TERS : Thermal Energy Recovery System

[Tzk]

Basically the cited regulations treat the engine as black box. As long as the torque output of the power unit matches the request from the driver (accelerator pedal). I don‘t get why it‘s so hard to believe that the accelerator pedal is not directly linked to the throttle bodies.

I didn‘t know about the exception under full load, thanks for pointing that one out hollus. This explains why some cars get slower at the end of the straights, when the energy recovery kicks in (brake light flashes), even before the drivers break.

[gruntguru]

The (power unit torque demand 5.5 -1 -2 -3 -4 and 5.6.1 I posted are all from 2019 regulations and they clearly show that a 50% driver throttle pedal demand and a throttle map giving 100% engine throttle is not allowed.

Absolutely not. The regs only talk about "torque", "torque demand" and "driven wheels". Nowhere do they mention the throttles of the ICE or the torque output of the ICE - it is always output of the PU which is a box containing the ICE, the MGUH, the MGUK, the ES and the electronic control system.

In summary the rules specify the relationship between the "throttle" pedal and the output from the PU.

[bergie88]

Very interesting dissertation I've only just started to read that I thought I'd share with everyone. This thesis was written in 2018 with the help and support of Ferrari. I'm only about 20 pages in after a quick scan through, but it seems very well written and offers a glimpse of what is allowed and not allowed.

If the driver is not requesting full power, the control system must deliver the amount of power requested by the driver, and the energy management system can only decide how to split it between the ICE and the MGU-K. Conversely, if the driver is requesting maximum power, the regulations allow the energy management system to decide the amount of propulsive power to provide and the split between the actuators. The degree of freedom offered by the thrust controller allows to trade off lap time against energy consumption and cope with the finite size of the battery and the fuel consumption limit of 105 kg per race.

https://www.research-collection.ethz.ch ... sAllowed=y

Promises to be a fascinating read. It will be interesting to compare it with the Limebeer paper. Limebeer modelled a limited range of the potential power options, on a quick skim this paper looks as though it deals with more.

On a first glance it seems that prof. Limebeer's papers are focusing more on the theoritical side of optimal control, while this one seems to be more on the practical implementation of it (off course with a strong theoretical background).

During my MSc I also worked on optimal control of hybrid vehicles and I was already wondering why I never saw terms like dynamic programming and ECMS in publications. Now there are, and it seems that Ferrari found the correct partner as prof. Guzella's research group at ETH is really the benchmark on this subject.

[henry]

As well as contributing positively to the torque from the ICE to match demand, the MGU-K can contribute negatively. So at 70% torque demand the ICE can make 90% and the K extract 20% to provide the target torque whilst charging the ES. (@tommy cookers pointed out above that this can lift the ICE to a higher efficiency output so depending on the relative efficiencies of the component parts may have little negativeor even positive effect on fuel consumption)

The MGU-K contribution is monitored by the sensor at the entry to the Control Unit for the K. The restrictions are +/- max 120kW and max 200Nm. These are conditioned by:

“A fixed efficiency correction of 0.95 will be used to monitor the maximum MGU-K power.” And “
The maximum torque of the MGU-K may not exceed 200Nm. The torque will be referenced to the crankshaft speed and the fixed efficiency correction defined in Article 5.2.2 will be used to monitor the maximum MGU-K torque.”

I struggle to get my head around the 0.95 factor for the generation side of things.

Does it mean that the monitored limits are +/- 126kW and 210Nm or +126, -114 and +210, -190 ?

Obviously this makes little difference to us as onlookers but the latter is a significant asymmetry for the engineers.

[saviour stivala]

The (power unit torque demand 5.5 -1 -2 -3 -4 and 5.6.1 I posted are all from 2019 regulations and they clearly show that a 50% driver throttle pedal demand and a throttle map giving 100% engine throttle is not allowed.

Absolutely not. The regs only talk about "torque", "torque demand" and "driven wheels". Nowhere do they mention the throttles of the ICE or the torque output of the ICE - it is always output of the PU which is a box containing the ICE, the MGUH, the MGUK, the ES and the electronic control system.

In summary the rules specify the relationship between the "throttle" pedal and the output from the PU.

"Torque" "Torque demand" and "driven wheels" What gets to the driven wheels is produced by the 'PU'/ICE-K-H-ES and CE, the product of which is controlled by the ICE throttle/s position (movement 0 - 100), of which in turn is controlled by the driver accelerator pedal demand (movement 0 - 100). 5.5.3:- At any given engine speed the driver torque demand map must be 'monotonically' increasing for an increase in accelerator pedal position.

[Tzk]

Simply no. You even quoted that the regulation states that the driver TORQUE demand map (note: this is not throttle position!) must increase monotonically with the accelerator pedal position.

Note that the torque delivered by the whole power unit isn‘t linked to ice only power (and thus not linked to throttle position). It‘s a hybrid powerunit so it can decide by itself how it generates the demanded torque/power through mgu-k, h and ice.

[gruntguru]

The (power unit torque demand 5.5 -1 -2 -3 -4 and 5.6.1 I posted are all from 2019 regulations and they clearly show that a 50% driver throttle pedal demand and a throttle map giving 100% engine throttle is not allowed.

Absolutely not. The regs only talk about "torque", "torque demand" and "driven wheels". Nowhere do they mention the throttles of the ICE or the torque output of the ICE - it is always output of the PU which is a box containing the ICE, the MGUH, the MGUK, the ES and the electronic control system.

In summary the rules specify the relationship between the "throttle" pedal and the output from the PU.

"Torque" "Torque demand" and "driven wheels" What gets to the driven wheels is produced by the 'PU'/ICE-K-H-ES and CE, the product of which is controlled by the ICE throttle/s position (movement 0 - 100), of which in turn is controlled by the driver accelerator pedal demand (movement 0 - 100). 5.5.3:- At any given engine speed the driver torque demand map must be 'monotonically' increasing for an increase in accelerator pedal position.

Not sure what you are trying to say there. It all sounds correct to me although if you are implying that the accelerator pedal controls the engine throttles you are wrong.
1. The accelerator pedal controls an electrical signal to the ECU to indicate the driver torque demand.
2. The ECU decides what combination of engine power and electrical power (+ve or -ve) will be used to meet that torque demand.
3. It decides the best way to achieve this combination given the current priorities set by the mode switches on the steering wheel.
4. It controls the numerous outputs at its disposal to optimise each process, for example:
- the desired ICE output will require a particular manifold pressure - the throttle position will be controlled and the MGUH motored or braked to achieve the appropriate compressor speed.
- the other vital ICE parameters like injection quantity, injection timing, injection multiple pulse parameters, ignition timing, cylinder skipping and variable intake length will all be controlled. Note that many of these will have a dramatic effect on ICE output regardless of engine throttle setting.
- The result of all these machinations is a "torque" at the gearbox input shaft. This torque is the value referred to in the rules - (must be controlled by the accelerator pedal, monotonically increasing etc etc).

[gruntguru]

Since writing the post above it has occurred to me that it doesn't deal with a key issue that may be causing some people to miss the point - here goes.

The main reason for the rule we are discussing, is to avoid electronic control of torque as a driver aid (eg traction control). With the current PU technology, throttle position (not accelerator pedal position) alone does not dictate output so even if the rules stated the engine throttles must be mechanically connected to the throttle pedal, the ECU could still control the torque eg:

- varying the MGUK motoring/braking amount
- varying the fuel delivery (these engines can maintain near peak efficiency over a a wide range of AFR)
- cylinder skipping
- varying timing or any of the other parameters although at a cost to thermal efficiency

So it becomes obvious that mandating a relationship between accelerator position and throttle position (or even mandating a relationship between accelerator position and ICE output) achieves nothing.

[saviour stivala]

Simply no. You even quoted that the regulation states that the driver TORQUE demand map (note: this is not throttle position!) must increase monotonically with the accelerator pedal position.

Note that the torque delivered by the whole power unit isn‘t linked to ice only power (and thus not linked to throttle position). It‘s a hybrid powerunit so it can decide by itself how it generates the demanded torque/power through mgu-k, h and ice.

Of course the torque delivered by the whole power unit isn't linked to ICE only. I never said so. But yes the torque delivered is linked to throttle position demand.
“monotonically:- nondegreesing function consistently increase or decrease but do not oscillate in relative value. Each member of a mouton increase sequence is greater than the proceding member. Each member of a monotone decreasing sequence is less than or equal to the preceding member”.

[saviour stivala]

Since writing the post above it has occurred to me that it doesn't deal with a key issue that may be causing some people to miss the point - here goes.

The main reason for the rule we are discussing, is to avoid electronic control of torque as a driver aid (eg traction control). With the current PU technology, throttle position (not accelerator pedal position) alone does not dictate output so even if the rules stated the engine throttles must be mechanically connected to the throttle pedal, the ECU could still control the torque eg:

- varying the MGUK motoring/braking amount
- varying the fuel delivery (these engines can maintain near peak efficiency over a a wide range of AFR)
- cylinder skipping
- varying timing or any of the other parameters although at a cost to thermal efficiency

So it becomes obvious that mandating a relationship between accelerator position and throttle position (or even mandating a relationship between accelerator position and ICE output) achieves nothing.

100% agreed “that the main reason for the rules we are discussing, is to avoid electronic control of the torque as a driver aid”. (eg traction control). Is but just one in a big can of worms if it wasn’t so. I do not blame you, because I know you couldn’t have read it, but I wrote and posted fully to that effect. But, with the current PU technology, throttle position(ICE throttle's position) which in turn depends on driver demand through the accelerator pedal position dictates the level of torque output of the PU (A mix of electric and ICE power resulting from the combined function of all 6 power unit elements). at a level in accordance of the deployment/harvesting map level selected by the driver from on the steering wheel. Also 100% agree that fuel injected/ignition timing and cylinder cutting plays an important part. And finally thanks for a very interesting discussion.

[Tommy Cookers]

...
The MGU-K contribution is monitored by the sensor at the entry to the Control Unit for the K. The restrictions are +/- max 120kW and max 200Nm. These are conditioned by:
“A fixed efficiency correction of 0.95 will be used to monitor the maximum MGU-K power.” And “
The maximum torque of the MGU-K may not exceed 200Nm. The torque will be referenced to the crankshaft speed and the fixed efficiency correction defined in Article 5.2.2 will be used to monitor the maximum MGU-K torque.”

I struggle to get my head around the 0.95 factor for the generation side of things.
Does it mean that the monitored limits are +/- 126kW and 210Nm or +126, -114 and +210, -190 ?

aren't the limits ? ....
126 kW for outbound electrical energy (ie 126 measured electric kW becomes 120 notional mechanical kW) and
120 kW for inbound electrical energy (ie 126 notional mechanical kW becomes 120 measured electric kW)

they don't seem to want to get very close to the 200 Nm torque limiting region anyway
the best EM efficiency would likely result from keeping at quite high rpm and correspondingly quite low current

[henry]

...
The MGU-K contribution is monitored by the sensor at the entry to the Control Unit for the K. The restrictions are +/- max 120kW and max 200Nm. These are conditioned by:
“A fixed efficiency correction of 0.95 will be used to monitor the maximum MGU-K power.” And “
The maximum torque of the MGU-K may not exceed 200Nm. The torque will be referenced to the crankshaft speed and the fixed efficiency correction defined in Article 5.2.2 will be used to monitor the maximum MGU-K torque.”

I struggle to get my head around the 0.95 factor for the generation side of things.
Does it mean that the monitored limits are +/- 126kW and 210Nm or +126, -114 and +210, -190 ?

aren't the limits ? ....
126 kW for outbound electrical energy (ie 126 measured electric kW becomes 120 notional mechanical kW) and
120 kW for inbound electrical energy (ie 126 notional mechanical kW becomes 120 measured electric kW)

they don't seem to want to get very close to the 200 Nm torque limiting region anyway
the best EM efficiency would likely result from keeping at quite high rpm and correspondingly quite low current

In the energy flow diagram the +/- 120kW is shown on the mechanical link from MGU-K to the “Engine”. So I don’t think 126 notional kW is allowed.

Keeping the revs high during braking would seem to be quite a challenge for the timing of gearshifts, with 5 or 6 downshifts in a couple of seconds. I wonder if the drivers “cheat” selecting the next downshift as soon as the previous one is made and relying on oversoeed protection to delay the shift.

[subcritical71]

...
The MGU-K contribution is monitored by the sensor at the entry to the Control Unit for the K. The restrictions are +/- max 120kW and max 200Nm. These are conditioned by:
“A fixed efficiency correction of 0.95 will be used to monitor the maximum MGU-K power.” And “
The maximum torque of the MGU-K may not exceed 200Nm. The torque will be referenced to the crankshaft speed and the fixed efficiency correction defined in Article 5.2.2 will be used to monitor the maximum MGU-K torque.”

I struggle to get my head around the 0.95 factor for the generation side of things.
Does it mean that the monitored limits are +/- 126kW and 210Nm or +126, -114 and +210, -190 ?

aren't the limits ? ....
126 kW for outbound electrical energy (ie 126 measured electric kW becomes 120 notional mechanical kW) and
120 kW for inbound electrical energy (ie 126 notional mechanical kW becomes 120 measured electric kW)

they don't seem to want to get very close to the 200 Nm torque limiting region anyway
the best EM efficiency would likely result from keeping at quite high rpm and correspondingly quite low current

In the energy flow diagram the +/- 120kW is shown on the mechanical link from MGU-K to the “Engine”. So I don’t think 126 notional kW is allowed.

Keeping the revs high during braking would seem to be quite a challenge for the timing of gearshifts, with 5 or 6 downshifts in a couple of seconds. I wonder if the drivers “cheat” selecting the next downshift as soon as the previous one is made and relying on oversoeed protection to delay the shift.

Interesting thought as while watching some in car video of Bahrain during the race I thought that they were waiting a bit longer than I remember in the past to do the first downshift. In my mind these shifts seemed to occur much earlier in the braking phase.

Also, I think the shift can be balked but not delayed.

[Tzk]

Of course the torque delivered by the whole power unit isn't linked to ICE only. I never said so. But yes the torque delivered is linked to throttle position demand.

Torque delivered by the ice is linked to throttle position. Torque delivered by the power unit however is not.

Torque delivered by the power unit is linked to driver torque input ( accelerator pedal), however torque delivered by the ice is not.

Now: You‘ve stated above that the driver request is linked to the throttle body position, which simply isn‘t true. I‘m not sure that you actually meant to say this though.

[gruntguru]

Yes. There is no doubt that a given accelerator pedal position - say 50% - does not always result in the same position for the engine throttles. Engine throttle position will depend also on the state-of-charge, mode settings, turbine speed etc etc.