2014-2020 Formula One 1.6l V6 turbo engine formula

[WhiteBlue]

Yes Wuzak, the waste gate use under standard operating condition would defeat the objective of the MGU-H. Hence my mulling over possible failure modes where it could be useful.

[autogyro]

In my experience it is very easy to stall a turbocharger.
Controlling the load/output on the H/MG will be very difficult.

[xpensive]

If you use a wastegate there will be no extra torque on the turbine, and thus no recovered power.

I think the conversion of exhaust turbine energy into electrical power will be one of the main areas of development. Particularly since the amount of energy recovered is unlimited by the regulations. The maximum power is defined by whatthe MGUK can use (ie 120kW), but it is unlikely that anybody will ever approach that.

Don't forget there's a limit on the battery charge, 4000 kWs, and discharge per lap, 2000 kWs.

And yes, I've had it with those "MJ", this is electrical energy and not some pellet-fueled heater.

[WhiteBlue]

If you use a wastegate there will be no extra torque on the turbine, and thus no recovered power. I think the conversion of exhaust turbine energy into electrical power will be one of the main areas of development. Particularly since the amount of energy recovered is unlimited by the regulations. The maximum power is defined by whatthe MGUK can use (ie 120kW), but it is unlikely that anybody will ever approach that.

Don't forget there's a limit on the battery charge, 4000 kWs, and discharge per lap, 2000 kWs.

We can safely assume that the bulk of the power from the MGU-H will never touch the battries. The energy management will always prefer the optimized route from the MGU-H to the MGU-K in order to avoid conversion losses. Only the small amount needed to top off the battery charge will be routed to storage.

[xpensive]

I'm not so sure if powering the MGU-K directly will make much sense, it would be xtremely intermittent, almost unpredictable.

[WhiteBlue]

In my experience it is very easy to stall a turbocharger. Controlling the load/output on the H/MG will be very difficult.

These electric servo units are incredibly fast in terms of control. They are much faster than any transient an ICE can produce under normal operating conditions. In machine tools their linear version is used to machine ultra high precision surfaces without any mechanical guidance. The operating head is cutting the geometry purely controlled by three dimensional motion control using three linear servo motors. These things use exactly the same technology as the MGUs in F1, brush less, permanent magnet, synchronous, water cooled AC servo drives.http://en.wikipedia.org/wiki/Servo_drive

Have you ever looked at the dynamics of digital controllers that provide the torque and speed control for these drives? http://publications.lib.chalmers.se/rec ... t/2386.pdf On page twenty it says:

A simple and efficient way to achieve fast and accurate current control in inverter-fed PMSMs is to utilize synchronous-frame PI controllers[38]. The mechanical dynamics can be neglected as they are typically much slower than the current dynamics.

I know no other mechatronic application that comes even close in terms of controllability and responsiveness. The finest hydraulic servo valves in the aerospace industry don't come even close to it. I know one or two things about electric drives and I would be very surprised if torque or rpm control of the MGU-H would be a problem at all, provided that the implementation is done with the required professionalism and bug free software. There could be teething problems initially when the engineers who have never used such technology get to grips with it. This has happened with KERS as well. But given the time for a proper learning curve this electric technology should be extremely accurate and effective.

[WhiteBlue]

I'm not so sure if powering the MGU-K directly will make much sense, it would be xtremely intermittent, almost unpredictable.

How do you arrive at that conclusion? Have you ever used two servo drives in a multiple axis application? This is standard design for machine tools and other type of machinery where you use one axis to regenerate electricity from braking and feed it to another axis that needs to accelerate. The drives use a DC type of intermediary circuit which typically uses some capacity (supercaps) to smooth out transients. The F1 application in fact is a very simple one as the MGU-K wants to be driven with the same torque profile as the ICE. The ICE in rising torque demand produces more exhaust gas which in turn stimulates the MGU-H to produce more electricity to keep the turbo charger balanced. The produced electricity raises the torque at the MGU-K which is exactly what the driver demands with his throttle pedal. The same applies in case of reducing the torque demand. The ICE and the MGU-K torque must be going in the same direction and this makes the control strategy very simple. There is almost no need for energy storage in the intermediate circuit in this application as long as you stay reasonably above the point where the turbo needs spooling up.

[wuzak]

I'm not so sure if powering the MGU-K directly will make much sense, it would be xtremely intermittent, almost unpredictable.

On the contrary, I should think that the MGUH's power generation will be entirely predictable and quite smooth.

The compressor's load will be predictable. The exhaust energy will be consistent for a given set of conditions, which will give a predictable turbine power.

The exhaust pulse causing the intermittent torque generation will be mitigated by the turbine's mass moment of inertia. The spooling issue that comes with that mass moment of inertia is overcome by using the MGUH to help spool the turbine.

The MGUH may help to slow the turbine down when the driver shuts off the throttle and boost is no longer required.

[wuzak]

This was Allison's turbocompound in 1944.



The 2014 engines are turbocompounds, but instead of a mechanical or hydraulic link, they use electricity.

The Wright turbocompound had fluid couplings to transmit the power from its three turbines back to the crankshaft. The Allison had its turbine connected directly to the crankshaft.

There were no issues with intermittent power, and teh Wright's fluid couplings were to compensate for differences in turbine and crankshaft speeds.

[WhiteBlue]

There were no issues with intermittent power, and teh Wright's fluid couplings were to compensate for differences in turbine and crankshaft speeds.

I expect the 2014 F1 electric compounding to use the intermediate DC link with a bit of supercap capacity as means of smoothing any dynamic issues. So that would be comparable to the hydraulic coupling in the Wright engine. Only that the electric compounding design would be far more elegant and versatile regarding the spool up option that you do not need in an aircraft.

[ringo]

I'm not so sure if powering the MGU-K directly will make much sense, it would be xtremely intermittent, almost unpredictable.

How do you arrive at that conclusion? Have you ever used two servo drives in a multiple axis application? This is standard design for machine tools and other type of machinery where you use one axis to regenerate electricity from braking and feed it to another axis that needs to accelerate. The drives use a DC type of intermediary circuit which typically uses some capacity (supercaps) to smooth out transients. The F1 application in fact is a very simple one as the MGU-K wants to be driven with the same torque profile as the ICE. The ICE in rising torque demand produces more exhaust gas which in turn stimulates the MGU-H to produce more electricity to keep the turbo charger balanced. The produced electricity raises the torque at the MGU-K which is exactly what the driver demands with his throttle pedal. The same applies in case of reducing the torque demand. The ICE and the MGU-K torque must be going in the same direction and this makes the control strategy very simple. There is almost no need for energy storage in the intermediate circuit in this application as long as you stay reasonably above the point where the turbo needs spooling up.

A servo drive is a DC motor, A generator must be alternating current.
But mostly, a motor/ generator is a load, it is not a relief valve, it cannot control exhaust mass flow through the turbine.
You also put yourself in a compromising position by the flow profile of the turbine and compressor if you cannot have any control of flow through it.
You will have only one compressor flow/ turbine flow profile for the turbine, but all you will able to do is to retard it by loading the turbine with the MGUH, and increase back pressure and heat soak by doing that. When you have a waste gate you will be able to change the profile of the turbine flow relative to compressor flow for various situations.
This is an imperfect world, you never know what situations may arise where you will need to drop the back pressure off the engine or change the turbine performance on a dime.
A load cannot do that, the most it will do is retard and stress things.
A waste gate is the simpler and cheapest form of auxiliary control. It is controlling flow, which is much different to torque therefore the likelyhood of stalling the turbine is much reduced. A generator runs a high risk of stalling. The waste gate is a fail safe and i wont be surprised if all the engines have wastegates on their exhaust systems. It's also another means of modifying engine breaking as well when the driver comes off throttle.

[ringo]

I'm not so sure if powering the MGU-K directly will make much sense, it would be xtremely intermittent, almost unpredictable.

On the contrary, I should think that the MGUH's power generation will be entirely predictable and quite smooth.

Depends on the storage, ie the state of the batteries, or the state of the loads on the MGUH. Let's say the MGUH powers the MGUK and the car is in a corner with wheel spin and engine speeed fluttering?

The compressor's load will be predictable. The exhaust energy will be consistent for a given set of conditions, which will give a predictable turbine power.

It's not safe to always assume this or depend on this.

The exhaust pulse causing the intermittent torque generation will be mitigated by the turbine's mass moment of inertia. The spooling issue that comes with that mass moment of inertia is overcome by using the MGUH to help spool the turbine.

The MGUH may help to slow the turbine down when the driver shuts off the throttle and boost is no longer required.

Your are increasing back pressure right there. The exhuast wants to do one thing, but the torque on the blades is doing another thing. This will more than likely have some effects on the compressor and also what is happening in the combustion chamber.

[ringo]

This was Allison's turbocompound in 1944.

http://www.enginehistory.org/Allison/V1710TC.jpg

The 2014 engines are turbocompounds, but instead of a mechanical or hydraulic link, they use electricity.

The Wright turbocompound had fluid couplings to transmit the power from its three turbines back to the crankshaft. The Allison had its turbine connected directly to the crankshaft.

There were no issues with intermittent power, and teh Wright's fluid couplings were to compensate for differences in turbine and crankshaft speeds.

This is a different animal, and i'm sure the performance, flexibility and response doesn't meet what is required on a modern road car muchless an F1 car. It's quite obvious the disregard for precision with that setup.

[Holm86]

So no one thinks there could be any advantage in using boost pressure for aero trickery??

[xpensive]

I'm not so sure if powering the MGU-K directly will make much sense, it would be xtremely intermittent, almost unpredictable.

How do you arrive at that conclusion?
...

Because it will work just like a wastegate, cut in only when the compressor has reached the targetpressure, before then, nothing.