Kinda want four of the 15kW units, one going to each wheel. I like AWD and one with automatic wheel-by-wheel traction control and turn assistance would be really cool, especially in a lightweight car.

I am totally not kidding about the safety/protective equipment. When I was young and dumb I did a pour on concrete. Some of the aluminum overflowed and when it hit the concrete, even though it had been poured 20 years earlier and we hadn't had rain in months, the concrete exploded like a landmine and I got to watch a drip of molten aluminum run down my face shield briefly before it froze. Ever since then I pour over dirt or sand.

One nice thing about 3d prints is you can break them into pieces if you need something bigger than your printer. Just print it in eight chunks that snap together and use wax to smooth the seams. Similarly you can print complex cored molds. I keep thinking about casting an OHV crossflow A-series head. Print the lower section including the chambers, place a pre-formed core for all the oil passages into it, and snap on the top section that has the lower bearing halves for the rocker. It's like green sand casting without the need for draft angles on everything, and no need for split molds. You can even print core boxes for making the cores, albeit you can't bake a printed core box so that'd have to be sodium silicate and CO2 cured.

That render looks amazing, but all the rest of the pictures I get "image you are requesting does not exist" from imgur, and I'd love to see how it came out.

Next time I cast I'll take some pictures. I have a few on my phone but right now my phone and my computer hate each other so it's difficult to get pictures.
I used the same process to make a custom thermostat housing with the nozzle pointed the opposite direction because that's how my stock spitfire radiator is set up. That would have been *particularly* difficult to sand-cast but was trivial to investment-cast.

@lamb_daquiri, somewhat belatedly: print the PLA, 5% infill and layer thickness of 2, cover any weird spots with hot wax, glue all the pieces (form, sprues) together with wax. Shove the whole thing in a bucket of water and roughly estimate volume for choosing how much aluminum to melt.
I mix plaster of paris and casting sand about half and half. What I should do is buy a big sack of Ransom and Randolph Plasticast, which is designed exactly for doing this, and I'll do that at some point. I usually use a plastic jug or a taped-up cardboard form about 1.5cm larger than the mold. Add water to the plaster, mix, vibrate to reduce bubbles, pour it over and inside the mold. Sometimes i weight the mold and pour around it. Sometimes i pour the plaster and press the mold down into it. It's so thick it won't float the mold out. Let that sit for a day, remove the coating/cardboard/jug so it's just a big chunk of plaster, and let that sit for a week to totally dry. (If there's much water still in it, it'll explode.) Ideally just after you pour the plaster you put it under vacuum to suck out all the bubbles. I don't do this because A: plaster sets up so fast my vacuum pump can't pull enough to be worth it, and B: because the PLA is mostly hollow, it would either collapse or get investment sucked into it if I vacuum debubbled it. So I just deal with air bubbles. Extra metal is okay. Not enough metal is the problem.

Casting day. I put the plaster mold upside down in a burnout oven that has a tray in the bottom, so all the crap melting out of the mold ends up in the tray not all over the oven bottom. Make this from steel NOT aluminum because you (I) feel really dumb when the tray melts and makes a mess all over the bottom of the oven. Burnout schedule for most of mine has been 1 hour at 100C, 30 minutes at 150C, 30 at 200C, and 30 at 240C. The wax comes out at 150C, and the PLA comes out at 225C or thereabouts. You have to make sure that your mold and vents are set up so the PLA can pour out. You can vaporize it out but that takes longer and the plaster/sand really begins to fall apart at the high temps you need to vaporize PLA. So you start your burnout, and then you go screw around with the foundry.

The foundry I have set up is pretty cheap. The furnace is an industrial metal paint bucket, like 20 liters. I packed in about 4cm thick of a mix of fire clay and sand and smashed-up bricks as insulation inside the bucket. There's a 30mm hole in the side at the bottom. My crucible is a chunk of steel pipe about 10cm in diameter, about 2mm thick, with a bottom welded on. I built a long pair of pliers to insert and extract it from the furnace. Originally I fired it with coal or charcoal with an air blower in the 30mm hole. This was an unbelievable pain in the butt. Now I'm using a homemade Reil burner running on propane. It takes about 45 minutes from when I light it to when I've finished pouring. Since I have about 50 liters of used engine oil I will at some point build an oil burner. One thing at a time. (There are good reasons to not use a steel crucible with aluminum. Graphite crucibles are cheap.)

So, light the burner by throwing some weeds and paper in the furnace, lighting them on fire, and turning on the propane through a nice needle valve. Usually I need to have the crucible in there because when it's cold, the flame blows off the end of the burner and extinguishes unless it has a little back pressure. I throw in whatever junk I have around, with the proviso that these days I have junk that was mostly cast already. Pop cans make lousy casting aluminum. Smashed-up heads and pistons and conrods make _fantastic_ casting aluminum. You can also cut it with bicycle parts and window frames and whatever else, just as long as at least half of it was originally cast aluminum. I have a top for the furnace. It's made of the same stuff, the same diameter as the paint bucket, but I put a pop can in the center of it when I made it and now that's the vent.

By the way. Mixing your furnace refractory out of fire clay and sand and broken bricks is really antiquated. You can buy castable refractory at home improvement stores, for repairing fireplaces, and it is WAAAAAY better than my DIY junk.

Put in just a little aluminum, like a muffin-sized chunk, up against the side of the crucible where the flame's hottest, and once that's melted add more. The increased surface area of the liquid aluminum transfers heat way faster and you save propane. Once you have as much as your displacement test said, plus 30%, you're ready to pour.

Back to the oven. Oh so carefully take out the mold. It's probably cracked. That's fine. I pour in a big container full of sand. I dig a hole in the sand, put all the pieces of the mold, which have cracked and fallen apart, back together and blow out any junk, and stick it in the hole in the sand, and scrape the sand up against the sides. This seals all the cracks so the aluminum will leak out a little but won't just pour out. I put some big steel plates on top so I don't float the top of the mold up on the aluminum if it cracks more.

The aluminum is ready when you can just barely see it glowing in the dark. There are other ways to test temperature that are more accurate but this works. Ideally you degas it by taking some salt substitute, potassium chloride, and wadding it up in aluminum foil and jamming this down into the melt so the KCl bubbles through it. You can also pour some borax laundry detergent on the top to help reduce oxidation. Just before you pour you scrape the oxides off to one side with a spoon or something. Then you put on EVERY BIT OF PERSONAL PROTECTIVE EQUIPMENT YOU OWN AND ANYTHING YOU CAN BORROW. Seriously. I have safety glasses and a face shield and a leather hat and a leather apron and big leather boots and a jean jacket. Coz when this goes wrong it shoots molten aluminum right at your face.

And then you pull the crucible and pour in a nice steady stream until it starts coming out the vent holes, and then you slow way down and pour very slowly for a second longer, and then pour the rest in muffin tins you bought at the dollar store because if you nick your wife's tins she will kill you. Turns out they have nonstick coating. Well, had.

And wait for 10 minutes and chuck it in a bucket of water.

Make all sections at least 2mm thick. Cast dimension-critical parts once first, because they're gonna shrink about 2-4%, and you want to see how much it actually shrunk and oversize your next print by that much. On this casting I printed the bottom and top plates separately to assure sizing.

Aforementioned plasticast won't crack into pieces during burnout. That's one good reason to use it not plaster-and-sand.

You can also do green sand casting, where you put your form in sand. It's reusable and works really well. But some mold shapes (like this one) are pretty complicated to do in green sand. It's best for hat-shaped things, not for stuff with multiple flanges and internal recesses. (That's perfectly doable, but takes practice.)

BTW historically fours had a weight and power advantage because a flat-plane crank could be forged rather than cast, so had a really good strength-weight ratio. These days industry can handle multi-axis forging, so that's no longer such a big deal.

I'd love to have longer runners on the intake, but there simply isn't anywhere to put them. I thought about doing something similar to the BMW 2002tii setup, with intake runners that sweep up 90 degrees, then bolt to a manifold of runners that go over the top of the valve cover and put the fuel system on the other side of the engine. I'd love that, as I have a lot of room over there and could some day fit a supercharger under it. But to get the airflow in the space I have, I'd need to cast elliptical tubes only about 22mm high and really wide. I'm still thinking about it. But mild supercharging and fuel injection seems to be a lot easier.

I did a previous model of these that had a pad on top for a balance tube. If these are trouble, there's a point where there's enough metal thickness that I could tap in a pipe fitting and connect them with a hose. I'm hoping that it'll do okay without.

Wowie yeah this motor definitely didn't come with the correct starter bolts. One's a 15 and the other's a 16, and the upper one is just about impossible.

I have an A14 in a Triumph Spitfire. There's about 4mm of clearance between the oil filler cap and the hood. This means there's a big hole in the hood to fit the downdraft carb.
I designed and 3D printed two intake manifolds out of PLA to adapt an A14 to a pair of sidedraft SU carburettors, coated the 3D prints in a plaster of paris/sand mix, burnt them out in an oven, and cast them in aluminum, then sawed the casting into its separate pieces and milled the mounting faces of the intake manifolds flat and parallel (and the same height.)
There were a number of issues: one bit of the mold didn't fill so I had to weld it up, the way I made the 3D print with switchable plates to fit the two differently-handed carbs ended up leaving weird divots on the inside of the casting, and the top face of the castings don't have the detail I'd like, plus just pouring the plaster over the surface meant there were big air bubbles on the bottom sides of the top plates.
However, they slip right on the A14, they'll seal well, and they fit the SU's.

I also designed them with fuel injector bosses so if I want to go down that route I have a good start. (In the long term I'll get a Hayabusa throttle body set and cast another pair of these with diverging rather than converging ducts to fit the Haya spacing.)
I hope the included pictures come through okay.

Attach file:

---

  PLA_print.jpg (22.10 KB)


  rough_casting.jpg (39.94 KB)


  sawed_loose.jpg (36.82 KB)


  on_engine.jpg (33.05 KB)


  milled_and_drilled.jpg (39.12 KB)