h a l f b a k e r yOn the one hand, true. On the other hand, bollocks.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Centrifugal superchargers (akin to the compressor part of a turbo) need to be running at a certain speed before they develop a decent boost pressure, meaning that they can be innefective at low engine speeds. Mercedes have got around this problem by having a 2 speed drive for the supercharger, this would
have the disadvantage of creating a 'step' in the torque curve where it switches between them. In order to eliminate this step, the supercharger can be driven via a continuously variable transmission so that it will always be spinning at exactly the right speed for a given engine rpm, possibly even with respect to throttle position, provided that the variator can react quickly enough.
I don't know what the best CVT type for this would be but I expect the cheaper and simpler, the better; perhaps the old 2 cones and a rubber belt thingy would suffice.
Anderson CVT
http://cvt.com.sapo.pt/acvt/acvt.htm
like this but minus the teeth [squigbobble, May 26 2005]
BMW Alpina B5
http://www.alpina-a...2775B5_evo_0605.pdf
Alpina seem to be doing something right. 20bhp drain at full boost! [kuatolives, May 27 2005]
[link]
|
| |
Why not gear it to the higher speed to start with and put a pressure regulator on the output pipe? Higher wear rate? |
|
| |
Yeah, you'd also be wasting power that way. The idea is to have the supercharger running so the engine is never under- or overboosted so that it doesn't drain any more power than is nescessary. |
|
| |
How significant are the frictional losses of the CVT drive mechanism? |
|
| |
What about the performance effects of spinning up that extra mass when accerating? Anything noteworthy there? |
|
| |
I'm led to believe that most superchargers are driven by a belt off the crankshaft so it shouldn't take a particularly hefty transmission, this should help keep the frictional losses down. My '2 cones and a rubber belt' would resemble the Anderson CVT (see link) but minus the teeth and with the belt transferring power by contact with the cones. There would also be a combined tensioner/derailleur that would shove the belt up and down the cones to change the ratio. |
|
| |
Hopefully this setup wouldn't create much more friction than a couple of normal pulleys and a tensioner. Obviously, the pulleys would be as light as possible to keep their inertia low but I expect it would be insignificant compared to the inertia of the crankshaft, flywheel etc. |
|
| |
I still believe this setup would be more parasitic than simply gearing it higher to start with. |
|
| |
To me, a clutch (or any other additional bits of internal gearing/power transmission) implies added friction which implies more parasitic power loss. What's the harm in blowing off the excess? |
|
| |
oh, and.. //Eddy current clutches operate reliably in clean environments. They are bulky, typically occupying twice the space of the induction motor itself.// |
|
| |
Good point about the clean environment. I'll take back the eddy current clutch stuff on that practical basis, assuming that it would also be impossible to engineer one to work in a less friendly environment. But, eddy current clutches don't present huge frictional losses like a mechanical clutch nor is it subject to the wear that a mechanical clutch is. The system you're talking about already has belts and pulleys. I can't see that replacing a direct drive pulley with, essentially, an electronically coupled pulley (power is not transmitted by mechanical contact) adds a lot of friction. One more set of idler bearings would be introduced, I suppose. The bulk might be an issue. |
|
| |
A supercharger with a blow-off valve to prevent overboost will obviously work since such things are already in use. There is no "harm" in blowing off the excess. But, the excess as described is not free. |
|
| |
I was exploring a way to prevent the excess in the first place without the losses inherent in a conventional CVT. It's a thought process along the lines of how a DC motor's speed can be controlled simply with a variable resistor or in a more complex manner via pulse width modulation. I prefer to take the PWM approach. |
|
| |
See, this is why hybrid cars work so good. The good thing about fuel engines is they make lots of power in a small package. The bad thing about them is they only have good efficiency within a small speed range. Hybrids work by running a fuel engine at one speed (wide open throttle), then converting that energy to electrical and storing whatever isn't used into the battery packs, and when the batteries are full the engine shuts off. If you want better efficiency at a single operating speed scrap the gas/diesel engine, use a turbine instead. That's why jet aircraft work so well! The turbine makes awesome power and is incredibly reliable so long as you keep it at one throttle setting. Hybrids are just crying out for turbine engines. Scrap those crummy transmissions, belts, chains, turbos, intercoolers, blah blah etc, they are just too inefficient.
Another good way to store energy is with a gigantic flywheel beneath the floorboards spun up to monstrous speeds, that way when you have an accident you get this enormous flying disc slicing through the floor cutting your feet off and ripping the interior to shreds... I JEST! |
|
| |
Why not drive the turbo using an electrical motor? The computer can determine how much amperage to send to it, taking into account the associated change in drag through the generator/alternator. CVT would be lots of moving parts, no? |
|
| |
I can see potential for the concept of a CVT driven supercharger, if it were used as a method of throttling the engine. One of the sources of power loss in a petrol engine is pumping against a restriction (the throttle). If the engine were throttled by reducing the speed of a positive displacement supercharger(relative to the speed of the engine), at low speeds some of the pumping energy could be recovered as the supercharger is driven by the induction vacuum and begins to drive the engine back through the CVT.
The CVT itself could be as simple as the type used in mopeds (split pulleys and belt). Alternatively, it could be a closed circuit hydraulic type. Both are proven subsystems known to work in dirty, wet conditions over a range of temperatures.
An electric motor to run continuously would need to be substantial and expensive. |
|
| |
It should also be noted that there are advantages and disadvantages to different technologies. Superchargers are not as efficient as Turbos when it comes to ultimate power capability but they produce much high torque at lower RPMS, turbos are the exact opposite(and have the added benefit of making More and More power the faster they spin). Neither one is as good as the other in their respective optimum ranges(pound for pound a SC will not provide the same power output as a Turbo at Higher RPMS) I really think that the parasitic loss of such a system will far outweigh the benefits. A Supercharger will draw about the same amount of power at 0 throttle with maximum bypass as it will draw at WOT and 0 bypass. The best bet is gear for optimum RPMs on the SC and bypass as needed. The inherent design of your typical roots type SC will lose efficiency as its speed increases(they pull more air than the intake can handle and power output levels off) This is why SCs are so good on diesels as they operate over a very narrow RPM range and so do not have the delay losses that turbos have in spin up time. |
|
| |
I say no benefit, possible hurt, as is far more complex(failure prone)and does not offer any real benefits. |
|
| |
However is way up there on the cool gadgety stuff o meter. |
|
| |
Hmm there's something about this idea I really like. I guess I came here too late to comment when this idea was fresh, but this idea would be great for cutting losses at idle and low throttle (gearing the CVT so the supercharger creates little-to-no boost), yet creating the same amount of boost throughout the entire powerband without wasting power. This would also make a centrifugal supercharger possible and efficient on a Miller cycle engine, since all its boost would be available right off idle. Bun from me. |
|
| |
Centrifigual superchargers are non-linear, boost level is based on RPM^2 (RPM squared). A CVT would compensate for that. |
|
| |
As for CVT efficiency, they are worse than gears, but the type used as car transmissions are better than 90%, with a 4:1 ratio. A supercharger designed for 15lbs of boost from 3500 RPM to 7000 RPM, could drop the boost at 7000 RPM to less than 2lbs with a 4:1 ratio. |
|
| |
Why worry? A supercharger isn't used all the time (as Corky Bell says, when you run full boost for 30 seconds, you are going really fast!). Reducing the superchargers parasitic losses would more than offset the CVT inefficiency. |
|
| |
How bad are the parasitic losses? On a 1.8 liter enginer, 10 HP is needed to provide 15lbs of boost at 3500RPM, 20HP at 7000 RPM. The supercharger is pushing that much air, even when it is not needed. Even though the air is recycled, there are losses due to air flow restriction, and bearing losses - both are RPM^2. |
|
| |
Now, does anyone know someone willing to build this? |
|
| |
This definitely gets a bun from me. |
|
| |
I've had this thought out in my head ever since the Antonov 2 speed SC drive came out. I thought, crickey, there must be a more efficient way to do this. |
|
| |
The theoretical benefits are huge, the engine will always have boost at any given rpm and the supercharger will always be within it's best operating range, greatly increasing it's efficiency.
It would theoretically produce a torque curve which is totally flat, which means more power, more efficiency and more fuel economy. |
|
| |
As for parasitic drag, well, there will be some, but the outright performance far outweighs the disadvantage, just as it would do in a Roots/Eaton positive displacement type SC. |
|
| |
I am giving this serious thought. I plan to do it to my first proper car (I live in italy, so a 500 Abarth is in order) |
|
| |
The hardest part of making something like this work is the control. It would have to be an electronically controlled cvt to pinpoint the exact transmission ratio as the rpms change. |
|
| |
The actual assembly is a no brainer, only slightly more complicated than an Eaton would be. A new crankshaft pulley might be necessary, that's about it. |
|
| |
I think this is totally possible.. anyone with significant knowledge in control systems wants to give a shout? |
|
| |