patsmx5

iTrader: (16)

Ok, this is me thinking out loud again. I know, I should probably stop. Just hear me out. Criticisms welcome.

The turbine takes energy from the exhaust and converts it into mechanical energy to spin the compressor. The compressor takes a certain amount of power to churn out air at a given volume and Pressure Ratio. Problem is, at low RPMs the turbine can't produce enough power to spin the compressor fast enough to make boost. Consider the following scenario at 3K RPMs at Wide Open Throttle.

The compressor needs X hp to sustain 10 PSI at 3K RPMs. However, the turbine can not produce this much HP because the engine is only taking in 1 Atmosphere of pressure (more or less). Let's us assume it can only produce 1/2 X hp which is not enough for the compressor to make 10PSI, but rather 2 PSI.

Now, if the turbine was magically given enough exhaust to produce X HP for a couple seconds, the compressor would spool and the system would sustain itself. This is the same case as when people spool their setup going down a highway, then use the brake pedal to slow the engine. Once it's spooled you can maintain boost at lower RPMs where you would normally not be able to make boost.

Of course, one way to magically give the engine the extra exhaust it needs is by using nitrous. A quick 50 HP shot of nitrous goes a long way to spooling up a turbo. However, a lot of people do not like nitrous for various reasons, mainly the system is dependent of having a supply of nitrous. Many disadvantages here as nitrous is expensive to refill and nobody wants to pop their trunk to turn on the bottle every time they start their vehicle. You just want to push the pedal and have boost.

The problem: We can't put X hp into the compressor at 3K RPMs.

This thread is to consider and develop possible solutions to the problem. I have a couple in mind.

The turbine imposes a restriction on the exhaust. At 3K RPMs, there is back pressure from the turbine. It's turning relatively slow because it's connected to a compressor that is using all the HP it can deliver. If the compressor was not there, the turbine would spin faster and back pressure would be reduced in the engine.

What if we were to bypass the compressor so that it did not use much HP. Say we shut a butterfly valve at its inlet so that it pulls a vacuum and is no longer compressing air. Simultaneously, we have a fancy intake plumbing system in place that switches the engine over to a regular intake so the engine is still getting air at atmospheric pressure. Now, the compressor is accelerating towards its death and the turbocharger as a unit is building up inertia.

Right as the turbocharger is spinning as fast as it safely can, we simultaneously open the bypass valve that was keeping the compressor under a vacuum and reconnect the turbocharger to the engine. At this exact moment, the compressor is feeding the engine boost. Energy in the exhaust is rising. Shaft speeds are falling as the rotational energy we stored up is being used to drive the compressor.

The question is would the system stabilize and maintain boost? Normally the turbine would have to accelerate the compressor to build up boost, and that takes 1-2 seconds. However, in this scenario it's already spinning and making boost. Would it 'instantly' stabilize more or less? Perhaps the turbocharger as a unit could not develop enough inertia for this to happen. If so, adding a flywheel to the turbocharger to store additional energy would make this possible.

Now, there are several problems and shortcomings for all this to work. We would need a nifty intake switching system that could instantly connect the intake manifold to the compressor, or to a separate intake. We also have no easy way to measure the shaft speed of the turbocharger. We would need something that could look at shaft speed and switch the intake system over right as it reaches its maximum safe speed. Another possibility is to have a timed device that bypasses the turbo for R seconds and then switches it back online. R would be small enough that the turbo could not reach unsafe shaft speeds. Also, adding a flywheel would be hard. It would need to be located between the compressor and turbine. If a flywheel were added, we could easily add a trigger wheel and a Hall Effect sensor to monitor shaft speeds.


Then, there is always the idea of building a mechanical device to spool the turbocharger.

Electric assisted turbochargers have been designed. They use an electric motor to add power to spool the turbocharger. This technology exists, but it has problems too. It's difficult to design an electric motor that can spin 100-150K RPMs reliably at high temperatures.

There is also the idea of having a gearbox and an overrunning clutch tied into the turbos shaft. I have a few different ideas on how to do this, but I'm tired of typing. I'll add them later. For this system to work, I would need the shaft sticking out of the compressor end of the turbo. This might could be done. It likely could, but would be difficult to design and build.

cardriverx

iTrader: (12)

Just one criticism is that if you spooled theturbo as fast as it could go and then switch the intake over to turbo, wouldent it create alot of boost too quick for a boost controller/wastegate to correct? Just thinking.

patsmx5

iTrader: (16)

Maybe, probably, I dunno. I guess either a brilliant physicist would have to run the math or find out by trial and error, beginning on the conservative side. I suppose a BOV in the intake would dump boost so it couldn't get too high. It would just act like a dump valve over say 20 PSI or whatever it's adjusted for.

opek

In my thoughts to get more boost earlier I have considered using a massively undersized turbine wheel and a large, priority fed wastegate to flow the exhaust charge. Sure the difference between the compressor and turbine wheel will slow spool, but with a small enough turbine could this be no problem?

Also, what about Ricardo's sequential setup? It was essentially a compound turbo setup, with a wastegate to bypass the high pressure turbine (what I'm describing here). This may be related to the fact that he used two turbos he had on hand, but he found more power when also bypassing the high pressure compressor. A nice benefit is you can easily remote mount the large turbo.

The_Pipefather

iTrader: (5)

I agree with your idea, but to make it work we need some microprocessor power and some amount of code written for it. The butterly valve you describe would have to be electronically controlled, or a really large solenoid valve of some sort.

Monitoring shaft speed directly is not required. It can be deduced from the compressor map because the manufacturer has correlated it to pressure ratio and CFM (both of which are measured rather easily with a MAF or MAP/VE combo).

The BOV, or compressor bypass valve as I prefer to call it, can be made to act in the reverse direction that it normally acts. That way, when you want the compressor spinning in a vacuum, it can be opened to feed the engine at atmospheric pressure. Of course, then that would have to be a really large solenoid as well.

Its a great project for someone well versed in controls. Though I am not that someone, I could very easily create a model in simulink which can predict system behavior before we put it on an engine.

BTW, the idea of running the compressor in a vacuum to improve its response is not a new one. Check out this picture of a 80's F1 turbo engine with an air cut-off valve upstream of the compressor:

http://gurneyflap.com/Resources/2ndpart.jpg

mazda/nissan

iTrader: (2)

yeah but going WOT the BOV wouldn't open when the throttle is closed, the BOV would open when no more air can be crammed into the motor, which may already be too late

mazda/nissan

iTrader: (2)

talk to lazzer about his electronic BOV, i could see it being used for this. I don't know if he finished it or not, a bunch of upper-crusters around here got termites in their vaginas and banned him before he was done

The_Pipefather

iTrader: (5)

Its already been done:

http://autospeed.com/cms/A_2188/article.html

Not such a hard thing to do except for the controls part of it. The valve is easily available at Home Depot for under $15.

The_Pipefather

iTrader: (5)

Looks like the average lifespan of a thread here is of the order of 2 to 3 hours.

patsmx5

iTrader: (16)

Well, it won't do much good to try to measure MAP or MAF with it bypassed. That won't work.

Was there any info on this already? I mean, did it make a big difference or was it a failure or what?

The_Pipefather

iTrader: (5)

With what bypassed? The MAF?

I've tried searching for info but couldnt find much on the subject. Even SAE does not have anything in its database. Guess this kinda stuff doesn't really get out.

patsmx5

iTrader: (16)

No, if you block off the inlet of the turbocharger then no air would go in it. A MAF sensor would do nothing. It's weird a lot of this stuff never gets tried. When I'm out of college I'll start experimenting with stuff like this, if not sooner.

TeamPLUR

iTrader: (1)

ok, ok . . . im drunk so bare with me, you could use a flywheel, and a centrifugal clutch . . like on a gas powered RC car to only have the compressor spool when the turbine is spooling at a set speed(controlled by spring rates in the clutch), and if you want to get bullet train style with that bitch, switch from wet floating/ball bearing to magnetic hovering bearings
or did i miss the idea completely . . ?

patsmx5

iTrader: (16)

Hmm, centrifugal clutch would do it. Design it so it grabs at a high enough RPM that when engauged the compressor will make boost. If you got fancy with this clutch, it could act as an overrunning clutch too. The following numbers are arbitrary. Have the clutch spin the compressor below 50K and above 100K. The centrifugal clutch would sling out releasing the compressor from the turbine at 50K. However, it would sling out and hit another set of springs. Now, it's got to hit 100K before it can overcome those and sling into the outer race that's attached to the compressor.

The_Pipefather

iTrader: (5)

Designing a clutch that even holds together at 100k+ rpm is a herculean task, one that would rather not be done.

Turbochargers have the response that they do only because of their minimal inertia. Design engineers are loathe to even make the turbine impeller out of steel, which is why most are made from inconel today. Don't even think about hanging something like a clutch off the end of it.

TeamPLUR

iTrader: (1)

ok, so apparently im not THAT drunk, and i was being mostly sarcastic about the magnetic bearings, but because the centrifugal clutch would most likely have to be in the center section, it wouldnt leave much room for anything else, so bringing in the magical hovering compressor, at least on the coldside, where at hotside should retain oil lube- and water cooling, as well at ball bearings for quicker response i assume, the amount of ridiculous sudden boost, could be controlled with some kind of variable geometry on the coldside so as to NOT go from atmospheric to 10 psi in half a second.