talk me down - DIY centrifugal supercharger content

So Ive recently begun toying with DIY aluminum casting. Ive been thinking about all the random crap I can make myself. Eventually I want to even produce my own metal working equipment - lathe, Mill, CNC etc. Is it completely out of the question to want to take a stab at a low pressure pulley driven centrifugal SC

essentially heres the recipe:

1. modified crank or other engine powered pulley
2. smaller pulley on the impeller shaft
3. proper Impeller design (copied from some other type of electric blower like in a clothes dryer or furnace blower)
4. some cartridge bearings
5. a cast snailshell (lots of designs available to emulate
6. lots and lots of trial and error

so if I shoot for low pressure (~2 or 3 psi so theres not a lot of tweaking EFI wise) is this really that retarded of an idea? I act kinda dumb most of the time, but every so often I strike gold...or at least copper of iron or some other elemental mine-able metal. If its dumb, say so, if its even remotely feasible, please gimme some hints.

Um, cast compressor wheel = fail IMO, you'd have to get the casting perfect, and any flash would have to be perfectly removed to prevent imbalance because at the speeds even these little bad boys turn you'd seriously have to worry about imbalance, not to mention shear stress on a cast part which I would be very, very leery of. Other biggie is compressor design, you're going to overflow/surge a wheel design not capable of the spin speeds you'll be generating even for 2-3psi of boost.

Whatever you do, make sure you have a couple of screens or something in your intake path to catch all the flying metal before it gets into the engine ;) otherwise why the hell not, if nothing else it's practice in casting complex parts and you'll learn something and or enjoy your time right?

I would think of looking at taking a turbo and casting the parts needed to make it belt driven. Yes it have been done before but it is something that could be done with the really critical parts being prefab. This would make success much more obtainable.

I agree with Raze on this one. A cast, rotating part, spinning at those speeds is going to tear itself apart and end up in the intake. Making a turbo belt driven seems much more feasible. If you do built it, please let us know how it goes :-)

Do a search for Gator Superchargers. DIY home built supercharger using a leaf blower impeller. The guy sells a book on building them. I think I got mine off ebay for about $10. Interesting read. I have got the shaft, bearings and impeller collected, just need to build the compressor housing. The impeller I was planning to use is off a tractor trailer truck. Should move enough air at under 20K RPM to do what I want to do. The bearings I am using are rated at 30K.

I thought maybe going the re purposed turbo route might be a better bet. I didnt think about the impeller explosion bit tho, good point. Im thinking that my pulley reduction would end up around spinning me around 150% of engine speed. Im a boost noob, but given I will be using a pretty small unit, Im guessing that at those speeds, 2lbs boost is pretty doable. Maybe not...dunno. At 2lbs, is a BOV necessary? I thought I remembered reaing somewhere that bov isnt really necessary at that boost. I dont really paln on doing anything else to the EFI except manually retarding my timing about 2° and running 91 octane or better. This is all speculation though as I still havent got a foundry Im happy with. I can melt aluminum, and have, but I havent yet gotten it hot enough to make aluminum fluid enough to move through an intricate mold without freezing before fully filling the cavity. Im going to try my hand at a blower for my furnace similar to this :
and if the blower performs well enough, I was wanting to use the same (or modified for higher efficiency) version adapted for automotive use.

Are you using the Gingery DIY books as a guide? (NFI, etc.)

http://www.lindsaybks.com/dgjp/index.html.

I was going to build one of his DIY furnaces, but found a used Ceramics kiln for $100. IMHO, I'd not bother with DIY impeller/housing. There are literally millions of used turbo's in the US that can be repurposed. (It's raining soup, all you need is a bucket). These are the results of several lifetimes worth of original R&D effort, and millions of $$ of manufacturing infrastructure, to create a commodity. Are you prepared to do your own DIY metalurgy?

Take it from an old guy, and instead make the most out of the most precious resource you have, your time and enthusiasm. Life is too freakin short to spend reinventing the wheel (unless that's your kink).

Concentrate on the overall project, and leverage your capability to envision and perform custom fabrication (casting for example) to integrate Off The Shelf technology into something new and useful.

In your case, that would be figuring a way to spin the impeller fast enough, not the creation of the impeller.

Carter

well said indeed!!!

So if I were to convert a readily available turbo into a Cent. SC, would you trust a base cast then lathe machined pulley attached to the impeller shaft via a flat ground onto the shaft and a set screw in the pulley, or should I go to the trouble of having a keyway cut into both by a real machinist?

These threads always go like this:

"I'm thinking of making X by some means other than conventional wisdom"

and a pig-pile ensues with everyone assuring the original poster a certain death, or at minimum, a painful failure following some huge and expensive effort. As someone who designs and makes things for a living and has supercharged exactly one car, were I to want to do this again I'd begin with one of the inexhaustable supply of Eaton superchargers that were original equipment on many, many cars. If I had to do some casting I'd make a manifold. Making the plug, cores, boxes copes and drags to sandcast that should only take maybe 100 hours. I have a big enough milling machine to face the mating surfaces. Then it's a relatively simple matter of making it all fit, sorting out the mounting brackets, belt tension adjusters, and remapping the ECU to work. Bigger injectors too if I'm going bigger than about 4-5 psi...

But not this time. Seeing as I've yet to see the fruits of a "Bonded aluminum tube and sheet Locost 7 powered by 2 V8s made of 4 FZR600 engines" or any of the other dreamer projects, I'm offering nothing but encouragement.

Cast your own centrifugal supercharger! Get started chop-chop and please keep us updated on the progress. If you complete it and can demonstrate that you made your own parts from your own molds and have a completed functioning unit on a running car making in excess of 4 pounds of boost, I will write you a check for $1000 USD. I'll give you 6 months to complete the project from 8-9-09.

I'm completely serious.

4cylndrfury wrote: so if I shoot for low pressure ~2 or 3 psi (so theres not a lot of tweaking EFI wise) is this really that retarded of an idea?

I can admit I was probably a bit beyond my means with the original post. It was just a wacky idea i dreamt up that seemed remotely feasible. However since my overhead will most likely be <$75 total in the furnace, and probably equal or less in parts (unless a remap is necessary) and you can accept ~2 psi as acceptable, I just may PM you next february

Yup.. you can do it. It's going to be tough. You're biggest issue will be casting a durable wheel with a efficient vane design devoid of inclusions and machining the tolerances needed to have a shaft rotate in the 20-30krpm range. I'd assume you'd have a slower shaft speed, but 70krpm and up is not uncommon on a normal turbo application.

Anyone who says a cast wheel won't work should go to Ross Alumium castings outside Cinncinnati to shut their doors because the product they've been making for over 30 years is going to blow itself to bits. Or you better go stop that man in his big rig because that turbo on his engine with a million mile warranty on it is going to ZOMG!!111 I READ ON THE INTERWEBS THAT YOUR TRUCK WILL HAVE AN EXPLOSION AND BABIES WILL CRY......

People.. How else do you make a wheel in the absence of 5 axis machining? You can't.

I don't have the time to do a full write up, but I can be of some help to your project.... Not a ton.. but some. I did product development engineering with a turbocharger company for a few years.

Ignorant is correct; a lot of commercial turbo wheels are cast. However, I'm pretty sure most are investment cast or possibly die cast, some method a bit more precise than sandcasting. Not to criticize the idea, only suggesting methods - you can do investment casting at home.

MadScientistMatt wrote: Ignorant is correct; a lot of commercial turbo wheels are cast. However, I'm pretty sure most are investment cast or possibly die cast, some method a bit more precise than sandcasting. Not to criticize the idea, only suggesting methods - you can do investment casting at home.

Correct.. Sand casting is used for the cases not for the wheels. Substitute a lot fo commercial turbo wheels are cast with Nearly all.

MadScientistMatt wrote: Ignorant is correct; a lot of commercial turbo wheels are cast. However, I'm pretty sure most are investment cast or possibly die cast, some method a bit more precise than sandcasting. Not to criticize the idea, only suggesting methods - you can do investment casting at home.

Exactly what I was getting at, most DIY casting isn't die, I assumed originally that he was speaking of sand cast which is why I spoke of tolerances and flash...

motomoron wrote: These threads always go like this: "I'm thinking of making X by some means other than conventional wisdom" and a pig-pile ensues with everyone assuring the original poster a certain death, or at minimum, a painful failure following some huge and expensive effort. As someone who designs and makes things for a living and has supercharged exactly one car, were I to want to do this again I'd begin with one of the inexhaustable supply of Eaton superchargers that were original equipment on many, many cars. If I had to do some casting I'd make a manifold. Making the plug, cores, boxes copes and drags to sandcast that should only take maybe 100 hours. I have a big enough milling machine to face the mating surfaces. Then it's a relatively simple matter of making it all fit, sorting out the mounting brackets, belt tension adjusters, and remapping the ECU to work. Bigger injectors too if I'm going bigger than about 4-5 psi... But not this time. Seeing as I've yet to see the fruits of a "Bonded aluminum tube and sheet Locost 7 powered by 2 V8s made of 4 FZR600 engines" or any of the other dreamer projects, I'm offering nothing but encouragement. Cast your own centrifugal supercharger! Get started chop-chop and please keep us updated on the progress. If you complete it and can demonstrate that you made your own parts from your own molds and have a completed functioning unit on a running car making in excess of 4 pounds of boost, I will write you a check for $1000 USD. I'll give you 6 months to complete the project from 8-9-09. I'm completely serious.

Actually threads go like this:

"I have little experience with metallurgy, and know very little about turbo systems, but I want to cast some of the highest stressed parts possibly out of a weak material" And then people point out that he has little experience, and wants to cast highly stressed parts out of a weak material.

And really, what is the hard in letting him know? You bring up the example of a locus chassis... You ever built a car? If a couple random people on the internet telling you it cant be done is stopping you there is NO CHANCE YOU WOULD EVER FINISH DESIGNING AND BUILDING A VEHICLE FROM SCRATCH, the nay sayers are absolutely the easiest hurtle to jump.

4cylndrfury, please don't take any of that as an insult... I am taking it from your own words that you know little about turbos, and are a beginner on casting.

I am just trying to point out that its not like we are here just discouraging any and all ideas that aren't common... Hell, this is the only website I have seen where people drool over geo metro builds. Its just that a freshly cast piece of aluminum that may or may not contain inclusions or voids spun at the speeds we are talking is likely to go to pieces, and the pieces are likely to end up in his engine. If you know enough to know that that is a very likely possibility, its kind of messed up to sit back and encourage that to happen.

Going in the face of convention wisdom isn't a bad thing, but its best to go with the flow, until you have the experience and knowledge to go against it. If you don't know how to swim, you aren't going to make it very far up stream.

Raze wrote: Exactly what I was getting at, most DIY casting isn't die, I assumed originally that he was speaking of sand cast which is why I spoke of tolerances and flash...

The only die casting I'm aware of that is used in turbocharger impeller wheels is a proprietary super secret process that isn't used in production. I can't speak any more.

Investment casting is used for wheels.

In reply to 4cylndrfury: I'd sure trust a DIY pulley over a DIY impeller. Flats work well, just need to make sure it's all reasonably well balanced. The impeller pulley will be small diameter anyway, so if you are using a multigroove belt, maybe an ex-alternator pulley.

Carter

In reply to ignorant: Yeah, a conventional multi-part die would have what, at least one section for each vane? Wonder what they use to make the wax plugs for the investment castings? Lot's of excitement now on using digital printers to create the wax plugs JIT under direct computer control.

Aside: So I'm down in Medellin Colombia on business. At the hotel, I'm watching the Latin American version of the Speed Channel, and there's this half hour feature on a big Brazillian tuning shop that specializes in turbo's. They manufacture their own wheels from billet. Nice shots of a 5-axis NC machining the vanes, followed by balancing, assembly, and testing. Technically, I guess that's DIY ....

Carter

What aluminum alloy do you plan on casting this with? I'm hoping that you aren't just planning on melting down aluminum cans (3104/5182).

If you're new to casting, I'd suggest trying the pulley machined to a turbocharger. Less RPMs to worry about there.

If you're dead set on doing your own impeller, i'd have it sent out to be balanced.

Thanks all for your input. Yeah, I new Id catch some flak for not drinking the "stay-in-line-and-dont-deviate-from-convention" koolaid. But most all of the info youve given me is realistically what I wanted to hear (i.e. opinions from those who have more experience than myself)

So Im thinking Im going to go with a retrofitted OEM type turbo cold side and look into having it machined out to fit a cartridge bearing in where a bushing wouldve been, unless anyone can recommend a small, oem, ball bearing type turbo that isnt ultra rare unobtainium.

My plan is to cast blanks for all my accessory pulleys and turn them on a lathe to balance them and remove as much unwanted weight as possible. As far as the crank pulley goes, I will probably have to have a machinist with precision measuring tools cut the keyway and mark the timing marksm as I am sure I will have to retard my timing some with a compression system in place. One of those pulleys will be a double pulley and will turn the impeller

The material I will be casting is cut up valve covers. From what I have been able to research, alloys used by the auto industry for casting have very little shrinkage upon cooling, so a backyard caster doesnt have to build much allowance for shrink in their patterns.

Yes I will be sand casting....but, Ive found a hybrid lost foam/investment method for achieving very high quality end results.

All this having been said, I still have to tweak my foundry furnace and I want to make/buy used an inexpensive metal lathe - I have looked a lot at the Gingery plans, and think thats entirely doable, but I think purchasing a small used lathe may save a lot of exasperation on my end

Like I said, thanks all for the feedback. I do place a lot of stock in the opinions of all you folks here in GRM land.

If you decide to make a lathe first my offer is retracted. That's just silly. And regarding buying a lathe (which is something I know a lot about) the bigger the cheaper, to a point.

Chinese machines ar all crap, but some can rebuilt to acceptable accuracy and repeatibility. The small home-shop sized older American machinery like South Bend 9s and 10s, Logans, and the ex-high school metal shop Clausings are all in demand as manufacturing no longer requires them and home shop machinists can fit them down the stairs into the basement. If you have space a larger machine will be the bargain.

Depending on where you are, HGR surplus has some exceptional deals on machinery. I know lots of "Iron Hoarders" but try to only have machinery I actually use. If I had space it'd be a struggle not to let multiple 4000# machines follow me home.

Yeah, Id really love to get my hands on a decommissioned lathe thats been replaced by some computer automated whizzbang gizmo. I just know that like you said, finding a great deal on one will get me either a well worn decent machine, or a new Chinese tool from Harbor Freight that will spend more time in the repairs bin than repairing things from the bin Im sure I could build something equally as unreliable as a chinese import for nearly free once I get my furnace and burner put together, but if my old high school shop lathe ever became available,Id snap it up in a heart beat.

I lucked up on an old southbend lathe, but my mill is one on the Chinese ones from Harbor Freight. I must say I have been impressed with it. I did pull it apart when I got it and cleaned up and lubed the slides. Don't discount them just because they are cheap Chinese crap. You aren't using it as a production machine so wear shouldn't be a problem. I have never had anything on mine fail but there is a company that sells all the gears and parts for them as well as a bunch of other stuff so you don't have to deal with HF. I will look for the link to them when I get home this evening.

http://littlemachineshop.com/

These guys stock all kinds of machining stuff.

wow, very nice. Thanks!