Super- and turbocharging - 'twincharging'

[doglegracing]

Interested in twincharging - where a supercharger assist with boost in low boost situations and the turbocharger taking over once it has spooled in high boost. This can be either in series where the S/C blows into the turbo or in parallel (where surge may be an issue).

Any suggestions on how to do this?

 

[TheBlacksmith]

Both Volvo Penta and Johnson and Towers has/had marine diesels that did this. VP used an electromagnetic clutch to engage/disengage the supercharger as needed, similar to an AC compressor. Supercharger feed the turbo. J&T just used flappers to isolate the supercharger as the turbo boost became high enough. Turbo feed the supercharger. Suggest you research along these lines.

Blacksmith

 

[spereira]

Hi, the Lancia Delta S4 was the first car to have this system. I think it was parallel arrangement, the S/C disengaging once the turbo spooled to operational velocity. You can find some information in the net regarding the S4. As an alternative you could place two turbochargers (parallel) one small for lower engine rpm's and a bigger one for high boost conditions. this eliminates lag and allows for a better balance between low speed and top boost performance. The Mazda rx7 has a system like this.
your idea for twincharging is to reduce lag or to have more power? Instead of twincharging why not use a variable geometry turbocharger? decision will mostly depend on engine size, though.

 

[racedoc]

Before going too far down this road I think it is important to define the operating parameters for your project. I have several supercharged/turbocharged maring applications here at the advanced transportation cneter and they work really well for what they were designed for. On the other hand I also had one of the Buick GN test bed vehicles, and still do have a twin-turbo Supra and likewise they both work amazingly well for their intended purpose. Ferrari, back in the 80's, tried a combo system I think they called the Comprex? and did both development testing and at-track testing and chose to run a more conventional twin-turbo package. Of course the whole purpose was to overcome the dreaded turbo-lag, but with today's systems I don't think that is such a problem. For one thing the thinking on camshafts in turbo applications has changed, and twins, increase the turbine/compressor mapping choices available to meet your design goals Even the Buick GN test vehicle would hit at/near max. manifold pressure at stall. What a ride! And, I am not sure that if I was Toyota I would make twin-turbos available to 18 year olds. I have obviously taken the position that you want the thing to go fast and you are not planning on putting it in a boat. On the other hand John Force would probably find it obsurd that we are even discussing turbos in the context of racing, as he defines racing.

 

[TZSiR]

You all have good points, but you info is a bit dated. This was first used in the Group B rallying championship in the Lancia S4, yes. No, it wasn't a parallel arrangement. The turbo was routed right into the S/C after an I/C then ran into another I/C.

Most currently it was tried on the Meguire's Integra Type-R. It was ran the same way using an HKS turbo, and a Jackson racing roots type. It made more power at the wheels at 3,200rpm than it did stock at the crank. 125HP by 2,000rpm. Quite an amazing feat if you ask me. Totalled 424HP with a very fat torque curve for a Honda. If you know Honda's torque isn't something they usually have, but it put down 225lb ft at 2,750. I'd like to take it for a drive.

 

[evelrod]

Just in case ya'll forget and just 'assume' this is all 'new' technology, let's not forget that duplex arrangements date to the 20's and 30's at least. Bugatti and several other makers. Some using as much as 10bar boost!!!!!

Rod

 

[GregLocock]

Comprex is not a combined supercharger/turbocharger, although the plumbing makes it look like one. The rotating element is a series of tubes. These are ported to the exhaust at one end, and the intake at the other. As it spins round (driven off the timing belt, typically) the pressure pulse in the exhaust compresses the intake mixture, which is then vented to the intake. You probably get a fair bit of EGR.

http://www.yet2.com/app/insight/techofweek/9609?sid=200

mentions using a variable speed drive for it, answering my own first objection to it.

Cheers

Greg Locock

 

[aorangi]

Detroit Diesels have commonly one or more turbocharger discharging into the Roots blower. OK, you'll tell me those are 2-strokers, so this doesn't count... Perhaps.

But as an aerodynamic compressor as a better efficiency than a volumetric one, the best way would be to use 2 centrifugal compressors in series with an intercooler in-between (or an aftercooler for better power but slightly decreased efficiency). They could be driven from the crankshaft by means of a CVT and step-up gears.

Then you install a good and efficient exhaust turbine (maybe a two-stages axial one) on the compressors shaft, so that power can flow in both directions from the turbine to the crankshaft at full load and high revs (instead of discharging the excess exhaust gas with a by-pass valve) and from the crankshaft to the turbomachinery at low revs.

You get a turbocompound engine. If you want to make it simpler you can suppress the valve gear and use a 2-stroke cycle, preferably Diesel because with double stage high pressure supercharging you would have to lower so much the compression ratio of a SI engine that its efficiency would be very low. Such a 2-stroke diesel would have a extremely high power density while being quite efficient.

I developed this idea about 20 years ago. Then I found out that an aircraft engine had been successfully built with that system: the Napier Nomad Nm6 ! A fantastic engine. You can find some photos and cutaway drawings of it on the web; just type its name on Google.

Cheers
Aorangi

 

[NeilRoshier]

One of the reasons such systems were developed was to overcome the inherent problems with driving a rootes blower at high rpm/boost and the turbos lag (and other problems). It would seem that both of these issues can be as easily dealt with by utilising a Lysholm blower? Mechanical losses are small, size is compact, effeciecy (adiabatic)is as good as or better than the turbo and it is much more effecient space wise.

 

[aorangi]

Neil,

According to the data I have the efficiency of Lysholm type compressors doesn't reach more than 75% while centrifugal ones attain about 84% efficiency (overall).

Another point is that you can't drive directly - on the same shaft - a Lysholm compressor with an exhaust turbine since they don't spin at the same revs. Even with a reduction gear between it wouldn't work well because their characteristics maps do not follow the same pattern.

With the Nomad's design, all the turbomachinery lies on the same shaft and the power of the turbine directly drives the compressor, just like in a turbocharger. So, the CVT between the crankshaft and turbocharger never transmit a huge amount of power to or from the crankshaft.

There would be no point in driving a (two stages) volumetric compressor from the crankshaft and then recover the full power of the exhaust turbine by means of a reduction gear and a hydraulic coupling to absorb the torsion vibrations. In addition to the lower compressor efficiency, greater mechanical losses would occur. If the Wright R-3350 Turbocvlone used that scheme (but with a centrifugal compressor) it was probably just to exploit the existing Cyclone basic layout.

Let me get a little bit out of my way now!

For wheels traction purposes an even better concept can be envisaged. Let's call it "Turbocompound High Pressure Differential Supercharging" and let's drive the two-stage centrifugal compressor by means of a planetary differential and a step-up gear train. The crankshaft(s) of the (preferably opposed pistons 2-stroke) Diesel would drive the planet carrier, the ring the compressor while the sun would be on the output shaft. The exhaust turbine is then geared to the output shaft. We get a kind of half gas-generator, an engine working as an integrated torque converter. A problem is the excess of compressed air at vehicle launch. At 6 bars (abs) of intake pressure, about 57% of the crankshaft power goes to drive the compressor and the Diesel can only rev at 57 % of its nominal rpm with the output shaft stalled. But the compressor is spinning at full speed and delivers more air than needed for scavenging and combustion.

There are several solutions. Just 3 :
- The excess air can be by-passed directly to the turbine or to a separate turbine.
- Variable pitch stator blades on the compressor inlet would help.
- A variable compression ratio engine could be used( specific design wanted…)

Of course there would be a heat exchanger and an intercooler between the two compressor stages to lessen the power needed to drive it and decrease the thermal load. With a variable geometry turbine inlet, the efficiency would be quite good from low vehicle speed. A clutch or one way clutch to lock the differential would prevent the crankshaft to rev slower than the output shaft and so avoid stall or even reversion of the compressor rotation. A 4-speed powershift transmission would be all right for up to about 40 tons rigs.

Two locomotives using the concept of Dr Leonhard Geislinger (see CIMAC 1955, "A locomotive with thermo-pneumatic transmission" or something like that) were built and successfully put in regular duty in Sweden in 1955 . The worked about the same way, but with two Hedemora-Pielstick 4-stroke engines and an auxiliary combustion chamber to utilize the excess air. Frank Wallace at Bath University also published several papers about his DCE or "Differential Compound Engine", but he apparently ignored the previous work and practical results of Geislinger.

Cheers
Aorangi

 

[NeilRoshier]

aorangi I believe that I was adressing dongleracing's original post and suggesting that the lysholm supercharger will perhaps provide a similar performance to that which he is interested in without the complication. We seem to have different information on the adiabatic eff of the lysholm blowers, I have a range of information tht suggests an effeciency of 70-85% with centrifugal blowers roughly 10% below this (with the attendant disparity of the low rpm operation not considered). I would be greatful if you could point me towards more information on the lysholm pro's/con's as the IHI lysholm blower from the Eunos 800 miller cycle engines interest me greatly.

 

[kimbo1]

If we were living in the 80's I would be trying to combine the two myself. But the 90's saw a massive leap in turbo design (spearheaded by Garret and Nissan) and even better in the 00's. Whilst the theory hasn't changed us mech engineers and materials guys have caught up. I wouldn't dream of trying to combine supercharging with todays modern turbo cars such as the nissan SR20DET's which make boost from 1500 rpm and respond almost instantly with no hint of lag from their ball bearing units. The response is even directly comparable to belt driven centrigugal superchargers (visit Capa.com.au for the type I am talking about).
But then again if you are playing around with old T3's and T04's I understand your dilemma!!

 

[spereira]

evelrod, 10bar boost!!! on just one or two stages!? thought that Bugatti's from the 20's were only supercharged, like the t35. to which bugatti are you refering to?
kimbo1, are you talking about variable geometry turbine or mostly improvement on impeller design?

 

[evelrod]

spereira, I am having a memory lapse today but, quite a few years ago I was given a private tour of Briggs Cunningham's collection of Bugatti stuff including a straight eight duplex supercharged engine. The turbo was a single carb draw through setup (I think) feeding a shaft driven roots type. I don't remember any IC or AC on that engine either.(???)I do recall that it was built like a 'Swiss watch'! The boost figures are from a pre war prototype(?) aircraft engine. A lot of boost for anytime and expecially pre war where metelurgy and design would have difficullty holding up. I don't remember much about it.
As to high boost, a couple of import drag racers using duplx turbos are approaching the 100psi range (Speed Channel). I have no idea how you keep a Honda together at that level. Amazing!

Rod

 

[aorangi]

Hi Neil,
Yes compressors adiabatic efficiency varies with their size. The overall (adiabatic x mechanical) efficiency figures I gave were for quite large ones used on medium speed Diesels. For smaller compressors they are lower.

For the Lysholm compressor on the Mazda Miller engine:

http://members.rogers.com/sofronov/Cars/Mazda/Room/MillerCycle.html

Lysholm website provides compressors maps (pdf), pictures of beautiful supercharged engines and other info:

http://www.rotor.se/lysholm/en/p_1600.asp

Evelrod,
I also wonder about a boost pressure of 10 bars, especially on an SI engine. At about 8 bars (abs) of intake pressure, the compressor power becomes equal to the engine (2-stroke diesel) power, so that all the crankshaft power is absorbed to drive the compressor. You get a gas-generator: an engine of which all the power is recovered on an exhaust turbine.

The 1983 BMW F1 turbo cars ran a modified M10 block with DOHC and 4 valve head, 1.5 liters displacement, merely 4 bars of boost… and produced about 1400 hp… but no one ever knew exactly because it's said the scale of the test bench went up to 1280hp only… gulp !

http://m10-power.320i.com/E21 cylinderhead modifications.htm

http://www.circlebmw.com/parts/mobile/9423036.htm

Cheers
Aorangi

 

[evelrod]

Wasn't that the engine that ultimately caused FIA to go back to NA in F-1?

Rod

 

[NeilRoshier]

aorangi thanks for the links. The published diagrams on the lysholm site show vastly different adiabatic figures to those I have for other manufacturers, an interesting fact in itself! The IHI unit is available used for around $150-200USD which makes it worth considering (for me at least, esp when compared to turbo's) for my application/interest in sprinting and hillclimbing. What do you think would be the outcome of swapping a turbo for a lysholm blower on a 'modern' engine ala' mitsubishi lancer evo 6 equalising operating pressures at maximum?

 

[kimbo1]

Neilroshier,
My comments may not be directly answering your question, sorry, if I can add my two cents worth - you are thinking of ridding EVO VI technology for a supercharger??
Superchargers are generally low boost units - my understanding is that your turbo will crucify any supercharger you may get your hands on considering your workable rev range and your intent to hillclimb, which is probably why no racing I can think of except dragsters use them. But if you want to pull tractors go for the supercharger.
Spereira, I was talking about improved impeller design and ball bearing units such as Garret and HKS.

 

[aorangi]

Yes, I agree with Kimbo. The Lysholm type superchargers might be a little more efficient than Roots ones, but not as much as a centrifugal one.
I see only few advantages in a volumetric compressor over a turbocharger : no lag, a better boost pressure at very low revs and no exhaust backpressure.
But the power to drive them has to be subtracted from the crankshaft power and so there's a greater load on the pistons and crankgear. Say your engine pumps out 120hp at the crankshaft but 20 are lost in the compressor drive, so you get only 100 hp while your pistons and crankgear are delivering 120.
With a turbocharger you recover some of the exhaust energy that would be otherwise lost and the boost pressure you get is more than the increase of exhaust backpressure. The turbine also acts as a silencer, so that a less restrictive silencer can be fitted.
Another great advantage is that boost pressure is a function of the load while with a supercharger it's more or less a function of rpm and remains the same at low load - unlees the compressor is by-passed and declutched.
Turbocharged engines bmep is limited only by detonation, max combustion pressure and thermal load, so that by lowering the compression ratio the limit is extremely high. With a supercharger the limit is much sooner reached as the efficiency drops more quickly when boost pressure is increased over 0.8 to 1 bar.

Cheers
Aorangi

 

[NeilRoshier]

Kimbo I was in no way planning to modify such a car as I do not own one. The question was posed to delve into the pro's and con's of both forms of forced induction. Considering the type of tracks and motorsport I am interested in the lysholm blower may have some significant advantages. It is interesting to me that of the many cars that have been released in the past 3 years with forced induction that a high percentage are now using mechanical forced induction instead of turbo charging (most using a form of rootes blowers) any suggestion why this is.
I am very interested in throttle response, pressure ramp times and packaging for a future project and the mechanical forced induction appears at first glance to have a few advantages. The boost pressure is only a part of the picture for forced induction performance and yet there would also seem to be some tuning advantages for constant boost pressure?