Lost Foam Casting In Action

Even though not all of us will do it, many of us are interested in the art of casting metal. It remains a process that’s not out of reach, though, especially for metals such as aluminium whose melting points are reachable with a gas flame. The video below the break takes us through the aluminium casting process by showing us the lost-foam casting of a cylinder head for a BSA Bantam motorcycle.

The foam pattern is CNC milled to shape, and the leftover foam swarf is removed with a hot wire. The pattern is coated with a refractory coating of gypsum slurry, and the whole is set up in a tub packed with sand. We get the impression that the escaping gasses make this a tricky pour without an extra sprue, and indeed, they rate it as not perfect. The cooling fins on the final head are a little ragged, so it won’t be the part that goes on a bike, but we can see with a bit of refining, this process could deliver very good results.

For this pour, they use a gas furnace, but we’ve seen it done with a microwave oven. Usually, you are losing wax, not foam, but the idea is the same.

19 thoughts on “Lost Foam Casting In Action

  1. The kind of guy every vintage bike owner wishes was his in-law. The casting quality on those fins is really impressive for shop-made, incredible that you can get that without big fancy vacuum machinery and whatnot

      1. You want rational behavior from someone saving a British motorcycle that’s not a Vincent?

        Lost foam is used to cast parts that are unmachinable. They glue multiple foam pieces together to make things like F1 engine blocks.

        Granting, the part shown is simple enough to just cut.

        1. “You want rational behavior from someone saving a British motorcycle that’s not a Vincent?”

          Look i want to hate you for this but you’re not wrong.

          Next im sure you’re going to tell me that a drive way shouldn’t have its own exxon valdez every summer and well you wouldn’t be wrong then too but still…I’m offended.

      2. 1. takes a much more significant CNC and more expensive tooling to machine aluminum than foam
        2. machining requires more expensive raw material and the waste of swarf, Casting can be accomplished with literal GARBAGE and near zero waste
        3. while not the case for this part, casting can produce complex geometries that cnc struggles with.

        1. ‘near zero waste’

          So long as you assume the first try at casting will be ‘good enough’ and ignore the post casting machining needed.

          Cheap HSS endmills last a long time on Aluminum.
          You know what rarely comes with an aluminum billet? Voids.
          How many hours labor equals a $100 hunk of a known AL alloy? Sorry about your billing rate.

          1. You, sir, have obviously spent not a single second in a machine shop, home shop, or any manner of workplace outside of your bedroom.

            Has it ever occurred to you that to make a part of that size, a $400 machine that can only cut foam and wood might make far more sense than a $40000 machine that can also do aluminum?

            Perhaps casting is the end goal, rather than machining. Just for the sake of argument, many parts in many applications are made with finished or NNS casting, and/or have been for likely twice as long as you have been alive.

            And, just for the sake of it, where did you get an 8″ cube of aluminum stock, of any alloy, for $100? I can make that in my backyard for about $50, but buying from anyone I’d expect $400 at a minimum.

            . . . Or, skip the hard/expensive work and just let the casting process itself do its thing. You do your thing though.

          2. Believe whatever you want lurker.

            Your estimates of stock size (8” cube? WTF, It’s a motorcycle head, not even overhead cam.) are as good as your estimates of experience.

            It’s at most a 8” square cut from 2” plate. Which gets you right back to $100, based on your numbers. Also consider the machine time wasted by using a 8” cube (that some half blind moron ordered).

            $400 machines cut aluminum all day (literally though). Aluminum cuts like butter. Not really a suggestion. $400 machines won’t have the reach to cut this in foam either.

            The end goal is producing a BSA head that _works_.
            That means it is absolutely going to get a machining pass. Not just on one side. First deck the bottom and drill the bolt holes. Then flip and 3 bolt to fixture while finishing the 4th bolt relief, repeat x3. etc etc,
            You think he’s casting the valve guides and seats too? Oil passages? Intake/Exhaust ports (you’d hope, but not shown)?
            Do you even wrench?

            Home casting runs porous. Better plan on doing it twice, at least (same as the Aussie in TFA). You won’t know if it’s good until the machining is done. Then you’re still hoping for compression.

            I doubt you could pour a 8” cube in your backyard. That’s a big pour.
            If you did, I bet it’s not homogenous and has voids. Especially if cast from melted beer cans.

          3. if the first try isnt good enough, you remelt and recast.
            Your post machining of this part can be accomplished with a drill press, a side grinder, a granite block and some sandpaper. If you dont have a couple of grand for an old bridgeport.
            A cnc is overkill for the task of producing or finishing this part.
            Do you think your car engine block is CNC from billet stock? Wonder why so many parts are cast if CNC is the answer to all.

          4. Not disagreeing with any of the points you make – for a limited run of an aluminum engine part that needs precise tolerances and no risk of voids or other structural issues, pure CNC is definitely simpler than lost foam casting, and you don’t need a fancy CNC machine that costs as much as a small car to pull it off. And sure, you end up “wasting” a lot of the aluminum billet, but it’s not like you can’t easily recycle or re-cast the remnants.

            I mean, I personally wouldn’t want to try and machine an aluminum cylinder head with a $200 genmitsu that I bought off aliexpress, but that doesn’t mean you *can’t*.

            .. but anyway – if you’re in Tony’s situation in the video, where the only CNC machine available belongs to a friend of yours (and that friend presumably doesn’t want to mill aluminum billet for you), this doesn’t seem like a bad option.

            Though I’d definitely feel better about it if he added some more sprues and pulled a decent a vacuum on the thing.

          5. If you’d watched the series of videos you’d know the first castings weren’t perfect, it was a learning process and one that’s been enjoyable to watch.

            Original heads for these bikes are cast so, if you want to keep it as close to original as possible without having some micron perfect CNC part that looks completely out of place, the only way to get ‘new’ ones is to either have them cast or cast them yourself so it’s worth the effort to have a go.

            Especially if you can nail the process and make it a viable way to get replacement parts on demand which are unobtainable through obsolescence.

            Apart form that, it’s fun to learn new skills and do ‘stuff’ the internet nay sayers like you say is impossible, ridiculous etc.

  2. Looks like he is trying to do a 1:1 and not taking in to account the shrinkage of aluminium.
    In an electric furnace this can be 6% in a gas furnace with no temperature control it can be well above that so the fins need to be a little thicker.

    The next issue is the pouring turbulance. Any square edges or sharp angles will result in tourbulance that will cause the molten metal to freeze before getting to where it needs to go. the Sprue should have been round not square to help with this.

    In addition there was no riser, this is important as it allows the metal to flow better. Kind of like filling a bottle of water with a hose. you end up getting a heap of splash back unless there is another hole that allows the gasses to escape.

    Other than that it was an absolultely brilliant pour!

    1. From what I know – also from HaD and YT – it seems that it would be better if he burned away the foam (or dissolved most of it with acetone or other solvent) and also if he heated the mould before the pouring (he complained about this).
      Using foam it doesn’t allow you to compact the sand around the model so I would suggest a 3d printed or made out of wood cast model.

        1. For starters I don’t believe that foam is good in this application for high precision parts. It is full of gas pockets that will get crushed during sand compacting.
          Also cheap foam won’t be machinable to good precision. However, there are two ways to bypass this: use tougher foam – it helps also while machining the template to a higher detail, or use a mix of silica sand and some liquid binder. You’ll get a tough shell, you can burn the foam away and also have it heated properly for aluminium casting. The net is full of how to’s.

  3. Re haha, the bantam engine is a two stroke engine so no valves, valve seats oil passages or inlet or exhaust ports as they are part of the main cylinder, but yes it will need some light machining to finish.

  4. Nycraft oven or small Scutt Kiln… Burn off the form. Cheap bell jar and vacuum pump… Air voids out of the investment/plaster mix. Bridgeport mill for cleaning up and basic surfacing.

    Technically you don’t need the CNC if you simply take an RTV mold of the original as a master and use a casting wax injector to make a wax version of the master. Two pieces of window glass, a U shaped piece of metal, and a 5/8″ wood dowel are enough to form the sprue and the RTV mold.

    A sprue can also be built with wax sticks and a cheap mini torch or soldering iron.

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.