12 replies to this topic

[Cociani]

Smog Pump Supercharger,

I read recently on the Internet that people have successfully converted automotive air pumps(smog Pumps) into superchargers for smaller engines. People have even used a series of these pumps to supercharge a V8.

I did a little research into these air pumps and as far as I can tell they are a vane type compressor. If that is the case they would work in a similar manner to the 1950's and 1960's vane type superchargers.

I could not find a single blueprint or internal diagram of a smog pump however so I am not 100% sure they are a vane type of pump. Does anybody know more about smog pumps and their internal design? Has anybody used one as a supercharger? Is there a source to find out what type of air volume these pumps produce and the differences between individual pumps?

I love reading about resourceful performance modifications does anybody else have any stories of dirt cheap performance modifications?

P.S. I did do a search before posting this thread and found the previous thread did not answer the questions I have asked and as I could not post to an archival thread I though I would start a new one. Here is the original thread for refferance.

http://forums.atlasf...light=smog pump

[Chevy II Nova]

Um.... Roots superchargers were used as air pumps in mining operations before they were bolted to V8's.

[Cociani]

I found a site which gives formulae for calculating the capacity of vane type superchargers here it it.

http://daphne.bio.fl..... Formulae.htm

I still have not found any conformation that a smog pump is a vane type pump. Anyone dissasembled one?

[McGuire]

An automotive air injection reactor pump (aka "smog" pump) does indeed employ positive-displacement, vane-type design. The vanes are invariably phenolic plastic, expanded radially by light coil springs. Most I have seen are four or six vane units. The rest of the rotating assembly is more substantial, with a heavy shaft and sealed ball bearings.

These pumps are typically designed to rotate about 1.25 times crank speed, and owing to their original use, develop only slightly greater pressure than exhaust backpressure (less than 2 psi). I wouldn't hazard a guess at the volumetric capacity anymore -- can't remember. They have been successfully used as superchargers in small, low speed engines, I understand (Briggs & Stratton etc). However, I am doubtful they could be used successfully in series in an automotive-sized engine. They aren't particularly efficient, and wouldn't get any better when pushed. I think you would need charge cooling stages to get any real boost, which would defeat the cost angle.

One interesting aspect of the most commonly found AIR pump design: while the pump exhausts through a conventional 3/4" pipe nipple on the back of the pump, the intake is through a fan mounted behind the drive pulley, which also serves as a centrifugal filter. Thus, for this type of pump you are essentially restricted to a "blow-through" intake layout.

[Cociani]

Thank you for your reply McGuire, would running several pumps parallel work rather than in series? You would of course reduce the frictional heating effect. Would a pre-supercharger water injection system on a series setup help with cooling the air temperature? If the system is sufficiently cooled would running the pumps (superchargers) at a speed faster than 1.25 x's of crank speed increase the boost pressure to a usable level like 5 or six pounds?

I 'm wondering if such a setup would be useful on something like a Harley Davidson Ironhead Sportster which has a displacement of 1000 cc and revs up to around 5000 RPM. These motors are fed by a single carburetor.

[McGuire]

Mounting the pumps "in parallel" will increase the volume but not the pressure. However, increasing the rotating speed will increase the pressure. I would guess the rotating assembly could take 3x crank speed, but I doubt if the pump's volumetric capacity would track proportionately.

Can an A.I.R pump (or two) be made to produce enough capacity to serve as a supercharger on a 1000cc Ironhead Sporty? I don't know, but I suppose there is one way to find out! Since the mounting and drive system for a motorcycle would be fairly involved, I would do some tests first.
One could build a simple blower dyno, but that would require an electric motor of at least 10 hp (big $$$ unless you happen to have one lying around). Easier, cheaper way: instrument for pressure and volume and test it by driving it off the pulleys on your street car. To make 7-8 psi of boost at 5000 rpm with your engine would require a pumping capacity of over 130 cubic feet per minute. I don't think a smog pump can do it, but I could be wrong about that.

[Cociani]

McGuire,

You're idea of testing a pump makes sense but I do have a few questions. What type of instrument would you use for measuring volume and how would you set up the test. A pressure gauge for an air line would work for measuring pounds of pressure but I am stumped on the volume question. Do you block the exhaust line from the pump? Sorry if my questions sound a bit stupid but I am having trouble visualizing how all this would be done. Using a car engine to drive the pump makes perfect sense to me it is the measuring apparatus I need help with.

The other thing I have read is that cars which used thermal reactors rather than catalytic converters used a higher volume air pump. Would those pumps possibly be higher pressure too? I remember reading that they were used on cars like BMW's. Were they also used on any American cars? I am just wondering if a thermal reactor air pump might be a better pump for my devious purposes.

Another thing I was wondering is that if one were to use water injection on such a setup if it would work best before or after the pump, (supercharger) I you inject the water before there may be some cooling effect for the pump but perhaps the injection after would cool the charge better before it reaches the carburetor. Any thoughts?


Thank you again McGuire for all your help.

[Greg Locock]

The cheapest way of measuring volume is an orifice plate (a plate with a hole in it). Use a U tube full of water to measure the pressure difference across it.

-------1---------------------------------2------------
|
|



-----------------------------------> air




|
|
-----------------------------------------------------------


ASCII CAD!


volumetric flow rate=C*A*sqrt(2*(p1-p2)/(airdensity*(A^2/a^2-1)))

C is around .62 to 0.7

a is the area of the orifice, A is the area of the pipe

if the height difference measured by the U tube is h then p1-p2=waterdensity*g*h

To make a U tube manometer just get a piece of wood and a long piece of fuel tubing. It's the cheapest bit of accurate gear you'll ever make.

[Cociani]

Originally posted by Greg Locock
The cheapest way of measuring volume is an orifice plate (a plate with a hole in it). Use a U tube full of water to measure the pressure difference across it.

-------1---------------------------------2------------
|
|



-----------------------------------> air




|
|
-----------------------------------------------------------


ASCII CAD!


volumetric flow rate=C*A*sqrt(2*(p1-p2)/(airdensity*(A^2/a^2-1)))

C is around .62 to 0.7

a is the area of the orifice, A is the area of the pipe

if the height difference measured by the U tube is h then p1-p2=waterdensity*g*h

To make a U tube manometer just get a piece of wood and a long piece of fuel tubing. It's the cheapest bit of accurate gear you'll ever make.

Ok so allow me to paraphrase in order to see if I understand. I take the same diameter tubing that I would run to the carburetor from the pump and put a plate at the end with a hole in it. How do I know what size to make the hole? Now inside the tubing before the hole I place a small tub of water with a u shaped tube in it which is comes out the side of the "manifold" and points straight out. I then have to graduate this tubing to measure how much the level increases when the pressure increases? How do I know what type of graduations I should make? Should they be in millimeters etc.? Do I still use a common air pressure gauge to measure the pounds of boost?

Sorry if my questions sound moronic you probably explaned everything very well with your formula, I am sadly not well skilled in math beyond arithmetic. In my line of work I use financial calculators and computers to do any calculations and simply never use algebra at all. I have worked in an automotive machine shop and have always repaired my own cars and consider myself pretty handy at the hands on practical understanding of mechanics. I also do understand the theory of how induction exhaust etc. work but I fail when it comes to the mathematics of the whole thing. One of my longer term goals is to improve my math skills. I have bought a computer tutor to help me with that but have not had much time to work at it. If you could give me some more idiot proof instructions I would be eternally grateful.

Thank you for your help.

[Greg Locock]

That damn ASCII CAD didn't work, the plate is supposed to be between the two pressure tappings.

One end of the U tube goes to each pressure tapping. The U tube needs to be rather less than half full of water.

h should be measured in m, for that formula to work with no constants in it, but obviously you'd use mm in practice and divide them by 1000.

Sizing the orifice and the U tube will depend on what volumetric flow rate you are expecting. If you make the orifice small compared with the carb then it will be more sensitive to flow rate, but will also choke the flow, not a good thing.

[McGuire]

That was an excellent suggestion regarding manometers. They never run out of uses. I think for output testing (not running into a carburetor on a vehicle) an orifice of oh, around 50% should give you a nice signal. The exhaust port of these pumps is typically around 1", so 3/4" (19mm if you want to work in SI) should do nicely.

However, the more I think about it, the more skeptical I am about smog pumps as superchargers for engines of any real size. I still haven't found any published pumping capacities. But they are just constant-displacement pumps, and I think a generous estimate of their displacement might be around 6 CID. That means that in order to feed a 61 CID engine at 7.5 PSI (1.5 atmospheres absolute, more or less) it would have to rotate at over 15 times crankshaft speed, or 75,000+ rpm at max horsepower rpm. Just to keep up with the engine's pumping capacity would require over 10x crankshaft speed. At that speed you can forget about driving it with a simple belt or chain setup, and there is no way that little pot-metal pump with plastic impellers could ever handle centrifugal loadings of that magnitude anyway. Ka-blam.

The visualization sort of reminds me of something Keith Black once (purportedly) said. Some friends called him up and said they had just patched together a dyno rig for testing Top Fuel engines and superchargers, and they had one loaded up with 98% nitro ready to go. Would he like to come over to their shop and watch the first pull, they asked. Black said, "Hell no! I'm too close now."

[Cociani]

Thanks again for your help McGuire and Greg Locock.

When I have some time I think I will procure a smog pump and test it at 3 x's crank speed to see what sort of results I get I don't think I will try the 75,000 r.p.m. test as the results seem obvious, a disaster.

The reason I thought it might work is that a regular smog pump has been used on 5 HP Birggs and Stratton engines and those engines displace around 300cc if I am not mistaken. I thought perhaps using the most effective pumps I could find two may be enough to work on a 1000 cc engine.

Whatever happens I think it will be fun to build the test rig and try it out. Perhaps if the smog pump does't work out I can design and build a Vane type supercharger of my own. More testing and more fun!

[McGuire]

Originally posted by Cociani
Whatever happens I think it will be fun to build the test rig and try it out. Perhaps if the smog pump does't work out I can design and build a Vane type supercharger of my own. More testing and more fun!

We can all use your attitude. It's not the destination; it's the journey.