43 replies to this topic

[Moon Tricky]

I know you're all read the subject line and are probably thinking "you're confused. A supercharger is mechanically driven compressor, a turbocharger is a compressor driven by a turbine in the exhaust." And you'd be right about the 2nd bit, only I'm not confused. Allow me to explain.

If you take a turbocharger, and then put a belt or chain on it so that it's driven mechanically AS WELL as by the exhaust, then what happens? You'd get the benefit of instant boost just like a supercharger. But also the turbine at the other end would be acting as a scavenging pump. When the exhaust pressure rises until it is sufficient to power the turbine, then it will do just that, and relieve the burden on the driveshaft - the only loss is in the friction of the mechanical linkage. But when the exhaust pressure increases even further, it can't spin the turbo any faster than it should, because it's geared to the driveshaft! So we don't need a wastegate anymore. All that exhaust that would otherwise have been wasted will instead go to push the turbine, and therefore the driveshaft, thereby directly adding even more power to the engine and increasing efficiency.

I've heard of turbocompounding (an extra turbine after the turbocharger) and twincharging (using both a supercharger and a turbocharger), but I've never heard of anyone ever just connecting a compressor/turbine pair to the driveshaft. But why not?

[Engineguy]

Originally posted by Moon Tricky
You'd get the benefit of instant boost just like a supercharger.

Not really. A centifugal impeller has very, very non-linear flow... the boost provided by a centrifugal supercharger increases with the square of the RPM, so you don't get any serious power increase until the upper 20% or so of your RPM range. A turbocharger, by being decouped, can be be running max rpm while your crank is turning 2000.

[m9a3r5i7o2n]

The only compounded engine I know of that was truly successful was the Curtis-Wright R3300 radial. This engine started out as the B-29 bomber engine. And wasn’t much of an engine at first with all sorts of trouble. At that time it was dually supercharged. Even the future President Senator Harry Truman got in on it. But eventually it was debugged and was used on the Lockheed Constellation. This took about 10 years of development before it became a great engine with terrific fuel consumption figures. A good study of this engine alone would keep one busy for weeks.
This may not be of direct use on your idea but it was a great engine killed by the advent of the jet engine in the 1950s. Maybe someone here has more on this engine than I and could add more to this!

M L. Anderson

[Moon Tricky]

Originally posted by Engineguy
Not really. A centifugal impeller has very, very non-linear flow... the boost provided by a centrifugal supercharger increases with the square of the RPM, so you don't get any serious power increase until the upper 20% or so of your RPM range. A turbocharger, by being decouped, can be be running max rpm while your crank is turning 2000.

That's true... but how about this?
http://www.turbodyne...?dA=TURBOPAC800

That's what gave me this idea in the first place, along with another scheme where a turbine in the exhaust is connected to an alternator. I thought a motor/generator connected to the turbine could be used to give instant boost running in motor mode, and absorb excess boost operating as a generator. Then I thought why not just do the same mechanically...

How about using a twin-screw type supercharger in the usual manner, and then using a turbine as well, also connected to the driveshaft? That would achieve a similar result, although you now need two mechanical linkages instead of just the one, and the power from the turbine now needs to go via the driveshaft to get back to the supercharger. Are there any turbochargers that use anything other than a centrifugal compressor?

[indigoid]

I found this on wikipedia. The usual internet-posting and wikipedia-reading disclaimers apply, and I know even less about 2 strokes than about other stuff, so I have no idea how true it is. It is IMHO interesting, though;

http://en.wikipedia...._diesel_engines

[J. Edlund]

Originally posted by Moon Tricky
That's what gave me this idea in the first place, along with another scheme where a turbine in the exhaust is connected to an alternator. I thought a motor/generator connected to the turbine could be used to give instant boost running in motor mode, and absorb excess boost operating as a generator. Then I thought why not just do the same mechanically...

That has already been done. In reality there are some problems though. The current 12V electric systems are not capable to handle the power requirements of the electric motor. The reliability of the system must also be increased while the costs are reduced. In order to function properly a powerful high speed motor able to tolerate temperatures of about 300 degC are required. Even for a small turbo the motor must be rated at a few kW.

There are mechanical systems too, for example based on hydraulics. But I think that it's too complicated to ever become a reality on production vehicles.

[Moon Tricky]

Originally posted by J. Edlund


That has already been done. In reality there are some problems though. The current 12V electric systems are not capable to handle the power requirements of the electric motor. The reliability of the system must also be increased while the costs are reduced. In order to function properly a powerful high speed motor able to tolerate temperatures of about 300 degC are required. Even for a small turbo the motor must be rated at a few kW.

There are mechanical systems too, for example based on hydraulics. But I think that it's too complicated to ever become a reality on production vehicles.

I know you can get electric superchargers, although you can only use them for short bursts because they draw from a set of batteries. You need about 30hp to get 15psi boost, which is quite a lot of motor. This is why I thought a mechanical system might be better, along with the inherent wastage in turning an alternator to power a motor. There must be some way to make use of that excess exhaust pressure.

How about a simple geared system? The engine is geared up to drive a twin screw compressor, which is in turn geared up to the turbine. Or in reverse, the turbine is geared down to drive the compressor, and in turn geared down further to turn the engine.

Possibly using one of these:
http://www.kennebell...re/BigBore.html

[Calorus]

Originally posted by Moon Tricky


I know you can get electric superchargers, although you can only use them for short bursts because they draw from a set of batteries. You need about 30hp to get 15psi boost, which is quite a lot of motor. This is why I thought a mechanical system might be better, along with the inherent wastage in turning an alternator to power a motor. There must be some way to make use of that excess exhaust pressure.

How about a simple geared system? The engine is geared up to drive a twin screw compressor, which is in turn geared up to the turbine. Or in reverse, the turbine is geared down to drive the compressor, and in turn geared down further to turn the engine.

Possibly using one of these:
http://www.kennebell...re/BigBore.html

Why bother? The major advantage of a turbocharger is the low power to weight and the lack of power dependency. By adding batteries, gears, belts and a host of other superchargers, you add weight, and complication. The best system would be be an infinitly varible compressor A/R ratio Turbocharger.

[Moon Tricky]

Originally posted by Calorus

Why bother? The major advantage of a turbocharger is the low power to weight and the lack of power dependency. By adding batteries, gears, belts and a host of other superchargers, you add weight, and complication. The best system would be be an infinitly varible compressor A/R ratio Turbocharger.

Why? For 3 reasons:

1) to get isntant boost at low RPM,
2) to counteract the parasitic draw on the engine to drive the compressor at medium RPM,
3) to make use of the exhaust gases that would otherwise be let through the wastegate at high RPM.

It isn't called a wastegate for nothing. It's a waste of power.

It'd only have one supercharger, and one turbine, and the purpose of doing it mechanically is precisely so that it doesn't need batteries or motors.

[Calorus]

Originally posted by Moon Tricky


Why? For 3 reasons:

1) to get isntant boost at low RPM,
2) to counteract the parasitic draw on the engine to drive the compressor at medium RPM,
3) to make use of the exhaust gases that would otherwise be let through the wastegate at high RPM.

It isn't called a wastegate for nothing. It's a waste of power.

It'd only have one supercharger, and one turbine, and the purpose of doing it mechanically is precisely so that it doesn't need batteries or motors.

That's the point, if you have a nice VGT supercharger, you can match the A/R ratio to the engine speed, elimating lag, and parasitic losses.

[Moon Tricky]

Originally posted by Calorus


That's the point, if you have a nice VGT supercharger, you can match the A/R ratio to the engine speed, elimating lag, and parasitic losses.

That's true. At least you reduce lag quite significantly (you can't completely eliminate it, but close enough). So it solves the first two problems, but you're still not using the excess exhaust pressure at high engine speeds.

[J. Edlund]

Originally posted by Moon Tricky


I know you can get electric superchargers, although you can only use them for short bursts because they draw from a set of batteries. You need about 30hp to get 15psi boost, which is quite a lot of motor. This is why I thought a mechanical system might be better, along with the inherent wastage in turning an alternator to power a motor. There must be some way to make use of that excess exhaust pressure.

How about a simple geared system? The engine is geared up to drive a twin screw compressor, which is in turn geared up to the turbine. Or in reverse, the turbine is geared down to drive the compressor, and in turn geared down further to turn the engine.

Possibly using one of these:
http://www.kennebell...re/BigBore.html

The turbochargers with a built in motor/generators already produce electricity from waste energy. The electric motors are also much lower rated than 30 hp. The power was just enough to help the ordinary turbine at low speeds/transients.

Some info about the Honeywell e-turbo:
http://www.eere.ener...deer_shahed.pdf
http://www.sae.org/a.../1-113-9-86.pdf

Gears aren't worth the trouble. The turbine only produce excess power at high loads, a condition where the typical engine spends little time. So if you use gears the friction losses at low loads will cause greater losses than gains.

I would avoid mixing turbines with lysholm compressors, the turbine is a much better match with the centrifugal compressor.

[Moon Tricky]

Originally posted by J. Edlund


The turbochargers with a built in motor/generators already produce electricity from waste energy. The electric motors are also much lower rated than 30 hp. The power was just enough to help the ordinary turbine at low speeds/transients.

Some info about the Honeywell e-turbo:
http://www.eere.ener...deer_shahed.pdf
http://www.sae.org/a.../1-113-9-86.pdf

That's some good info. So if the engine speed and load are high enough, the turbo can generate about as much or more electrical power than your alternator, so you could "indirectly add" power to the engine by taking the load off the alternator. It's only a couple of horsepower, but it's free.

I'd kind of hoped there was more free power to be had than that though.

[Calorus]

Originally posted by Moon Tricky


That's true. At least you reduce lag quite significantly (you can't completely eliminate it, but close enough). So it solves the first two problems, but you're still not using the excess exhaust pressure at high engine speeds.

If the A/R is desgned to match at redline, there isn't any excess.

[Moon Tricky]

Originally posted by Calorus
If the A/R is desgned to match at redline, there isn't any excess.

But then you'd be creating far too much boost. The theory of VGT, as far as I understand it, is that it captures as much exhaust energy as possible at low engine speeds so it can spin up the compressor as quickly as possible, but as the engine speed increases it doesn't need to capture all of it to generate full boost.

[J. Edlund]

Originally posted by Moon Tricky


That's some good info. So if the engine speed and load are high enough, the turbo can generate about as much or more electrical power than your alternator, so you could "indirectly add" power to the engine by taking the load off the alternator. It's only a couple of horsepower, but it's free.

I'd kind of hoped there was more free power to be had than that though.

I think you must include the system in a "bigger picture". Say that we remove the conventional 12V system with its alternator, starter motor and 12V lead acid battery and replace it with a high voltage AC motor/generator built into the flywheel and a high voltage lithium ion/lithium polymer battery. Now we can really take advantage of the electric motor in the turbocharger. First off, we have a battery that is able to deliver the kind of power levels needed to speed up the turbo with the motor, the load on the battery can also be reduced by using the flywheel motor to generate power for the turbo. The power generated by the turbo-generator from excess exhaust energy can also be added directly to the crankshaft through the motor built into the flywheel. The motor built into the flywheel can also help compensate for the lack of low speed torque. This means that we could build a very small highly boosted engine that has the performance of current engines, but with the fuel consumption of a very small engine. Perhaps half the displacement (or even less) could be enough for a given power output.

Originally posted by Calorus
If the A/R is desgned to match at redline, there isn't any excess.

A typical turbocharged 2 litre inline four has a power excess from 2000 rpm and up (at full load).

Originally posted by Moon Tricky


But then you'd be creating far too much boost. The theory of VGT, as far as I understand it, is that it captures as much exhaust energy as possible at low engine speeds so it can spin up the compressor as quickly as possible, but as the engine speed increases it doesn't need to capture all of it to generate full boost.

Variable Geometry Turbines or Variable Nozzle Turbines allows to control the output from the turbine over a wider range. The difference between VGT and VNT are that the latter is technology specific, even though in practise, they usually use the same technology; tht is variable nozzle area. The caracteristics of a turbine is similar to that of a nozzle. Reduce the area and the pressure difference over the nozzle goes up, along with the gas velocity in the nozzle. Variable nozzle area therefore allows us to control the pressure difference over the turbine, aka the expansion ratio and therefore the power output of the turbine. In general variable nozzle area turbines are used in a way that keeps the area as small as possible at low speeds and low loads. This gives a high turbo "idle speed" (for short lag) and a high turbine power output with low massflows (boost at low speeds). As the mass flow increase, so is the area of the nozzle as this reduce turbine power to a level needed to produce the required boost. While increasing the nozzle area, the exhaust back pressure (and the associated losses) can be kept low even at high speeds.

[Moon Tricky]

Originally posted by J. Edlund


I think you must include the system in a "bigger picture". Say that we remove the conventional 12V system with its alternator, starter motor and 12V lead acid battery and replace it with a high voltage AC motor/generator built into the flywheel and a high voltage lithium ion/lithium polymer battery. Now we can really take advantage of the electric motor in the turbocharger. First off, we have a battery that is able to deliver the kind of power levels needed to speed up the turbo with the motor, the load on the battery can also be reduced by using the flywheel motor to generate power for the turbo. The power generated by the turbo-generator from excess exhaust energy can also be added directly to the crankshaft through the motor built into the flywheel. The motor built into the flywheel can also help compensate for the lack of low speed torque. This means that we could build a very small highly boosted engine that has the performance of current engines, but with the fuel consumption of a very small engine. Perhaps half the displacement (or even less) could be enough for a given power output.

Now that's more like it. I'm not so keen on using enormous batteries for reasons of weight, and as I'm currently discussing elsewhere I'm a little skeptical of the hybrid concept. But a combined starter motor/alternator/boost motor would be a good idea, especially if it would help make use of that excess exhaust power at high speeds without compromising low-end power.

I'd suggest this is another job for the Ultracapacitor - they don't have quite the energy density of a battery just yet, but can store and release it very quickly.

[Greg Locock]

I don't think UC will ever take over from batteries, although they are good for reducing spike currents.

I think they currently manage around 4 Wh /kg, and there is a new technology that will give you about 10 fairly soon, but that is a long way short of PbH at 30, Nimh at 60 odd and LiPoly at 120

PbH can be made to handle high peak currents well, albeit not especially efficiently, so there is an interesting tradeoff there.

[Moon Tricky]

Originally posted by Greg Locock
I don't think UC will ever take over from batteries, although they are good for reducing spike currents.

I think they currently manage around 4 Wh /kg, and there is a new technology that will give you about 10 fairly soon, but that is a long way short of PbH at 30, Nimh at 60 odd and LiPoly at 120

PbH can be made to handle high peak currents well, albeit not especially efficiently, so there is an interesting tradeoff there.

They're not a viable replacement just yet for long-term energy storage. But I was reading something yesterday about the BMW X5. Although the battery pack of a Toyota Prius can store about 10 times the energy of the equivalent weight in Ultracapacitors, it never actually fully charges or discharges, as it is only efficient to do so within a certain range of its capacity. So you don't actually need to store that much, as you never use it all anyway.

Although I anticipate Ultracapacitors being able to exceed batteries in energy density "eventually", possibly within the next 20 years (don't quote me on that), the total capacity isn't really what interests me. As I said in the electric car thread, the energy you could possibly get from regenerative braking isn't enough to warrant an enormous amount of storage, as even your car's electrical system would be able to use anything recovered in a matter of minutes, let alone enormous boost motors. But they are advantageous in being able to charge quickly enough to actually capture a significant amount of the energy.

In an electric turbo application it's the other way round. You might need to only charge the battery or capacitor fairly slowly, but you need to get it out again in short bursts.

For long term or high density storage, you could always use a battery as well. But in summary I just don't see the need to store large amounts of electrical energy. It's more important to be able to store and retrieve it efficiently and quickly.

[J. Edlund]

Originally posted by Moon Tricky


Now that's more like it. I'm not so keen on using enormous batteries for reasons of weight, and as I'm currently discussing elsewhere I'm a little skeptical of the hybrid concept. But a combined starter motor/alternator/boost motor would be a good idea, especially if it would help make use of that excess exhaust power at high speeds without compromising low-end power.

I'd suggest this is another job for the Ultracapacitor - they don't have quite the energy density of a battery just yet, but can store and release it very quickly.

A certain battery capacity will be required in order to use the system efficiently. For example, we must reduce the number of components directly driven by the engine, such as the power steering and AC compressor and making these electrically driven. Part of the saving with the combined starter motor/generator (sometimes called ISG, integrated starter generator) is to be able to stop/start the engine when required; the engine shoudn't run when the car is stationary (estimated fuel consumption reduction from this alone is 5-7%). Under those conditions it's important that the battery has enough capacity to power all electric driven accesories. It could also be beneficial to run only the electric motors at low speeds, since the internal combustion engine is very inefficient at those conditions. The battery must also be able to handle the power levels generated by regenerative braking.

Ultracapacitors are still not mature enough to be used in production vehicles, and we also need the possibility to store energy for longer periods of time. Ultra capacitors are however possible as supplementary energy storage in future hybrid vehicles, since they are able to handle fast charge/discharge rates. On a side note, there are also flywheel contructions which offers a high capacity per weight and a fast charge/discharge rate. These are investigated for F1 hybrid use.

It's also important that the energy comsumed by the car during it's whole lifecycle can be reduced. Current hybrid vehicles such as Toyota Prius are still beaten by larger cars on this point.

The weight of the complete powertrain should also not be increased beyond the current level. For that to be possible it's important that we integrate the system well. For example the use of ISG's instead of conventional starters and generators as well as high voltage system and high capacity battery replacing conventional electric systems and batteries. ISG's are not only able to replace the conventional starter and generator, but when used as flywheels, they can replace the "dead weight" of the flywheel. Making smaller engines is another important step to keep the weight down.

The battery of the Tesla roadster is said to weigh 450 km and give the car a range of 400 km. Perhaps a battery with a capacity of 1/20 of the Tesla battery is enough for a hybrid, that means a weight no more than 20-25 kg. Not significantly heavier than a current lead acid automotive battery.

[Moon Tricky]

Originally posted by J. Edlund

Part of the saving with the combined starter motor/generator (sometimes called ISG, integrated starter generator) is to be able to stop/start the engine when required; the engine shoudn't run when the car is stationary (estimated fuel consumption reduction from this alone is 5-7%).

The weight of the complete powertrain should also not be increased beyond the current level. For that to be possible it's important that we integrate the system well. For example the use of ISG's instead of conventional starters and generators as well as high voltage system and high capacity battery replacing conventional electric systems and batteries. ISG's are not only able to replace the conventional starter and generator, but when used as flywheels, they can replace the "dead weight" of the flywheel. Making smaller engines is another important step to keep the weight down.

The Toyota Prius engine is actually rather large for its power output, owing to its use of the Atkinson Cycle. Although they've made efforts to reduce its weight as much as possible. Personally, I'd prefer a Miller Cycle engine using an electric turbocharger. Since the first part of the compression is done more efficiently by a supercharger than by the piston, this ought to make it even more efficient as well as increasing its power to weight ratio, allowing it to be downsized further. The disadvantage is in the increased mechanical complexity.

The battery of the Tesla roadster is said to weigh 450 km and give the car a range of 400 km. Perhaps a battery with a capacity of 1/20 of the Tesla battery is enough for a hybrid, that means a weight no more than 20-25 kg. Not significantly heavier than a current lead acid automotive battery.

The battery out of the Toyota Pris weighs about 68kg. But does anybody know how much its motors weigh? If considered as part of the transmission, I'd expect it to add up to quite a lot as far as transmissions go, automatic or otherwise.

[J. Edlund]

Originally posted by Moon Tricky
The Toyota Prius engine is actually rather large for its power output, owing to its use of the Atkinson Cycle. Although they've made efforts to reduce its weight as much as possible. Personally, I'd prefer a Miller Cycle engine using an electric turbocharger. Since the first part of the compression is done more efficiently by a supercharger than by the piston, this ought to make it even more efficient as well as increasing its power to weight ratio, allowing it to be downsized further. The disadvantage is in the increased mechanical complexity.

The Prius also tend to comsume more fuel than claimed. I doubt that the Atkinson cycle is the way to go. Personally I would try to reduce the friction losses as much as possible instead (mainly by making the engine smaller). Then I would prefer a high compression ratio, direct fuel injection and turbocharging so the engine can deliver the neccesary power output. With an engine designed for a high octane fuel such as methanol the thermal efficiency should be high (max efficiency above 40%) while the weight of the engine is kept low. The engine complexity would also be kept to a minimum, since the engine is conventional. At part load it should be possible to use internal exhaust gas recirculation and lean mixtures, while at full load no EGR should be used while running the engine at air/fuel mixtures around lambda 1. A possible development on this is the use of HCCI.

Originally posted by Moon Tricky
The battery out of the Toyota Pris weighs about 68kg. But does anybody know how much its motors weigh? If considered as part of the transmission, I'd expect it to add up to quite a lot as far as transmissions go, automatic or otherwise.

Don't know the weight of Prius motors but I think the whole installation can be done a lot better, both to reduce the weight and improving the performance.

It seems that you can get a 25 kW motor for 19 kg:
http://www.zytekgrou...lt.aspx?tid=152

GM/DaimlerChrysler/BMW has made a gearbox where the two motors of the hybrid system are built in, unfortunatly, they don't mention the weight of it.

[Moon Tricky]

Originally posted by J. Edlund


The Prius also tend to comsume more fuel than claimed. I doubt that the Atkinson cycle is the way to go. Personally I would try to reduce the friction losses as much as possible instead (mainly by making the engine smaller). Then I would prefer a high compression ratio, direct fuel injection and turbocharging so the engine can deliver the neccesary power output.

Oh yes, all these things. I'm an enormous fan of Miller Cycle, but Atkinson cycle does seem like a bit of a waste of engine.

Also with the advent of gasoline direct injection it should become plausible to build a clean two-stroke automotive engine, and then we'd definitely be seeing smaller, lighter engines with the same power output.

Don't know the weight of Prius motors but I think the whole installation can be done a lot better, both to reduce the weight and improving the performance.

Definitely. It's like all they wanted was a car they could call a hybrid.

GM/DaimlerChrysler/BMW has made a gearbox where the two motors of the hybrid system are built in, unfortunatly, they don't mention the weight of it.

I've seen that before. The CVT I designed would be a fraction of the size and weight. You could probably even have four of them (one for each wheel) and it would still be lighter, and you wouldn't need differentials either.

I'm still waiting to get contact details of some professors from the innovations centre. With a bit of luck this thing might actually get built. I want to put it in a Mini Cooper (the original one not the BMW one). They only weight like 650kg, about half the Prius, so if we can't beat their stated fuel economy figures I'll be a bit disappointed...

[Greg Locock]

You gonna seat 4 and pass crash in 650 kg? Props.

[McGuire]

Originally posted by J. Edlund


Personally I would try to reduce the friction losses as much as possible instead (mainly by making the engine smaller). Then I would prefer a high compression ratio, direct fuel injection and turbocharging so the engine can deliver the neccesary power output. With an engine designed for a high octane fuel such as methanol the thermal efficiency should be high (max efficiency above 40%) while the weight of the engine is kept low. The engine complexity would also be kept to a minimum, since the engine is conventional. At part load it should be possible to use internal exhaust gas recirculation and lean mixtures, while at full load no EGR should be used while running the engine at air/fuel mixtures around lambda 1. A possible development on this is the use of HCCI.

Yep. I think that is pretty much the future.

[McGuire]

Moon, I think what you want to accomplish with the boost curve can best be achieved with a supercharger and turbo in tandem. That is the best way to accommodate the two different operating speeds.

[Moon Tricky]

Originally posted by Greg Locock
You gonna seat 4 and pass crash in 650 kg? Props.

The original Mini DOES seat four, surprisingly comfortably. I don't know what its safety rating is, but I'm counting on not crashing it.

[Moon Tricky]

Originally posted by McGuire
Moon, I think what you want to accomplish with the boost curve can best be achieved with a supercharger and turbo in tandem. That is the best way to accommodate the two different operating speeds.

Twin charging is an interesting technique. Although I was lately thinking of simply having a completely seperate turbine that is only used to make electricity, along with a supercharger. If the turbine is generating enough power it could be routed to the integrated starter/generator, otherwise it could just take the strain off the generator. While not 100% efficient, it saves on gears, and anything better than 0% efficiency is better than nothing.

The other idea I had last night was to put a flywheel on the supercharger, and an overrunning clutch, so that when the engine speed drops the supercharger can still keep going just under its own momentum for a while before the engine catches it up again.

[Wuzak]

Originally posted by Moon Tricky
Also with the advent of gasoline direct injection it should become plausible to build a clean two-stroke automotive engine, and then we'd definitely be seeing smaller, lighter engines with the same power output.

Ralph Sarich and his orbital engine company developed just such an engine several years ago. Not sure where it is at now. There were some Fords (Focus' IIRC) running around Perth for a while with triple cylinder 2 strokes.



Also interesting is that Harry Ricardo schemed a two stroke Diesel aero engine in the early '30s which was given to Rolls Royce to help develop. In the late '30s the Air Ministry requested that the engine was converted to petrol operation, and it was run with direct (mechanical) fuel injection and stratified charge. It was the RR Crecy, a 26l sleeve valve 2 stroke V12 which was tested up to 1800hp. Ricardo tested his single cylinder devlopment engine to powers which would have equated to 5000hp in the Crecy.

The crecy had much exhaust energy, which could be recuperated in direct thrust from ejector exhausts, or, in an alternative configuration, could drive a turbine which would power the supercharger and crank.

[imaginesix]

Originally posted by Moon Tricky
...
Also with the advent of gasoline direct injection it should become plausible to build a clean two-stroke automotive engine, and then we'd definitely be seeing smaller, lighter engines with the same power output...

Oooo, that tickles the imagination. Let's see, a 750cc 2-stroke 4cyl boxer that could produce maybe 100-150hp. Smooth, lightweight, compact, efficient, yeah baby! With the state of CFD these days, what's holding it back? Would tractability be an issue? Noise?

[Moon Tricky]

I was thinking about the Miller cycle, and how during all that time where the intake valve is left open part way during the compression stroke, nothing is really being done. I wondered if it could be doing something else during that time, and the obvious thing it could be doing is a scavenging cycle. So here's a graph I made demonstrating this:



As you can see, the power and compression strokes are exactly the same as for a Miller cycle engine, only it gets twice as many of them in the same time. As opposed to a normal gasoline two-stroke which doesn't use the full length of the stroke for the power stroke, as the exhaust port necessarily opens before BDC.

The question is, is the length of time long enough in that gap for the exhaust gases to be replaced by fresh air?

The other question is, how would you actually achieve that timing? Well we'd have to move back to poppet valves for the exhaust, as per a diesel two-stroke. But that's simple enough - plus you get to use all four of them for the exhaust. The intake could be controlled by something like a sleeve valve near the bottom of the cylinder, where in a diesel two-stroke it would be simple piston porting. It would also rely on a supercharger, but then so does Miller cycle.

NOTE: the piston motion in the graph was calculated using an implausibly short rod/stroke ratio, so the timing is a bit exaggerated, but it demonstrates the idea.

[Wuzak]

Sarich's two stroke engines came in two different types.

One was the conventional induction (air only, no fuel) through the crank case and the other was an engine with a sealed bottom end and a blower for exhaust scavenging and air intake. Is that sort of what you are proposing?

[Greg Locock]

I /owned/ a real Mini. It was a terrific car. It weighed more than 650 kg, did not seat 4 normal sized human beings comfortable, and would have been lethal in any crash event.

It was also one of the best cars I have ever owned, don't get me wrong.

[Moon Tricky]

Originally posted by Greg Locock
I /owned/ a real Mini. It was a terrific car. It weighed more than 650 kg, did not seat 4 normal sized human beings comfortable, and would have been lethal in any crash event.

It was also one of the best cars I have ever owned, don't get me wrong.

Wikipedia lists the kerb weight as 617 kg to 686 kg. Of course it could be wrong. And it was designed to seat someone over 6 foot tall, which is taller than I am. At least, in the front anyway. Maybe the back is just for children.

Anyway I found this while trying to look it up:
http://www.treehugge...hybrid_mini.php

and 690kg for the 1992 version:
http://www.carfolio....=48224&id=13970

[Moon Tricky]

Originally posted by Wuzak
Sarich's two stroke engines came in two different types.

One was the conventional induction (air only, no fuel) through the crank case and the other was an engine with a sealed bottom end and a blower for exhaust scavenging and air intake. Is that sort of what you are proposing?

I've never been sure of the wisdom of using the crank case for air intake, other than it makes it very compact for your chainsaw. Miller cycle relies on a supercharger anyway, the idea being that it's more efficient to perform the low-level compression with a twin screw than with the piston.

I wouldn't use piston porting for the exhaust because the airflow is less than ideal without using funny shaped piston heads, and then the combustion chamber shape is less than ideal.

[J. Edlund]

Originally posted by Moon Tricky


I've never been sure of the wisdom of using the crank case for air intake, other than it makes it very compact for your chainsaw. Miller cycle relies on a supercharger anyway, the idea being that it's more efficient to perform the low-level compression with a twin screw than with the piston.

I wouldn't use piston porting for the exhaust because the airflow is less than ideal without using funny shaped piston heads, and then the combustion chamber shape is less than ideal.

These days it's common to use turbochargers as spool pumps on large two strokes as those are most efficient. On a small engine it's still simplest to use the crankcase as the spool pump.

No need for funny shaped piston crowns when using exhaust ports in the liner. Schneurle porting, which is used on most small high performance two strokes uses a flat, or almost flat piston usually combined with a hemispeherical combustion chamber.

[Moon Tricky]

Originally posted by J. Edlund


These days it's common to use turbochargers as spool pumps on large two strokes as those are most efficient. On a small engine it's still simplest to use the crankcase as the spool pump.

No need for funny shaped piston crowns when using exhaust ports in the liner. Schneurle porting, which is used on most small high performance two strokes uses a flat, or almost flat piston usually combined with a hemispeherical combustion chamber.

I see how that works. It doesn't give the desired timing shown in the graph though, i.e. a power stroke longer than the compression stroke. Plus is there not a danger of burning oil? It might be simpler to use the crankcase, but is it better?

[J. Edlund]

Originally posted by Moon Tricky


I see how that works. It doesn't give the desired timing shown in the graph though, i.e. a power stroke longer than the compression stroke. Plus is there not a danger of burning oil? It might be simpler to use the crankcase, but is it better?

A two-stroke engine has an expansion stroke which is shorter that the compression stroke, if that wasn't the case we would blow exhaust back into the crankcase/spoolpump.

When the crankcase acts as a spool pump we can't use a conventional lubrication system. Instead the oil must be injected to the air stream, or the engine must be made "oil free", using solid lubricants, sealed ball bearings, hard materials etc.

[Moon Tricky]

Originally posted by J. Edlund

A two-stroke engine has an expansion stroke which is shorter that the compression stroke, if that wasn't the case we would blow exhaust back into the crankcase/spoolpump.

That's only the case for a simple piston-ported engine. If you use valves at one end or the other, or both, you get more control over timing. It makes it more complex, of course. But then the four stroke engine in a car is already more complex, and nobody seems to have any quarms about making them more complex to improve their performance.

When the crankcase acts as a spool pump we can't use a conventional lubrication system. Instead the oil must be injected to the air stream, or the engine must be made "oil free", using solid lubricants, sealed ball bearings, hard materials etc.

That's what I thought. Not a very good idea for automotive use. Dirty, filthy.

[J. Edlund]

Originally posted by Moon Tricky
That's only the case for a simple piston-ported engine. If you use valves at one end or the other, or both, you get more control over timing. It makes it more complex, of course. But then the four stroke engine in a car is already more complex, and nobody seems to have any quarms about making them more complex to improve their performance.

Even if you fit a two stroke with valves you have to open the exhaust before the inlet. The difference is that you could close the exhaust valves before the intake ports or keep the exhaust valves opened longer in relation to the intake ports. But if that has any benefit I don't know. It's however also possible to adjust the exhaust port timing on two strokes with the ports in the cylinder, this have long been used in motorcycles to improve the low speed torque.

Engine design is all about tradeoffs, if you can gain more than you lose by using a certain solution then use it. On the other hand, if you end up with more disadvantages than advantages, don't use it. For example, in a small two stroke, which is the biggest problem; the current porting or the frictional losses as well as the costs and weight increase associated with the extra valvetrain?

Originally posted by Moon Tricky
That's what I thought. Not a very good idea for automotive use. Dirty, filthy.

Burning oil isn't a problem. Rotary engines are oil burning (so are four strokes to a certain extent) and I haven't seen anyone complain about that. You simply make a oil that is suitable for the task such as a biodegradeable ester oil.

[Moon Tricky]

Originally posted by J. Edlund

Even if you fit a two stroke with valves you have to open the exhaust before the inlet. The difference is that you could close the exhaust valves before the intake ports or keep the exhaust valves opened longer in relation to the intake ports. But if that has any benefit I don't know. It's however also possible to adjust the exhaust port timing on two strokes with the ports in the cylinder, this have long been used in motorcycles to improve the low speed torque.

If you can only adjust the exhaust port timing, you can open them before the intake and shorten the power stroke (the wrong stroke to shorten), or you can close them after the intake and shorten the compression stroke, and waste some of the boost from the supercharger. If both exhaust and intake can be controlled you can move the whole scavenging/intake so it doesn't start until BDC, thereby using the full piston travel for the power stroke, which is what I've tried to show in the diagram.

The first idea I had was to put the exhaust at the bottom, piston ported, and the intake on the poppet valves. That way you could delay the intake timing and get the exhaust opening first and the intake closing last, reduing the compression stroke but leaving the power stroke the same. But it turns out there's a reason it's not done that way round, and it's because of air flow.

Engine design is all about tradeoffs, if you can gain more than you lose by using a certain solution then use it. On the other hand, if you end up with more disadvantages than advantages, don't use it. For example, in a small two stroke, which is the biggest problem; the current porting or the frictional losses as well as the costs and weight increase associated with the extra valvetrain?

I'm not really talking about a small two stroke, but rather, one that could replace the 3.0l four stroke in a luxury sedan or sports car. That's quite large for a gasoline two stroke, although tiny for a diesel.

Burning oil isn't a problem. Rotary engines are oil burning (so are four strokes to a certain extent) and I haven't seen anyone complain about that. You simply make a oil that is suitable for the task such as a biodegradeable ester oil.

I've seen people complain about it. Dirty or not, you've got to carry a supply of it (mixed into the fuel or otherwise), and I gather it's expensive. It seems less than ideal anyway.

[J. Edlund]

Originally posted by Moon Tricky
If you can only adjust the exhaust port timing, you can open them before the intake and shorten the power stroke (the wrong stroke to shorten), or you can close them after the intake and shorten the compression stroke, and waste some of the boost from the supercharger. If both exhaust and intake can be controlled you can move the whole scavenging/intake so it doesn't start until BDC, thereby using the full piston travel for the power stroke, which is what I've tried to show in the diagram.

The first idea I had was to put the exhaust at the bottom, piston ported, and the intake on the poppet valves. That way you could delay the intake timing and get the exhaust opening first and the intake closing last, reduing the compression stroke but leaving the power stroke the same. But it turns out there's a reason it's not done that way round, and it's because of air flow.

Most two stroke diesels use intake ports in the cylinder and exhaust valves in the head, this is called uniflow scavenging.

Since you also will need an efficient gas exchange you can't use the full piston travel for the expansion stroke. Beyond a certain point, making the stroke longer also won't give much of a benefit. The downsides on the gas exchange will be much larger than the benefits.

Originally posted by Moon Tricky
I'm not really talking about a small two stroke, but rather, one that could replace the 3.0l four stroke in a luxury sedan or sports car. That's quite large for a gasoline two stroke, although tiny for a diesel.

Gasoline two strokes of this size have been used in marine applications, so it's not unusually large for a gasoline two stroke. But these days new marine engines of this size tend to be four strokes.

Originally posted by Moon Tricky
I've seen people complain about it. Dirty or not, you've got to carry a supply of it (mixed into the fuel or otherwise), and I gather it's expensive. It seems less than ideal anyway.

You need to carry a supply of oil anyway, so whats the difference if you need to add some once and a while instead of having the oil changed by a shop? The oil needed isn't that different from current motor oils. With direct injection, the oil will not be mixed with the fuel, as that would be pointless.

[Moon Tricky]

Originally posted by J. Edlund
Most two stroke diesels use intake ports in the cylinder and exhaust valves in the head, this is called uniflow scavenging.

Yeah I know, I'm just saying it doesn't work the other way up (i.e. the intake ports in the head and the exhaust port in the cylinder). I guess it's a bit like reversing with a trailer. Which is a shame as it would allow you to open the exhaust port before the intake without shortening the power stroke. Uniflow is still a better option if you don't mind the extra complexity.

Since you also will need an efficient gas exchange you can't use the full piston travel for the expansion stroke. Beyond a certain point, making the stroke longer also won't give much of a benefit. The downsides on the gas exchange will be much larger than the benefits.

Why would the gas exchange be any less efficient just for happening slightly later? Obviously there's a limit to how much longer the power stroke can be compared to the compression stroke. Miller cycle shortens the compression stroke by 20% to 30%.

You need to carry a supply of oil anyway, so whats the difference if you need to add some once and a while instead of having the oil changed by a shop?

I'm too lazy.

[J. Edlund]

Originally posted by Moon Tricky
Why would the gas exchange be any less efficient just for happening slightly later? Obviously there's a limit to how much longer the power stroke can be compared to the compression stroke. Miller cycle shortens the compression stroke by 20% to 30%.

As the piston is approaching BDC the pressure and temperature in the cylinder drops, thus producing less and less torque at the crankshaft. At the same time, when you move the exhaust opening to a later time there will be less time availible for the gas exchange. This is nothing stranger than the valve timing in four stroke engines. If you open the exhaust valve 60 degrees before BDC instead of 30 degrees before BDC there is the possebility of a higher volumetric efficiency and lower pumping losses; meaning more power.