was thinking the exact same thing – that spindle needs to be preloaded and would benefit greatly from a nice set of angular contact bearings.

      Brass is also tricky to machine. It’ll be cutting smoothly and then suddenly grab (which is what happens right as the chuck lifts in that last video).

      That said, this machine seems pretty capable.

      I’m curious how much it cost though. You can often find the cheap, but totally capable, Harbor Freight (aka Sieg, Enco, Grizzly, Real Bull – they have many brand names) mini lathes for $300-400 used, and not much more new. With some tuning you can do quite a bit with those machines.

      1. His point was that a “gingery lathe” is not a concrete lathe. David Gingery wrote books on how to bootstrap lathes, but they’re all about green sand casting, not concrete as the material.

    2. And he is using what appears to be self aligning flange mount bearings like you would use on an exhaust fan or something. So there is very little rigidity in the spindle which can be seen with all the chatter when he is trying to face something.

    3. Yes, clearly tapered roller bearings would be better, but I also suspect a problem with the deflection of the chuck is that it is mounted to a plate of 1/4″ aluminum. The 3/8″ plate ONLY serves to center the chuck; that quarter-inch plate that it actually mounts to probably accounts for much of the deflection.

      Welding has its own problems. Gingery states in his lathe book that he first tried to fabricate a lathe by welding, and found that he couldn’t control the warpage as it cooled, leading him to try casting aluminum and zinc, with superior results.

  4. What about adding epoxy so it’s smooth-edge fills the gaps , needs to be set under pressure tho , so not sure how your going to do that. but it would provide sturdiness maybe reenforce it with steel bars.

  5. Square tubing is astonishingly straight. Much straighter and stiffer than gas pipe. My Dad had commented about this before he died, but it was only after that I put a precision straight edge to a piece of 3/8″ wall, 2″ square tubing he left me. It was as straight as the straight edge which was specified at 0.001″ in 3 ft. You would be most unlikely to match that by welding. In fact, the chief difficulty one would face making ways using square tubing would be alignment.

    The concrete work shown is an example of what not to do.

    1. How so? To my limited experience in concrete, it looks like he did right by not having too wet a mix, making it stronger, but did not work it enough to get the cement and fine aggregate into all the nooks and crannies. He does seem to be a bit in denial, deciding that the deep voids are “cosmetic”. But his choices at that point were 1) break up and dispose of 400 lbs. of concrete and try again, or 2) use it as-is and hope for the best.

      Noob mistake, but I’m wondering what you mean by “what not to do”.

      1. That is absolutely terrible concrete work. Everything he did that has anything to do with concrete is wrong.

        First off, he should have just bought a sheet of concrete forming plywood; it already has a waterproof finished surface specifically designed for contact with concrete. It would cost less than that sheet of MDF and he wouldn’t have had to seal or oil anything.

        Second, why use quickcrete bags when you can design your own, much stronger mix for a fraction of the cost by just buying a bag of portland cement and some aggregate.

        Third; while a dryer mix may end up being stronger, it can be marginal and you run the risk of what happend here. Any strength benefit is outweighed by the fact that all the air voids have made the structure incredibly weak and prone to cracking. A properly compacted mix is much more important than a slightly higher final Mpa. Strength is useless if the concrete is not workable. Concrete should be able to flow.

        Regardless, for this case, the strength of the concrete is irrelevant. All you want is a big heavy block that won’t fall apart. He could have used concrete so weak that it wouldn’t even be suitable for filling in a whole and it would have still worked.


        Yeah, sure he lives in Canada – and so do I. I know a little bit about canadian winters. Considering that I operate tower cranes on high rise buildings, I also know a little bit about pouring concrete in canadian winters.

        Concrete work doesn’t stop in the winter and you can rest assured that there is no heating for the concrete that is used for walls. Floors are heated from underneath but only for two days. After that the heating is turned off and moved up for the next pour. Every single building in Canada that takes longer than one year to build has concrete in it that did not see temperatures above freezing during it’s entire cure time.

        If you want to pour concrete in the winter in you garage – GO AHEAD. Once you pour it, cover the form with some plastic to keep the humidity in and some thick insulated blankets to keep the heat from curing in. you can strip the forms the next day if you want to – but don’t leave it anymore than three days or you’re just making it harder on yourself to strip the forms.

  6. uh…. I’ve read what I thought were all of the Gingery books. None of them cover making a lathe out of concrete. the lathe in his series is made out of aluminum, which is melted and formed using the techniques learned in the first book, the charcoal foundry. Sorry for the flame, and my apologies if you can point me towards such a book.

    1. Was a book ever mentioned? I can’t tell if you just want a book for use as a reference or if you’re implying that without a book this can’t be an effective way to make a machine.

      I’ve built the entire Gingery series of metal working tools and can say with some authority that making a Gingery lathe is magnitudes more difficult than what I saw in those 5 videos. If all you want is a lathe, this is a pretty good way of going about it. The Gingery series is more about building a comprehensive metal working shop from scratch, the lathe being only one part of that.

      1. His point was that a “gingery lathe” is not a concrete lathe. David Gingery wrote books on how to bootstrap lathes, but they’re all about green sand casting, not concrete as the material.

        1. Right. But the article never says you’ll find a Gingery book on concrete lathes. It says if you search for ‘concrete lathe’ you’ll find results for Gingery lathes. It then goes on to talk about pre-Gingery concrete lathes.

          I still don’t understand if the person I replied to was looking for a book, felt like there was a book the builder of this concrete lathe referenced, or if they were trying to be snarky and suggest if it’s not found in a Gingery book, it ain’t right.

          To your comment – did you even read what I wrote? We all agree that the Gingery series is all about casting metal parts for metal working.

          1. From the article:

            “[David Gingery]. His series of books on building a machine shop from scrap begin with a charcoal foundry, and quickly move to creating a metal lathe out of concrete”

            Did *you* even read the article? None of Gingery’s books deal with concrete to my knowledge.

          2. From the article:

            His series of books on building a machine shop from scrap begin with a charcoal foundry, and quickly move to creating a metal lathe out of concrete

            Did *you* read anything? Article, comments, anything?

    2. There is a book, but not by Gingery. This build is based on the Yeomans lathe, which was designed to ramp up US production for WWI by introducing an inexpensive lathe that could be built easily and quickly. I can’t find a link to it right now, but I’ve done so before and found the original book in pdf form.

      1. Except the Yeomans lathe is not made from conrete. It’s a lathe where steel guideways are cast in place instead of the usual way of making a rough casting, followed by milling, and then by painstakingly scraping the ways.
        The original article is this:
        Note that this article has no reference to conrete, cast here clearly means cast iron.

        Then some guys came along that for some reason (maybe because of the rough cast look) thought that it was made out of conrete. This is the article that is the original source of the WW1 conrete lathe story:

        As for as I know there is no original documentation about any WW1 conrete lathe, and all articles about it derive from this one “multimachine” article. Considering the quality of concrete at the time I very much doubt it existence.

        This concrete lathe here would be a lot better as a wood lathe. Considering the spindle design (or lack of design) and the rather unsuitable characteristics of concrete. It’s clear from the brass turning video that it can barely turn non-ferro metal, let alone steel. If you just need to do occasional, not too accurate metal turning,as well as freehand wood turning it might be a good lathe.

        It’s a nice effort but seems like a lot of work for what it is though.

  7. Why does HaD keep featuring junk projects like this? Do you not have a mechanical engineer or machinist on staff? Proving something can be done, whether it should be done, and whether it is a practical learning experience/there are lessons to be learned, are all entirely different things, and this project ticks none of those boxes.

    There’s a wide variety of reasons why this project is moronic that would be apparent to any sophomore mechanical engineer. Chief among them is that concrete is a terrible material to be building machine tools from. It’s not even a practical material for that task.

    1. Ticks none of those boxes?

      They proved something can be done. I saw the videos. I have several “real” lathes, from tiny little ones to a giant beast of a lathe. This guy made a machine that can do most of the things my “real” machines can do. Box ticked.

      Should it be done? You are not the authority on what should or shouldn’t be done. Nor am I. If the guy who made the lathe thought it should be done, then yes, it should be done. Box ticked.

      You assert that concrete is a terrible material. Let’s go with that. You don’t say why it’s terrible so let’s also assume you’re talking about its long term stability and shifting dimensions as it cures. Let’s ignore the affordances for adjustments over time the designer of this project clearly points out as he builds the lathe.

      6 months from now the maker of this machine might discover that the concrete has changed enough to cause alignment issues. They may not know that right away, but will certainly spend some time investigating. Will they learn something in a practical way? Yes. Box ticked. Will they gain experience? Yes. Box ticked.

      If you’re going to trash a project, and cite that even a sophomore mechanical engineer knows it’s moronic, don’t be as sophomoric to not point out why. You’re not helping anyone by just being a dick.

      1. Concrete:
        -low Youngs Modulus, so large strain (deflection) under load
        -hygroscopic, so it absorbs moisture from the atmosphere, which is one of the primary causes of dimensional instability
        -largely amorphous microstructure, so brittle
        -very poor damping characteristics, especially compared to grey cast iron, epoxy granite, or granite

        And that’s ignoring all the mechanical design fails (some key points would be, spindle design is atrocious, ways are poorly supported, line Hertzian contact area between carriage and ways, ways poorly supported, cross slide design flimsy…)

        I doubt this lathe can do the single important thing your ‘real lathes’ are designed to do – and that’s precision. This is just another example of cargo cult design.

        1. I wasn’t aware this had become a site where people came to see well-engineered machinery. Last I heard, this was Hack-a-Day.

          Am I in the wrong place? ’cause I’m not here to see what “Real” engineers can do.

          1. Really? Because here I was thinking I was on a site that celebrated the use and reuse of technology to achieve practical ends. From the sites own ‘About’ page: ‘This highly creative activity can be highly technical, simply clever, or both’.

            This is neither highly technical, nor clever. There’s a gulf of difference between ‘professionally engineered’ and ‘having put even the barest minimum of effort into understanding the problem you’re attempting to solve’. This is a HaD fail, not a hack.

          2. Fail? Fail is defined by the customer. If you build something for yourself, and it meets or exceeds your own expectations, then it is most definitely not a fail. The fact that it doesn’t meet YOUR standards means you shouldn’t try to buy one from him, or build one in the same manner. It doesn’t appear that you had any notion of doing either, so just what WAS the purpose of your post?

            The fact that the builder of this project already had access to CNC machine tools tells me that the primary objective was not to obtain a lathe as cheaply as possible. Also, he did not blindly follow the instructions for building Lucien Yeomans’ lathe – he modified the design to fit his needs and expectations. And while the end result is a machine that probably doesn’t perform as well as a commercial lathe he could’ve bought for the same money, I’m certain he has learned a lot from this process. I am also sure that his lathe is not finished.

            Hacking is the act of deciding that what you can buy is not what you want, and doing something about it.


        2. Dude.. It is surely a first go in what will become a body of work. (After all it looks likely to vibrate to pieces, but that does not matter)

          The amount of work and some obviously decent craftsmanship that went into this is impressive.

          1. Replace ‘concrete’ with ‘polymer concrete’ or ‘epoxy granite’ and I’d fully endorse it. Steel reinforced polymer concrete is a great material for machine tool bases, and combined with inexpensive linear rails, can be the basis for the best pound for pound, dollar for dollar performance.

      1. It’s quickly turning that way. It seems almost every day now that the ‘hack’ of some ‘hack’ is being featured, that is more regressive than progressive. This is a prime example.

        Gerrit Coetzee keeps writing articles basically calling out exactly the type of ‘cargo cult’ design that is featured in this ‘hack’, where ‘fuck the engineering, if it feels right, do it!’ seems to be the overriding design philosophy. Obviously the rest of the staff writers aren’t reading them.

        1. IMO Hacks involve clever creation or repurposing of something to solve a need or problem. Kluges usually involve duct tape, zip ties and inapropriate uses of things as inadequate fixes that are accidents waiting to happen.

          Using a stack of pennies wrapped in tin foil for repairing a blown fuse is a KLUGE
          Building your own custom niche spot welder is more in line of a hack – unless you use duct tape and zip ties

          This lathe build is an accident waiting to happen.

    2. Assuming you are not just trolling…

      Why do you think you are to decide if a project is junk or not?
      Why would a mechanical engineer be needed to _post_ about a project? Are you assuming that there should be an in-depth analysis of each posted project?
      [Also you should browse machinist websites – the “cargo cult” is strong there! The most obvious one is the worship of weight rather than proper design]
      This is obviously a practical learning experience, claiming that it isn’t smells of bovine droppings.

      Moronic? Sophomore? No, how about you just go away? You obviously don’t know shit but try to brag about yourself _without_ giving any clue why you “know” something isn’t good enough.

      Concrete have been and is(!) used to make machine tools so even that is wrong.

      In retrospect (after writing this) I hope you were trolling…

      1. I’m not trolling at all. I’ve sat and had a critical think about the information presented, and using my experience as a mechanical engineer and machinist, I made a value judgement of the project – and the outcome was not positive.

        No, but I think that someone should sit and think critically about what a project says before they mindlessly post it. This project provides good learning examples in many regards – structural design for stiffness, spindle design, understanding the properties of materials – but none of the lessons to be taken from it are positive. It’s basically a ‘how to’ manual of how not to design a machine tool. That’s why I’d call it a HaD Fail, rather than a post to highlight.

        I regularly browse machinist sites, and there’s a very good reason why mass is worshiped that’s immediately apparent to anyone who understands machine design. All (and that’s a large generalization, but holds for anything under a $100k VMC) metalworking machine tools are made from grey cast iron, for its damping capacity. Therefore if the material and its basic properties are fixed (focus on the density property, in this case), without performing an in-depth structural deflection analysis, the best estimate of the stiffness of a machine is its mass. Machine tools by and large have comparable form factors; grey cast iron always has the same density. The only way then, for machine tools of comparable size, to increase the mass, is to increase the section thickness of the castings or add other internal bracing, and as a consequence, increase the machine stiffness. If a machine tool is heavy, it’s not because it’s full of lead.

        ‘Cargo cult’ design is not the raw worship of a single property; it’s the aggregation of properties that resemble the desired result. It’s grabbing self-aligning bearing blocks and using them as spindle bearings, ‘because lathe spindles have bearings’, without understanding the multitude of reasons why those particular bearings are nearly completely unsuitable for this use. It’s using allthread as a leadscrew, because lathes need leadscrews, right? And on it goes.

        Concrete isn’t used to make machine tools. Composites are used to make machine tools – specifically epoxy granite/polymer concrete – when weldments or cast irons are not. But you will not find a machine tool made this century, where the main structural components are based on a portland type cement concrete. For the reasons I’ve stated already. Concrete has only ever been used as an expedient in machine tool design, because the engineers that design machine tools understand its limitations. Portland concrete is far cheaper than grey cast iron or polymer concrete. If there were any genuine reason to be building machine tools from it, people would be doing it. It’s a lot more than ‘tradition’.

        I’d simply say that the reaction to my comments shows just how deep the ignorance of mechanical engineering and machine design runs in this community. Clearly I’m rubbing against the grain, but the question then remains, why is this community writ large defending what is fundamentally poor engineering and design – and as a corollary, bad hacking? Why is my observations, as a professional in the field, being dismissed as trolling? I certainly wouldn’t have gone to this level of effort to respond just for a lark.

  8. I was the project engineer for the procurement of a custom built 52-ton, 2 story tall, ultra precision diamond turning lathe. The shell of the machine was made from welded plate steel. The components were turned upside down and fill with an epoxy-granite slurry and cured at an elevated temperature. The steel shell acted as a form for the slurry. When completed, the machine looked like steel but had better damping. The machine was stable and performed very well. Large castings would have been more expensive and taken long to procure.

  9. This is neat, though it looks of questionable short or long term stability in it’s current form.

    Somewhere on there is a long series of posts about building CNC mills, mostly vertical style, with epoxy and common concretes. A very thorough interesting series of threads from sometime around 2004.

  10. Large lathes were built out of concrete during WWII if I remember right. They were for turning ship’s gun barrels. The idea is sound but the engineering in this design is sorely lacking. The spindle itself might be fine for a wood lathe but never for metal.

    1. Tapered bearings for a car axle would do better than this. Get some preload on the bearings in this lathe so it can handle boring the ends of a heavy wall steel pipe to hold the outer races for the wheel bearings to make a sort of cartridge spindle for the next concrete lathe you build.

      That’s a neat property of lathes, one that’s good enough can be used to turn parts to a degree of precision for building a better one, as long as the spindle isn’t flopping around.

  11. Wow, another nice discussion going on over here (I’m Curt, I built this machine). One thing I would love to correct, this is not based on David Gingery’s work, its based on Pat Delany’s work around the Yeomans concrete lathe from the WW1 era. Both guys have great designs but lets give credit to the correct person :).

    Was a nice surprise to stumble across my own project on hackaday. Thanks guys!

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.