DIY Aluminum Heat Sink Casting – Hackaday

[Peter Wirasnik] has been casting his own aluminum heat sinks. He’s working on capturing the heat from a car’s exhaust system and turning it into electricity, kind of like the candle generator. In the photo above a standard heat sink is bolted to one side of a Peltier cooler with [Peter’s] own casting on the bottom. That casting will connect to the exhaust pipe and transfer heat to the Peltier while the other heat sink keeps the opposite side relatively cool. What results is a voltage between 600mV and 1V.
We’re not quite sure what the end product will be but the casting process is fascinating. He carves the shape of the piece he wants to cast from Styrofoam and embeds it in a box of sand. He then melts salvaged aluminum in a cast iron frying pan using what looks like a propane torch. Once molten, he pours the aluminum into the mold and it burns away the Styrofoam as it fills the void. A little cleanup and he’s got the heat conductive mounting bracket he was after.
ok this might be a stupid question, but could you make the whole exhaust a peltier? by the way a much better way of doing one off pieces and getting WAAAAY better accuracy is to make a wax piece, put it in the sand that has a binder mixed into it. compress the sand, with something as simple as a stick, bake the mold to remove the wax and harden the binder in the sand. pour your aluminum in. aka the “lost wax” method. also its good to see that it was done outside, fumes off of molten metal can be quite nasty.
/ex-iron molder
That is seriously cool!
How many pop cans would that take?
Umm… free air aluminum casting? Maybe I’m missing something but doesn’t this require an oxygen free environment?
No not really, but removal of dross would help a lot. Dross being the metal that was oxidized.
Couldn’t you just strap the thing to a wider part of the exhaust system, like the muffler? That should be relatively flat enough to make good contact on such a small surface as the Peltier, and still plenty hot.
Not that I get the idea here anyway. Pulling 1 volt out of the heat hardly seems worth it, could probably do better spinning a little turbine in the tip of the pipe.
He’s casting a big chunk of aluminum, what’s the big deal? I thought that the article was about those nice extruded heat sinks – that would have been cool since they usually cost a lot.
Doesn’t the oxygen sensor after the catalytic converter require it be heated to 600F? Leeching heat may ruin his emissions quality or cause spurious sensor faults.
I agree about the turbine in the exhaust, it’d probably be a more efficient way of generating electricity as long as it didn’t constrict exhaust flow enough to affect engine performance.
I suspect this is one of those “because it’s there” things 🙂
Interesting method though, you can’t say it’s not creative!
And short of styrofoam this is how things have been cast in sand for centuries.
He is simply doing things the way every foundry has been doing it forever. Although the Styrofoam is a bad idea, it’s contaminating the metal and possibly causing gas pockets in the metal casting.
Metallurgy 090 stuff here. I can build you a forge for $20.00 in junkyard crap.
@MS3FGX I think part of the point is that now he’s made the thing he can add as many peltiers as he can fit in meaning that he can get a useful amount of power. On a secondary note it’s quite hard to make turbines, they have to be efficient (well crafted) and tiny making them very hard to create, so while the idea works it’s too difficult to implement.
Cant wait to get my own car so I can modify it!
I would think that the added weight would negate the voltage gained vs just using a high-output alternator.
However, good idea, and nice execution.
Impurities in the metal affect the heat transfer a lot and this casting method probably won’t yield a very pure metal. Probably good enough though.
One thing that probably would improve the efficiency of this greatly is fins inside the pipe. Cast a whole pipe section with fins inside and flat surfaces for the peltiers on the outside.
@MS3FGX: Peltiers requires high pressures to work well so clamping them against a thin pipe wouldn’t be ideal.
when making the casting, I’d also make sure there were run-off pads. just pockets that would fill up first with the impurities and porosity. In forge type casting like what he’s doing its not really all that necessary for them to be there. But when you’re using injection molding, it’s a must. Worked in an aluminum/magnesium plant for 2 years. fyi for all the tards out there, NEVER mix hot molten metal with water. Even if you’re cooling down the casting. A large hunk of aluminum like that could stay gelatinous in the center for up to 10mins at least. I’ve saw 10 pound slugs of explode and almost kill a man because a water line broke open on it. if water is trapped in molten aluminum it can expand into steam to be at least 10x the original size, and magnesium can instantly explode and flash back at you. Dangerous stuff, just take the proper steps for safety and you’ll keep all your fingers and eyes.
Whats the gain of all this? Surely it is much less than the impact of added weight to the car.
Just the energy needed to cast the aluminum heat sink is going to wildly exceed the energy derived from the peltier.
Nice post on hardware hacking. Last year I read an article how the commercial side of things, is getting closer to using exhaust heat, to be able to supplement a vehicles alternator to the point where a smaller alternator could be used, reducing total vehicle weight. In my mind heating one side of a peltier device is the easy part; cooling the other side sufficiently the tricky part of the job job. I have ideas on how to use the sun for heating, and the earth for cooling for stationary power production. When I do the pricing for modules that spec out the power output, I’d be better of buying plain old solar panels for solar.
Lost foam casting is a pretty neat process. It’s being increasingly used for things like engine blocks and other complicated high-precision parts. It works better if you coat the foam with a thin coat of plaster before packing it in sand. There’s not really any contamination — the metal is so hot it quickly drives off the foam.
As for his furnace set up, dear god. It doesn’t take much to build a decent one.
I didn’t read all the postings, so shame on me if this has already been said. You can leach the heat after the cat and it shouldn’t cause any problems. Putting a fan in the end of your exhaust has two problems, first being inconsistent air flow after the muffler. And the second if back presser. After the muffle wouldn’t be a big deal but befor the muffle would upset the timing of the overlap.
HAHA why put a turbine at the end of the pipe when you could put one attached to the exhaust manifold, inthe car world we call that a turbo!!! BOOST! haha who needs 1v extra power when you have 80 extra horsepower
the heatsink seems kind of thick…
@ Alan The advantage is after the input energy no additional energy is needed for these devices to produce energy in motor vehicle applications. Unlike the a conventional alternator, that adds a load to the engine, a peltier generator use energy that normally dissipated to the air, as the engine goes about it’s prime purpose of moving the vehicle. electrical load. Will a person, lacking deep pockets be afford to build a good peltier generator in their home shop? Probably not, but in this case it’s personal recreation. Most likely using less energy than other recreational activities.
The energy to originally create the aluminum, then re-melt exceeds exceeds the power he’ll produce by about 100,000x. Also, the weight of the battery he plans on storing that energy in will eat more MPG away than he thinks he’ll be getting. Another fruitcake green idea. Things like this are destroying our planet.
All that for a weak 1V source?
@vince. idk(about temp. requirements) but he’s really insulating the exhaust, so it wont get cooler.
this would go alot better with a much thinner heatsink, he is only pulling upto 1v because of that massive hunk of aluminum. copper would transfer the heat better, and the aluminum being used on the cool side to dissipate heat faster.
a much more efficient method would be to have copper (or aluminum) (or any metal that wouldnt melt really), formed to be flat on one side, and hugging the exhaust pipe on the other, about 1 inch thick, so you have a flat mounting surface for the peltier, cutting out the additional weight and the extra distance the heat has to travel to reach the peltier.
Damnit, I was intending to do something like this last year! I accidentally found out that my ceramics kiln can melt aluminum.
Those of you interested in melting metal:
I do lost foam casting at home. Not sure why his casting looks so bad. It helps a lot to start with things that were cast in the first place like engine blocks, scrap hard disks, etc. Extruded stuff like lawn chairs, window frames don’t work so well.
The reason the casts are so rough is he isn’t using green sand, he’s just using straight sand. Green sand is fine sand mixed with either water or oil bound clay to hold its shape and give smoother castings.
Also depending on the source of the aluminum there could be mass impurities. It’s really best to cast from ingots rather than directly from scrap.
It takes 20-32 aluminum cans to make one pound of metal, excluding losses to dross. 20 if they’re all from beer cans, 32 if its from lighter weight soda cans. If you’re going to cast from light gauge scrap it’s best to get a pool of metal going and add the scrap to that to limit the exposure to oxygen, otherwise you’ll get mad amounts of dross.
And as others have said you can make a charcoal foundary for well under $10 if you already have a leaf blower or other high volume air source. Propane ones aren’t that much more expensive if you build your own burners.
Also unless he laps the exhaust mounting block the heat transfer is going to be abyssmal.
Check out for some indepth discussion and lots of pictures.
Dont forget that though it’s only 1v in this case (and that’s fairly low in comparison with whats possible), TEGs are fairly high current devices. It might only be 1v, but it could well be 1A too, and they can be stuck in series and/or parallel to make any combination you like. Just horrifically ineffient (4%ish). But if it’s wasted heat that’s extra for nothing (assuming your pumping losses don’t increase due to more dense gas in the exhaust system).
The reason for not doing other ideas that would net more power- like an exhaust turbine, or finning the exaust pipe- is because this hack theoretically only recovers lost heat and converts it back to usable energy without interfering with the normal operation of the vehicle.
The main detraction of an exhaust turbine or fins in the exhaust that it would really hurt combustion- the whole idea is to get as much air into (air intakes, headers) the engine and as much c02 out of (cat-back exhaust) the engine.
When you restrict that air-flow you disrupt the air/fuel ratio in the EFI (or carburetor if you’re old-school) and sacrifice efficiency, i.e. MPG. So basically you’re recovering heat at the expense of the original fuel.
Though there are other factors to control like the weight of the sink and the battery and aerodynamics, this hack is in the right thinking.
Enjoy your alziemers disease later in life
Good way to get rid of all you old hard drives, and the data on them… You’ll end up with a bunch of steel crap at the bottom though. The glass platters usually don’t melt very well either.
what kind of hard drives are you using? Last time I checked the platters were metal, platinum plated to be more specific.
IBM, Fujitsu, Toshiba, Seagate. Pretty much all of them use glass in their laptop drives. SCSI drives are usually glass as well. Check the end of this video for the glass blobs.
@ not me, There is no concrete link between aluminum and alzheimers disease. Yes alzheimers patients have elevated aluminum in their brains but no one has proven that it got there and caused the alzheimers it could very well be due to the disease.
In a healthy person the aluminum atom is too big to pass through the blood-brain barrier, let alone an aluminum ion.
Besides the amount of aluminum vapor cause by hobby casting that would be absorbed by a person would be so small compared to other environmental sources ie; deoderant, pill filler, paint, cooking utensils &ca.
PS: Teflon pans don’t hurt you either.
here is a better idea…instead of adding shit to your exhaust that looks nastier then you mommas asshole. how about make new products that are already installed on cars to generate power….a bit crazy but for example, head gaskets that have the ability to convert heat into usable energy…..another way is to make use of all those spinning pullys on the car, magnets and a simple cam shaft sensor mounted to an idler pully with magnets embedded into it to get free energy. probably more then just one volt. there isnt a damn thing inside a car that this concept is good for….. as for melting down scrap aluminum….make yourself some detailed sand moulds and have fun.
and there is a unlimited supply of scrap aluminum….check out places like automotive repair shops (old pistons, warped cylinder heads, seal carriers and so on), glass shops, the 2 pane windows are housed in a aluminum frame, they usually give the old ones away if they are not taking it in as scrap for money.
Whenever a heat is just let to dissipate in a car (especially engine block), it should be re-routed to a stirling engine top roduce useful power. For an ICE, 80% of energy is lost as heat anyway… ICEs are destroying our planet, not green ideas.
here is another idea i just had, if you had a big enough peltier, could you put it on your radiator?
@ barry99705
bah, hate that quicktime bs. Heres the link to the vid for any windows users who are smart enough to keep quicktime as far away from their pc as possible.
I tried this years ago. This is not an easy project and will require a heck of a lot more than installation of a heat recovery system. You need to greatly reduce the electrical consumption of your vehicle to make this worthwhile. LED conversion for all lights would be the easiest. There have been a number of studies and a ton of research in industry has gone into this exact application for peltier devices. I feel that this is worthwhile however the physics are a bit disheartening. The surface area in terms of consumer-grade peltier elements that is required to drive a typical car is impressive to say the least. After I had already bought a whole slew of peltier elements I ran the numbers. If memory serves the area required was measured in square meters. Not reassuring.
Now, for all the braindead morons who are shitting on this guy saying that this has no application; an automotive alternator consumes a massive amount of energy. 10 – 15% on a typical car. …more if it’s a small, efficient car. The engine will rev much more freely and your vehicle will consume less fuel if the alternator is eliminated. Today manufacturers are scrambling for fuel savings that amount to less than 1 mpg. Elimination of the alternator amounts to much more than this. There is huge incentive to eliminate the alternator entirely.
Volts do not equal power. 1 volt could provide enough power to run a city if you had a lot of amps behind it. Practically yes, this might be difficult to accomplish however theoretically it’s possible. I don’t see why you guys are bitching about how 1 volt is not significant. I mean, James called you guys out on the series connection thing earlier but he shouldn’t have had to. That is amazingly basic electrical theory.
Magnets on any rotating equipment is simply creating a generator. That is not free energy. A joule of energy will cost you at least as much in gasoline as a joule from the alternator.
First of all, the exhaust absolutely cannot sustain a peltier element. It’s a hateful environment and although it seems good at first you won’t be able to get it to work. Scavenging waste heat from the cooling system is much, much better. Your peltiers will live a much longer, happier life and will thank you for it. You should be able to achieve at least an 80C temperature differential across the peltier, which is much healthier.
Funny enough, cooling the exhaust is actually a bad thing, even on a naturally aspirated engine.
As for my experiment it ended in complete failure. I got my ass handed to me by the laws of physics. Oh well. My alternator still ended up being a fixture on the garage floor. My car’s electrical system is now solar powered. It works well and has been for nearly 2 years.
If you want to increase fuel efficiency, turn your car into a plug-in hybrid (of sorts) simply by installing a switch in series with your alternator field windings. Use the switch to deactivate the alternator (with no current through the field windings the alternator will just spin freely without drawing or producing power). Use a dash mounted voltmeter to monitor battery voltage.
I made the above mods to my car in a couple hours. I drive with the alternator inactive most of the time except for long trips that would drain the battery completely. Most of my trips are short and don’t come close to draining the battery, so I can charge the battery at home instead. I have a 4A smart charger permanently connected to the battery and mounted in the engine compartment. A 3 prong electrical plug hangs out the grill and looks just like a block heater plug. I just plug it in when I park. When I’m leaving, have it set up so I don’t even have to unplug it. I just back away and it unplugs itself.
In my small Chevy Sprint, electrical power accounts for around 10% of fuel consumption, so this method gets me about 10% better fuel economy just by producing my electricity from a source other than gasoline. Yes, the electricity to charge the battery still costs money and may be produced by burning coal, but if you run the numbers, both your cost and your carbon footprint are actually significantly reduced.
The energy density of gasoline is about 9 kWh/L. A litre of gasoline costs about $1.40 where I live, or about $0.16/kWh. Divide by typical engine/alternator efficiency of 10% and I pay about $1.60/kWh for alternator produced electricity vs only $0.08/kWh for electricity from my home. Even allowing 50% battery losses charging from home is still 10 times cheaper than using gasoline to produce the same electricity.
Mods like those above above are MUCH simpler and MUCH more effective than converting exhaust heat to electricity. The above mods will actually pay back on parts, labour, and carbon footprint in under a year
Granted, casting stuff is fun, but saving actual $$$ at the pump from mods you did yourself is more fun in my books. With this and other simple mods, I average about 60 MPG in the city (up to around 70 MPG in the summer and down to 50 MPG in the winter) and sit comfortably in the Top 10 at, (except in the winter when southerners have a significant advantage).
It’s not rocket science. It’s just high school physics. Anyone can do it, and no self respecting hacker with a car has any excuse not to. Figure out what 10% of your annual fuel bill is and see if it’s worth a couple hours of your time to eliminate it.
PS. For an extra 5-10%, consider running your water pump from an electric motor controlled via thermostat to match engine need rather than running continuously like a typical belt driven mechanical water pump.
Big rig trucks use these “modifications” on exhaust and get up to 1KW. They just remove the alternator completely to save on gas. A google search for TEG (thermoelectric generator) is the place to start if interested. Recovering waste heat from fossil fuels will always (almost always for the argumentative people) increase efficiency the question is, can the recovered energy by reused effectively. is even trying to go mainstream with the idea on all sizes of trucks.
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.
By using our website and services, you expressly agree to the placement of our performance, functionality and advertising cookies. Learn more


Leave a Comment

Your email address will not be published. Required fields are marked *