What does GRM think of the Supercharge vs Turbocharger debate?

Here's something I've always wondered about, regarding turbocharging vs. belt-driven centrifugal superchargers:

Centrifugal superchargers produce boost based upon engine speed while turbochargers produce boost based upon engine load. Consider the following example:

We build two identical cars, one with a turbocharger and the other with a centrifugal supercharger. Both produce identical power and both cars have a 10 lb wastegate.

In neutral, rev both engines to 3000 rpm and note the boost. The supercharger’s boost is 10 psi due to being related solely to engine speed and because the throttle is nearly closed (because it’s in neutral with no load.)

The turbocharger’s boost will be very low because the engine is producing little exhaust pressure and because the throttle is nearly closed.

Now apply a load by driving both cars up a step hill, still at 3000 rpm. Turbocharger boost will be 10 psi because the engine is burning a lot of fuel and the throttle is wide open. The supercharger boost will be far less, perhaps 5 psi, because the boost is based upon crank speed and the open throttle decreases the measured boost.

Net result: the turbo car will walk away from the supercharged car. Is this a fair comparison? The centrifugal supercharged car can downshift, increasing the rpm of the supercharger. Or, the gas pedal can be pushed further to builder higher engine rpm to bring the supercharger up to the point that it’s producing full boost. However, this takes time, which is the same as lag.

The key issue is that the pumping efficiency of a centrifugal compressor, be it in a turbocharger or centrifugal supercharger, is proportional with the square of the speed of the turbine. That is, its efficiency will increase by a factor of four by doubling the rpm of the impeller.

Because of this, the centrifugal supercharger, at low rpm, is producing little boost. Either it must be geared to spin faster at low rpm or the engine rpm must be increased until the compressor impeller gets to its “sweat spot” and produces full boost. If it geared to spin fast enough to produce full boost at low rpm, at higher rpm it is severely heating the intake charge.

The point is, under identical loaded conditions, a traditional turbocharger can be not only equal to a belt-driven unit, but superior. This is why big trucks tend to use turbochargers; they work great under load, independent of engine rpm.

Raze wrote:

MA2LA wrote: The biggest problem we are seeing with the VGTs is soot building up and sticking the vanes. Gas engines don't have this problem anywhere near what we do with our diesels. one of the biggest reasons alot of diesels are having issues is that they don't get run hard enough to prevent the buildup. I have been researching vgts for my jeep project.

I agree, I know this is partially true because on our Holset, it had alot of soot deposit buildup from the diesel rig, now after running in a gasoline engine you can eat out of the turbine housing, and even the 3" downpipe that came with that was all sooted up has been magically cleaned out...

soot is from the doser. Some chuckle head put the doser before the turbine stage instead of after like on a proper cummins engine. Ok.. Actually the story goes that on the dodge ram the frame rail was too close to the downpipe and in an accident the fuel line, on a downpipe mounted doser, would be severed and cause a massive fuel leak.... So.. they put the doser before the turbine stage. This sprays raw fuel into the turbine housing. The resulting combustion soots up the turbo.

It's a good simple design on paper, but when you get down to it. It's really tough to make work at an acceptable production price. Don't get me started on Nozzle Ring pushrod cooking or properly dimensioning the vanes on the nozzle rings.... The GD&T still gives me damn headaches.

kb58 wrote: Here's something I've always wondered about, regarding turbocharging vs. belt-driven centrifugal superchargers: Centrifugal superchargers produce boost based upon engine speed while turbochargers produce boost based upon engine load. Consider the following example: We build two identical cars, one with a turbocharger and the other with a centrifugal supercharger. Both produce identical power and both cars have a 10 lb wastegate. In neutral, rev both engines to 3000 rpm and note the boost. The supercharger’s boost is 10 psi due to being related solely to engine speed and because the throttle is nearly closed (because it’s in neutral with no load.) The turbocharger’s boost will be very low because the engine is producing little exhaust pressure and because the throttle is nearly closed. Now apply a load by driving both cars up a step hill, still at 3000 rpm. Turbocharger boost will be 10 psi because the engine is burning a lot of fuel and the throttle is wide open. The supercharger boost will be far less, perhaps 5 psi, because the boost is based upon crank speed and the open throttle decreases the measured boost. Net result: the turbo car will walk away from the supercharged car. Is this a fair comparison? The centrifugal supercharged car can downshift, increasing the rpm of the supercharger. Or, the gas pedal can be pushed further to builder higher engine rpm to bring the supercharger up to the point that it’s producing full boost. However, this takes time, which is the same as lag. The key issue is that the pumping efficiency of a centrifugal compressor, be it in a turbocharger or centrifugal supercharger, is proportional with the square of the speed of the turbine. That is, its efficiency will increase by a factor of four by doubling the rpm of the impeller. Because of this, the centrifugal supercharger, at low rpm, is producing little boost. Either it must be geared to spin faster at low rpm or the engine rpm must be increased until the compressor impeller gets to its “sweat spot” and produces full boost. If it geared to spin fast enough to produce full boost at low rpm, at higher rpm it is severely heating the intake charge. The point is, under identical loaded conditions, a traditional turbocharger can be not only equal to a belt-driven unit, but superior. This is why big trucks tend to use turbochargers; they work great under load, independent of engine rpm.

A few things about this - and I'm probably going to make some mistakes :)

• the SC won't make more boost free revving than it will when under load. You don't measure boost pre-throttle, you measure it post-. So if the throttle's closed, you don't see the boost. I've not seen a centrifugal blower with a wastegate (or more accurately, a pop-off valve), their boost is generally regulated through pulley sizing and nothing else. At least, that's how I've always seen them implemented.
• in the example of pulling up the hill, if the centrifugal is geared to make 10 psi of boost at 6000 rpm, then it will make 5 psi at 3000. Yes, the turbo car will make a bunch more boost and walk away in this case.
• winding a centrifugal engine up to 6000 rpm to reach 10 psi is not the same as the short lag required to spin a turbo up to 10 psi at 3000 rpm. You'd double the speed of the vehicle, for one thing Think of the centrifugal's behavior as analogous to a naturally aspirated car coming on cam. In fact, centrifugals do a good imitation of a rip-snorter of a good naturally aspirated engine.
• you're comparing a centrifugal to a turbo and talking about mid-range torque. You should be comparing a positive displacement blower to a turbo, as they produce a more consistent boost level across the engine's rpm range.
• big trucks use turbos because diesels love turbos, that's your main reason

Keith wrote: ... in the example of pulling up the hill, if the centrifugal is geared to make 10 psi of boost at 6000 rpm, then it will make 5 psi at 3000...

I'm comparing it to a centrifugal-type, basically a turbo compressor with a gearbox that's belt-driven off the crank, not a roots-type positive-displacement blower. (I would agree with you in that case.) Centrifugal compressors work with the square of the rpm, so 5 psi would actually occur at 6000 / 2^0.5, or 4242 rpm, making it an even worse performer, comparatively.

Keith wrote: ... I've not seen a centrifugal blower with a wastegate (or more accurately, a pop-off valve), their boost is generally regulated through pulley sizing and nothing else. At least, that's how I've always seen them implemented.

I was giving them the benefit of the doubt, but if this is true, they're even a worse performer due to the "x^2" boost curve because it's shackled to the crank.

An interesting idea I've had would be to fit a small CVT to the Blower to have it run at a constant RPM regardless of the speed of the belt it's attached to, or does something like this exist?

kb58 wrote:

Keith wrote: ... in the example of pulling up the hill, if the centrifugal is geared to make 10 psi of boost at 6000 rpm, then it will make 5 psi at 3000...

I'm comparing it to a centrifugal-type, basically a turbo compressor with a gearbox that's belt-driven off the crank, not a roots-type positive-displacement blower. (I would agree with you in that case.) Centrifugal compressors work with the square of the rpm, so 5 psi would actually occur at 6000 / 2^0.5, or 4242 rpm, making it an even worse performer, comparatively.

I don't see where we disagree, other than where we get 5 psi of boost - and I'm generalizing in that case. The centrifugal is not the torque monster option.

The positive displacement blower makes boost by essentially stuffing the engine with air. Every time the blower turns over, it moves (for example) 62 cubic inches of air. If the engine normally only swallows 31 cubic inches during that same crank movement, then you've doubled the amount of air going into the engine and that's basically 1 bar of boost. Note that engine speed isn't a factor, so theoretically it'll make the same boost off idle as it does at redline. Of course, compressor efficiency does become a factor but it's still going to be a better choice. In our example, the positive displacement (which also includes twin-screws, not just roots) will make 10 psi of boost at 3000. Again, boost is controlled by the relative speeds of the crankshaft and the blower (technically, a roots-type is a blower while a twin-screw is a compressor). Your boost limits are generally the speed limitation of the supercharger's bearing or the efficiency limits of the supercharger. For example, a 1.2l twin screw (1.2 litres of air per revolution) on a Miata will make more and more power with higher supercharger speeds until the bearings self-destruct. But an Eaton M45 (45 cubic inches per revolution) will get to the point where it will simply thrash around and spend more time heating the air than pumping it. I digress.

The Rotrex centrifugals have a planetary gearbox inside that apparently improves their efficiency somehow when compared to other centrifugals. I think it's because they can spin the impeller faster. Don't know a whole lot about them though, and this could simply be marketingspeak.