33 replies to this topic

[Rosssco]

 

Would a Rotrex SC not be better suited to compound charging rather than the Eaton?

 

Yes, and at that point you may as well throw away the turbo and just fit a massive Rotrex SC as you've already got past the issue of inertia. 

 

 

Yep, benefits of a positive displacement SC is the ability of produce reasonable boost levels at low - med RPM.

 

Old school torque fill before electric motors became all the vogue..

[slindborg]

TBH compound changing in the traditional sense isn't really sensible anymore (imho), "just" fill in the torque hole with electric motors either on the drive output or to help the turbocharger (ala F1/Lemans)

[Scuffers]

 

 

Would a Rotrex SC not be better suited to compound charging rather than the Eaton?

 

Yes, and at that point you may as well throw away the turbo and just fit a massive Rotrex SC as you've already got past the issue of inertia. 

 

 

Yep, benefits of a positive displacement SC is the ability of produce reasonable boost levels at low - med RPM.

 

Old school torque fill before electric motors became all the vogue..

 

if you look at the detail, nobody in their right minds would use a mechanically linked centrifugal supercharger, to compound one makes zero sense.

 

TBH compound changing in the traditional sense isn't really sensible anymore (imho), "just" fill in the torque hole with electric motors either on the drive output or to help the turbocharger (ala F1/Lemans)

Ultimately, this is the way it's going, but as there are no (cheap) off the shelf systems to do this, it;s a bit of a moot point.

 

there's also the other reason you compound charge, when you're looking for a PR over where a single turbo can efficiently manage. 

[Rosssco]

Never understood why a manufacturer (especially one like VAG with their twin charge unit) have not designed in integral turbo / supercharger unit which in its most basic sense is a turbo unit with an extended compressor shaft driven off the accessory belt..

 

Or a bit like a Rotrex unit , but with a exhaust turbine side. The 'supercharger' aspect would feature a geared clutched attachment that would engage in low RPM situations to provide compressor rotation, which would then de-clutch when sufficient exhaust gas pressure was there to drive the compressor via the turbine.. I guess there would be some technical trade-offs in general design, packaging and heat management, but modern engine management systems would appear to be (to the untrained eye) perfectly capable to this..

 

I guess its a potential spin off (no pun) benefit of this technology used in the likes of F1 that will eventually filter down..

[Scuffers]

like I implied before, an engine driven centrifugal compressor is a stupid idea on an engine that's designed to have any significant rev-range.

 

the only way that could work is to have an infinitely variable drive so that the compressor shaft speed was totally un-related to the engine crank speed, problem is that would be complex and expensive to make and keep efficient, which is why nobody has done it.

 

The Rotex solution (friction drive) is a bodge, they are very limited on shaft power ratings, and horribly inefficient in terms of a transmission system, quite apart from them being hopelessly unreliable.

 

As said, an electric compressor is the ultimate, BUT these are only just becoming feasible with the advances in small high power motors, however, they still need a hell of a lot of power to drive them.

 

what your missing here is that the shaft speed of the compressor needs to be load related, not engine speed related, hence why a turbo actually works really well, the shaft speed is a factor of load, not engine speed.

 

the current thinking is just to electrically 'boost' turbo's to get the response times down, then, taking this a step further, using the motor to regulate the shaft speed when the turbine is powering it (instead of using the wastegate), syphoning off the additional power to either store for boosting later or feeding back into the powertrain (be that powering engine ancillaries or the main powertrain itself).

 

in time, once the OEM's are doing this, the parts will become much cheaper.

 

 

[fezzasus]

But the parts may not become any more usable even if they are cheaper. The industry is also moving to very tight packaging of the exhaust manifold and aftertreatment to avoid getting penalised by warm up times in the NEDC cycle. There might be plenty of electronically boosted turbos out there, but if you can't make them fit anything then they're not that useful.

[slindborg]

what about "simple" small turbo and big turbo...

[Scuffers]

what about "simple" small turbo and big turbo...

a good solution, but notoriously hard to implement well.

 

you need very good components and an ECU with the ability to control it all (not common).

 

using a PD SC is much easier, especially when it can be done with off the shelf parts, and the control system are easier (though still complex).

[Rosssco]

what about "simple" small turbo and big turbo...

  Looks "complex"... http://www.cobaltss....-anyone-305631/

[Rosssco]

like I implied before, an engine driven centrifugal compressor is a stupid idea on an engine that's designed to have any significant rev-range.

 

the only way that could work is to have an infinitely variable drive so that the compressor shaft speed was totally un-related to the engine crank speed, problem is that would be complex and expensive to make and keep efficient, which is why nobody has done it.

 

The Rotex solution (friction drive) is a bodge, they are very limited on shaft power ratings, and horribly inefficient in terms of a transmission system, quite apart from them being hopelessly unreliable.

 

As said, an electric compressor is the ultimate, BUT these are only just becoming feasible with the advances in small high power motors, however, they still need a hell of a lot of power to drive them.

 

what your missing here is that the shaft speed of the compressor needs to be load related, not engine speed related, hence why a turbo actually works really well, the shaft speed is a factor of load, not engine speed.

 

the current thinking is just to electrically 'boost' turbo's to get the response times down, then, taking this a step further, using the motor to regulate the shaft speed when the turbine is powering it (instead of using the wastegate), syphoning off the additional power to either store for boosting later or feeding back into the powertrain (be that powering engine ancillaries or the main powertrain itself).

 

in time, once the OEM's are doing this, the parts will become much cheaper.

 

 

 

I guess its easy to underestimate the inefficiencies of the some of the potential components (such as the gearing drive required to convert crank speed to compressor shaft speed), and of course cost and maintenance factors of mass-produced engines quickly make such a system appear more unrealistic.

 

And I dunno what you mean about feasibility of electric drives - you can get them off fleabay for ~£25...  

[FLD]

How about VNT or VGT turbo's?  I ran these some time ago with great results.  I had 3psi of boost at tickover running right up to 24 psi of boost at 3k.  Don't get me wrong, there are issues with these that need to be overcome.  For example, the garret system operates over quite a low rev range so you need a wastegate as well.  This means there's lots of control systems to get just right (ball ache!).  Then there's the EGT to watch, teh vanes tend to melt unless you get the ceramic version which costs.  Get it set up right and it's pretty good though.

[Scuffers]

How about VNT or VGT turbo's?  I ran these some time ago with great results.  I had 3psi of boost at tickover running right up to 24 psi of boost at 3k.  Don't get me wrong, there are issues with these that need to be overcome.  For example, the garret system operates over quite a low rev range so you need a wastegate as well.  This means there's lots of control systems to get just right (ball ache!).  Then there's the EGT to watch, teh vanes tend to melt unless you get the ceramic version which costs.  Get it set up right and it's pretty good though.

good as they are, nobody has managed to get one to survive hard use, Borg waner have the current best (used on the 911 turbo) but still have issues.

 

Much easier on Dismals with lower EGT's.

 

the Saab setup is actually pretty elegant, but it's very specific, note the big turbo is not very big.

[Exmantaa]

 

what about "simple" small turbo and big turbo...

  Looks "complex"... http://www.cobaltss....-anyone-305631/

 

 

Sorry.. I posted that over there... (It's a modified Diesel twin turbo system if I remember correct)

 

@ Nev; like Scuffers said, the bypass opening is plenty big enough for you. Most of the air is going through the spinning rotors and only the excess air is routed back to before the SC.

(If it works OK for in a normal SC @ ambient pressure, it will also work the same if you pre-compress the air entering the SC unit with a turbo. )

 

 

[FLD]

 

How about VNT or VGT turbo's?  I ran these some time ago with great results.  I had 3psi of boost at tickover running right up to 24 psi of boost at 3k.  Don't get me wrong, there are issues with these that need to be overcome.  For example, the garret system operates over quite a low rev range so you need a wastegate as well.  This means there's lots of control systems to get just right (ball ache!).  Then there's the EGT to watch, teh vanes tend to melt unless you get the ceramic version which costs.  Get it set up right and it's pretty good though.

good as they are, nobody has managed to get one to survive hard use, Borg waner have the current best (used on the 911 turbo) but still have issues.

 

Much easier on Dismals with lower EGT's.

 

the Saab setup is actually pretty elegant, but it's very specific, note the big turbo is not very big.

 

 

I ran a pair on a cosworth GB.  Worked fine but it ran a thermocouple in the exhaust for egt.  Knowing how I drive I doubt it recieved 'hard use' though.  I believe the Holsett sliding ring system has had greater interest in the US.  It's just what I've heard but might be worth looking into (nev, )