Twincharged setup, Supercharger THEN turbo?

[Warmington]

Hi,

I have just been directed to and enjoyed reading the following post regarding twincharging setups, and correct SC/TC sizing:

thread71-174690

Some excellent words of wisdom from Warpspeed and Pat!

I have just designed and almost finished installing my own twincharged setup on a very small engine. It is 993cc, 3 cylinder, 12 valve that was originally TurboCharged from the factory. It has a IHI RHB5 turbo charger with a larger than standard compressor from a 1.6 litre RHB5 equipped car.

I have installed an eaton M24 (tiny!!) SuperCharger which is constantly driven without any clutch mechanism at 2 x engine speed. The SC feeds directly into the TC, providing around 7psi of boost at the TC inlet. When the engine manifold pressure (post Throttle Body) exceeds 10psi, a butterfly valve opens to allow the TC to draw air around the SC from a common filter. The SC is not blocked at input or output and continues to supply air.

After the TC there is an InterCooler, then the Throttle Body.

In addition I have used a recirculating dump valve from a large engined car which is connected to the pipework just after the TC, but before the IC. It's return is plumbed back into the SC intake port. The dump valve allows the boost from the SC to continually pass through the TC, keeping the compressor spinning, but does not allow the boost to affect the MAP sensor at idle or small throttle openings.

The engine has not been run properly yet, so it is purely theoretical, but soon to be tested!

I read with interest the theory of boost multiplication (or rather pressure ratios) when the TC provides the SC with pressurised air. I realise I will not benefit from this phenomenon in my setup due to the TC not being a positive displacement pump. Also I realise that my turbocharger should still be sized for max boost and flow as if the engine were TC'ed alone.

What I am hoping for is a lift in bottom end torque and quicker spooling of the turbo. The top end performance was previously about as far as I wanted to go with regards to traction issues with the TC alone, so outright power is not a huge concern (it was producing a reliable 135 Bhp at the wheels with 16psi in a very light car).

Would it really be worth considering rearranging the pipework so that I have the SC post TC, and choosing a larger turbo? Could I achieve similar power outputs whilst increasing engine responsiveness and efficiency by doing this? Is the decrease in exhaust back pressure from using a larger turbo going to give me a better 'feeling' engine with the SC?

Any advice or opinions would be gratefully recieved!

Greg

 

[Warpspeed]

Hello Greg, it sounds like an interesting project.

One thing I have never tried is putting a positive displacement supercharger ahead of a turbo, so all I can do is speculate on some of the problems that may possibly arise.

The first observation is that the supercharger being a positive displacement air pump virtually meters the air volume going into the engine. So placing a turbo compressor after the supercharger cannot increase mass flow beyond what the supercharger is displacing. All it can really do is reduce the discharge pressure at the outlet of the supercharger once the turbo compressor starts developing some serious pressure differential of it's own. But you are already fully aware of all that.

The second observation is that a standard turbo blow off valve arranged as you have it, will be very violent in operation. It will either be fully open or fully closed, leading to some fairly dramatic changes in boost pressure upstream of the throttle during driving.

With a slightly open throttle, there will be massive boost pressure developed upstream of the throttle, because with the blowoff valve closed, the supercharger displacement will meter the air volume going into the engine, rather than the throttle position. Think about that for a while...

A sensitive and very progressive air bypass that fully unloads the supercharger at small throttle openings is an absolute necessity with the throttle located upstream of a positive displacement supercharger. A crude turbo blowoff valve is not going to work very well for you at all doing that job.

Thirdly, the additional proposed throttle body air bleed valve ahead of the turbo may also be somewhat problematic and erratic in it's opening and closing, it may even flutter, leading to additional sudden surges in pressure ahead of the throttle body.

The turbo may be very unhappy, and realise that any negative pressure at the turbo compressor inlet is guaranteed draw oil into the compressor housing. Very few turbos these days have positive oil seals on the compressor shafts that can tolerate a vacuum, including the RHB5.

My thoughts are, that this may not feel very nice at all to drive, as the air pressure ahead of the throttle is very likely going to vary wildly with throttle opening, and the operation of the blowoff valve, and throttle body turbo bypass system.

If the turbo and supercharger are already mounted on the engine, some pipework changes should not be too difficult to implement.

I would much prefer to see the turbo compressor ahead of the positive displacement supercharger, with a fully progressive air bypass system arranged around the supercharger as described in the earlier twincharging thread. It is really the simplest system, and it has absolute minimal problems, and the drivability will be superb.

The RHB5 turbo may be a bit on the small side, especially the turbine, but try it first and see. I suspect that the turbo will be up to full boost at surprisingly low engine rpm. That is good, but the total exhaust back pressure may be higher than desirable, and the RHB5 internal wastegate may be too small to be able to control boost right at the very top end. But try it first, and monitor some pressures.

 

[patprimmer]

The logical way is turbo feeding positive displacement blower, feeding intercooler feeding engine.

Whether or not you are past the point of no return, only you can judge, But I think there will be a significant advantage in the simplicity and efficiency of the Turbo, blower, intercooler, engine route.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[wwest]

Off-beat suggestion..

Are you aware of how the e-CVT works in the Prius, FEH, MMH, RXh, and HHL?

May I suggest you convert your project to ONLY a positive displacement SC, a VARIABLE SPEED SC using the e-CVT concept?

Place an open diff'l between the two driving, motive, sources for the SC. Use a belt drive from the engine itself for one diff'l input and an AC synchronous motor (permanent magnet rotor) via a 4:1 gear reduction into the opposite end of the diff'l.

The traditional problem with an engine belt driven SC is that in order to bring on a goodly level of boost at low RPM, more and more of the SC output must be bypassed as the engine RPM rises.

With the e-CVT implementation as described above used the boost level of the SC can readily be modulated, moderated, as the engine RPM rises.

As an extreme example, at 5000 RPM engine speed the SC speed could be brought to ZERO speed via turning the AC synchronous notor at 20,000 RPM but in a direction OPPOSITE the engine drive belt.

Assuming the maximum HP level from the engine belt drive to the SC were 5HP, the AC motor would only need to be capable of 1/4 of that, 1.25HP, a total of 6.25HP maximum to the SC.

Back in the fifties Studebaker, YES, Studebaker, implemented a form of this SC concept using a mechanical CVT, two V-belt pulley's.

 

[Warmington]

Hi,

Thanks for the replies! I mustn't have explained my setup very well as it seems there is some misunderstanding. There IS a fully progressive bypass valve around the SC.

Also, the dump valve merely allows off throttle boost to be recirculated through the TC to keep it spinning, it is not the cahngeover valve for SC to TC.

I have drawn this diagram to help explain my setup:

http://i23.tinypic.com/juv5vb.jpg

Again, thanks for the replies, hopefully you will have another look!

Greg

 

[Warpspeed]

O/k Greg, that is a lot clearer. But I still have to agree with Pat that the turbo compressor should be located ahead of the supercharger.

I expect getting a throttle body to work as a supercharger air bypass will be quite challenging. I tried for a very long time to get something like that to work, but without any success.

There are problems with flow versus throttle plate angle, and how you control it with an actuator. The systems I tested flopped open and flopped shut, or fluttered. I had insurmountable difficulties using this to control supercharger discharge pressure. It was always highly non linear and very unstable.

A far better solution is a poppet valve. Flow area increases in direct proportion to movement, and the control of supercharger discharge pressure is fast, sensitive, and stable. An external turbo wastegate when fitted with an appropriate spring is ideal. This has worked for everyone that has tried it, and is the recommended method.

When the turbo compressor is placed upstream of the supercharger, the supercharger air bypass system can feed air right back to the turbo compressor inlet. That eliminates the need for a turbo blowoff valve, because the air bypass opens fully on closed throttle.

The only problem with this is that when the supercharger bypass is open, a great deal of supercharger rotor noise will be transmitted back to just behind the air filter. These noise can be objectionable at light throttle mid rpm, but a lot depends on the pipework layout and the acoustics.

 

[patprimmer]

Warmington

That is the layout I envisaged.

Like Warpspeed says, it will be very difficult to control the valves and you will rapidly fluctuate between over boost and under boost as the drive from the turbine and the load on the compressor vary at different rates as the wastegate and the bypass valves open and close at different rates.

wwest

As I recall, the Studebakers had a Paxton type centrifugal blower which is a belt driven turbine, just like a belt rather than an exhaust driven turbo. As I recall, these were on the Golden Hawk and the Avanti models and maybe the Hawk GT.

The Eaton mentioned by the OP is a positive displacement supercharger. These deliver a very flat boost level with it tending to drop off a little at high rpm. Without bothering to check, I am about 99% sure that an Eaton is a high helix Roots type blower, but they just might be a screw type. Either way, positive displacement with almost linear output to rpm up to the point where their efficiency starts to drop off.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Warmington]

Hi,

Thanks for the replies.

I am still not completely sure what you mean by having the TC compressor ahead of the SC. It actually is ahead of the supercharger in my application. Although the TC wastgate and the butterfly throttle valve will open at different rates, they are both governed by actual manifold pressure, therfore linked fluidly...

I have seen systems with the same configuration work very well with just a flap valve in place of my actuated butterfly valve.

I understand using a poppet valve wastegate mechanism upstream of the SC in a TC-SC configuration to control max pressure, but I dont quite understand what you mean by fluctuations in this configuration.

Of course the blow-off valve and the TC wastegate and the butterfly valve will have their own open/close speeds and characteristics, but this will only result in fluctuations before the dumpvalve/recirculating valve, i.e. none of the fluctuations should ever reach the manifold as it will all be recycled and kept within the SC bypass system off throttle.

The problem of over boost (ie the SC keeping the boost at a constant level, and the TC adding some) will just be fed out of the dump valve. If there is under boost (ie the engine cannot drag enough air past the recirculating system) should be overcome by the fact that there will always be *some* positve pressure after the intercooler as the dump valve is sized smaller than the actual main hose diameter.

Could you possibly help with an actual illustration, e.g. "when the foot is off the throtle, a happens at this valve and B happens at this valve, and manifold sees this pressure"?

Thanks
greg

 

[Warpspeed]

Both the topic line of this thread, and your diagram shows the supercharger as being ahead of the turbo compressor.

Yet the first paragraph of your above post suggests exactly the opposite. I have now become so confused by all this I can be of no further help. Sorry.

 

[Warmington]

Hi,

I think I understand. If the SC is AHEAD of the TC it is upstream, i.e. it the airflow must pass through it to reach the TC initially. Yes, it is ahead of the TC.

I apologise for clouding the water over that point!! I was assuming that ahead, meant that it was more towards the inlet manifold!

So all the points above remain warpspeed. I have read and digeted what you had siad and find it very understandable, you have obviously had experience of this setup and dont think it will be very good.

I will definitely change to the other configuration or even change the butterfly valve to the poppet valve if needed, but at this point, I still cannot envisage a case when there would be an overboost situation, flutter, or underboost.

As the dump valve bypass is also manifold vacuum actuated, there will still be boost until the Dump valve opens, and even then the boost is still circulating. This is exactly how the TC would work alone (plus the lag to spool it up again).

Is it this dump valve/recirculating valve then you suggest to be swapped for a poppet valve? In order to maintain more linear control of the boost reaching the engine, this valve would have to be vaccum activated, and I suspect would still actually behave like the dump valve anyway.

Warpspeed, could you please illustrate a case where there may be an unwanted boost level, explaining what the pressures would be at each point please? I really would find that beneficial to my understanding of the problem that I will encounter.

Many thanks
Greg

 

[patprimmer]

This has already been done to death.

The correct layout as stated several times in this thread, and several times in the thread referenced in the OP is

From atmosphere to cylinder head

Air cleaner.
Turbo charger.
Supercharger.
Intercooler.
Cylinder head.

The throttle body can be inserted in several places.


Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Warmington]

Just one other point, my butterfly valve 'bypass' does not have to be realted to total maximum boost at all. That is not what it is achieving. It is merely allowing the TC to suck air around the SC when it needs to. It opens at the SC max boost level, not the total combined maximum, so therefore does not control max boost. That is still controlled by the TC wastegate actuator, which will still open when it is overcome by manifold pressure like ussual.

I cannot see how having a maximum pressure of 7psi before the TC at all times (which is really the ONLY difference my system now has from a fluid standpoint over the original TC only configuration) can cause any problems. The dump valve existed before in this configuration as on alot of cars.

Greg

 

[Warmington]

Pat, you have actually not offered anything useful in the way of advice!! Could you possibly answer the question above if you understand the flow through my system?

I am well aware of the system suggested, and have viewed such systems as on the lancia S4 and nissan micra Superturbo. I understand that it is the accepted method. I understand why it works. I DONT undertsand where you feel the problem will arise in my system.

Care to illustrate a scenario where there may be a problem?

Thanks
Greg

 

[patprimmer]

As said several times before, but obviously not clearly enough.

The supercharger will provide enough air until the turbo starts to pull a vacuum in the airduct between the supercharger and the turbocharger.

When this happens you intend to rectify this with a butterfly valve that will open and allow additional airflow into the airduct.

When this butterfly valve is open, the supercharger will still consume power to drive it, but it will serve no useful purpose as the air pressure in the airduct will be at atmospheric or less, or worse still it will be at atmospheric pressure but at a higher temperature due to heat transfer, shear heat and friction while in the blower.

The main problem I envisage is there will be positive pressure in the airduct, which will decline until the valve opens. If the valve opens a fraction to soon, boost will be lost giving a flat spot in performance. If the valve is opened to late, a vacuum will develop then an inrush of air.

This might introduce an overreaction in the butterfly which will open then close then open then close like a trailer swaying.

If you are to far along the path, continue and try it, however if it is still an easy option to change follow the tried and true and efficient layout.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Warmington]

Hi Pat,

I completely understand what you are saying. I did think of this scenario. My *solution* is manual adjustment of the actuator spring pressure on the butterfly valve until the changeover point is set very slightly ahead of the SC max boost pressure. This will ,as you say, result in a flat spot as the TC momentarily sucks a vacuum between itself and the TC. The closer the point can be brought to the point where SC boost and TC boost cross, the less this effect will be. I will just have to keep adusting the spring pressure.

I will start off with it at quite a large amount over SC pressure (say 2psi) and come back until the boost curve is as smooth as I can get it without losing SC boost through the valve. I understand that the heat of the incoming air will have an effect on actual boost pressure, and effectively continue to move the ideal SC changeover point around, but I am sure that accepting a *very* small flat spot in the boost pickup will effectively throw a blanket over that area on the curve.

The other problem you mention, with respect to the hot air from the SC, is surely a consideration in both setups (TC into SC (without SC clutch), or SC into TC) so may be ignored for the point of comparison, although, if inlet temps are undesireable additional intercooling can be added, or even water spray.

As you say, time will tell. I have not seen this setup on another engine yet (probably for a good reason!) but at the moment, I am convinced I can get it to work. I my try both methods, this one first and then the TC into SC method and compare the differences.

I dont like the idea of having to control max boost ahead of the TC into SC setup by venting the boost to atmos through a pop of valve, even if it is re-routed back to the air filter. The pressure being wasted has been compressed at the expense of the engine power, cooled, and then thrown away.

In the case of the TC into SC, then a bypass around the SC at higher revs, what is your preferred max pressure control mechanism apart from a poppet valve?

Thanks
Greg

 

[Warmington]

One other point I forgot to address from your post pat:

When the bypass is open, and the TC is supplying more than the SC max boost, the SC will consume almost NO power due to the fact that it has a vaccum ahead of it i.e. it will just be freewheeling. There is no restriction, therefore no resistance, and no forces to reolve. Its power consumption will merely be a function of its inertia at this point, although it will still heat the air al little due to the temperature of the SC itself.

Greg

 

[patprimmer]

I think you might be underestimating the power absorbed by a roots blower turning at neutral boost

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[wwest]

Also keep in mind that modern day use, adaptation, of the Atkinson Cycle has now made it obvious to all that the exhaust energy used to run the turbo was never "free" after all.

You can make much better use if that energy by adapting your system to the atkinson cycle, late/delayed intake valve timing and ~13:1 physical compression ratio. In your case actually the later derivative thereof, the Miller Cycle.

And...

Am I wrong...?

It has long been my understanding the in order for a positive displacement SC to provide boost at low engine RPMs they must PUMP more air than the engine intake/RPM can "handle". Then, absent somehow throttling the SC intake itself, as the engine RPM rises higher and higher a portion of the sc output must be bypassed to prevent overboost.

The latter, absent a method for moderating the SC boost, being the major shortcoming of an SC over a TC.

 

[Warmington]

Pat,

Please feel free to enlighten me...

As far as I am aware, assuming a constant rotational speed for the SC, if the pressure accross the SC is negative, it does no compression (positive displacement), the only energy needed is that which is consumed in moving the air from the inlet to the outlet plus frictional losses. This is negligible compared to power drain if there is a positive pressure differential.

Do you see it differently to that?

wwest,

The rootes type super charger displaces a certain amount of air per revolution. Its rpm ratio to engine rpm is fixed by pulley size. Similarly an engine displaces a certain amount of air per revolution. The ratio of SC rpm vs engine rpm determines a pressure ratio that is contant throughout the entire rpm range. There is no need to moderate boost using only a SC as it does not rise past a set level. The bypass is used to allow the engine to run in vaccum off throttle.

Greg

 

[wwest]

First..

Then why does almost positive displacement EVERY SC installation use some sort of boost bypass as the engine RPM rises.

Is it maybe becuase the SC has NO inlet restriction but the consumption/flow into the engine is "moderated" by intake valving? Unline a turbine, or the SC, the airflow consumption of an ICE is not constant, it occurs in "pulses" as each cylinder comes up for "air".