2025/2026 Hybrid Powerunit speculation

[Tommy Cookers]

This new rule might limit the amount of fuel the ICE can burn to recharge the battery:

5.4.5 At partial load, the fuel energy flow must not exceed the limit curve defined below:

• EF (MJ/h) = 380 when the engine power is equal to or below ‐50kW
• EF (MJ/h) = 9.78 x engine power (kW) + 869 when the engine power is above ‐50kW

the fuel energy flow rate limited to 3000 MJ/h
3000 MJ/h is 833 kW if the fuel conversion to energy is performed at an efficiency of 100%

at 3000 MJ/h with ICE efficiency of 45% and MG efficiency of 95% MG-generated power is 362 kW (limited to 350 kW oc)

btw
https://www.racefans.net/2023/03/25/exp ... el-future/
doesn't explain that (as currently the carbon-neutral fuel's energy of production is 6 times its energy content) .....
the resultant global warming can easily be greater than a conventional fuel's ... though paradoxically ...
if produced with (less) and otherwise surplus renewable electricity it might be seen as some equivalent of the battery

[Juzh]

When we say 45% efficiency we mean at the crank, right?

[saviour stivala]

ICE "Thermal efficiency" is. i.e. percentage of the potential energy in the fuel delivered to the CRANKSHAFT.
PU "efficiency" is. i.e. ICE thermal efficiency + ERS efficiency also delivered to the CRANKSHAFT.

[mzso]

When running the engine against the MGU under braking the fuel consumption goes up. (burning fuel to charge ES).

This new rule might limit the amount of fuel the ICE can burn to recharge the battery:

5.4.5 At partial load, the fuel energy flow must not exceed the limit curve defined below:

• EF (MJ/h) = 380 when the engine power is equal to or below ‐50kW
• EF (MJ/h) = 9.78 x engine power (kW) + 869 when the engine power is above ‐50kW

Agree that 5.4.5 seems to might limit the amount of fuel the ICE can burn to recharge the ES. But by agreeing/saying that ''running the engine against the MGU'', the original said "under braking", the ICE would be burning 'EXTRA' fuel to charge ES, WE are risking another down-vote by those that doesn't think so.

If you use more fuel to later use more efficiently with the MGU, then you still end up saving fuel. What is there to argue?
The only thing in question is how much prospect is there for doing this.

[saviour stivala]

While risking another downvote. Still to me the extra burning of fuel to recharge ES is not only comparable to robbing Peter to pay Paul, but having to add interests to Paul's payments.

[Stu]

While risking another downvote. Still to me the extra burning of fuel to recharge ES is not only comparable to robbing Peter to pay Paul, but having to add interests to Paul's payments.

I think that this is the theory behind the teams pushing back against the reduction in fuel allocation (to 60kg) for the race.
Considering the limitations on fuel delivery at low ICE outputs it would be interesting to see the real effect on lap time of the increased fuel allowance (either 100kg or 110kg - cannot recall atm which figure was settled upon) with ‘sacrificial’ KERS boosting versus the reduced starting mass of the car.

This maybe a stupid question, but in order to push the K regen during the braking phase, will teams change their downshift philosophy with regards to the clutch operation (what driver would be comfortable with the the PU trying to ‘push them on’ in the braking zones) & what fall-back/fail-safe procedures will be required in the event of a software glitch/hardware failure while operating the ICE for regen maximisation?

[saviour stivala]

Re the subject of using the ICE under braking to charge ES. Quoting from the Bianchi tragic accident report. "During the 2 seconds Bianchi's car was leaving the track and traversing the run-off area, he applied both throttle and brake together, using both feet. the failsafe algorithm is designed to override the throttle and cut the engine, but was inhibited by the torque coordinator, which controls the rear brake-by-wire system''

[Vanja #66]

When ICE is charging the battery via MGU on straights, it's doing so in a regime where it provides near constant output and this regime can be made quite efficient. On the other hand, when the car is accelerating ICE inefficiently generates huge amount of excess power - which is also effectively wasted no matter what. MGU provides acceleration a lot more efficiently and this is where massive harvesting on straights makes sense in a 50-50 ICE-MGU architecture. Is this a correct assumption.

[Stu]

When ICE is charging the battery via MGU on straights, it's doing so in a regime where it provides near constant output and this regime can be made quite efficient. On the other hand, when the car is accelerating ICE inefficiently generates huge amount of excess power - which is also effectively wasted no matter what. MGU provides acceleration a lot more efficiently and this is where massive harvesting on straights makes sense in a 50-50 ICE-MGU architecture. Is this a correct assumption.

I would say that this is a perfect description for the MGU-H operation currently in use, but not sure that it is for MGU-K.
Maybe I’m just not able to ‘see it’, but if the K MGU & ICE are connected before the clutch system any charging under acceleration will have a negative impact on performance (and is a time when you expect the MGU to be in output mode).
I can ‘see’ that a new version of ‘clipping’ will be possible once the required speed is achieved on any straight, prior to the braking event (something similar in effect to lift & coast, but systematically induced).

[Tommy Cookers]

While risking another downvote. Still to me the extra burning of fuel to recharge ES is not only comparable to robbing Peter to pay Paul, but having to add interests to Paul's payments.

this 'robbing Peter to pay Paul' is wrong

it would be right if ICE efficiency was the same (or better) at lower power than at higher power ...
but it isn't (engine efficiency falls 0% at idle)
without the fall in efficiency due to throttling the normal SI vehicle would be undriveable
(the economy advantage of the CI engine is in large part due to the air supply always being unthrottled)

yes the ICE control of the 2014-2025 F1 car is somewhat unknown (and there's cylinder cutting of course)
cartainly it has throttling and the above characteristic is anyway reinforced by ....
fuel rate rules dictating reducing fuel per cycle with rpm below 10500 (and for rpm above 10500 of course)

AMuS (Gruner) has quite recently written that the F1 cars use 30% of the fuel to generate electricity

[wuzak]

I can ‘see’ that a new version of ‘clipping’ will be possible once the required speed is achieved on any straight, prior to the braking event (something similar in effect to lift & coast, but systematically induced).

The rules allow for the MGUK output to fall by 450kW during full throttle application, and the recently released revised regulations allow the MGUK to recover up to 100kW when the driver is at full throttle.

This would be the "clipping" you talk about.

[wuzak]

While risking another downvote. Still to me the extra burning of fuel to recharge ES is not only comparable to robbing Peter to pay Paul, but having to add interests to Paul's payments.

I think that this is the theory behind the teams pushing back against the reduction in fuel allocation (to 60kg) for the race.

Given that they currently have a maximum 110kg for the race with fuel that has energy density of 45MJ/kg or greater, and the new rule specify a energy density of between 38.0MJ/kg and 41.0MJ/kg, reducing the race fuel allocation to 60kg would be extreme.

Or, to look at another way, the ICE output is reduced by ~1/3, but fuel allowance is cut by 45%.

I was thinking about the fuel requirements for a race the other day.
They could specify the amount of fuel energy for the race (say 3,000MJ), or they could specify fuel by weight (say 70kg).

But if your fuel is at the lower end of the energy density range, you are at a disadvantage either way.
If the fuel is specified by mass, you could be as much as 7.3% down on fuel energy.
If it is by energy, your fuel mass would be around 7.9% more.

[Tommy Cookers]

When ICE is charging the battery via MGU on straights, it's doing so in a regime where it provides near constant output and this regime can be made quite efficient. On the other hand, when the car is accelerating ICE inefficiently generates huge amount of excess power - which is also effectively wasted no matter what. MGU provides acceleration a lot more efficiently and this is where massive harvesting on straights makes sense in a 50-50 ICE-MGU architecture. Is this a correct assumption.

I would say that this is a perfect description for the MGU-H operation currently in use, but not sure that it is for MGU-K.
Maybe I’m just not able to ‘see it’, but if the K MGU & ICE are connected before the clutch system any charging under acceleration will have a negative impact on performance (and is a time when you expect the MGU to be in output mode).
I can ‘see’ that a new version of ‘clipping’ will be possible once the required speed is achieved on any straight, prior to the braking event (something similar in effect to lift & coast, but systematically induced).

all stored energy has been bought from the ICE - that's how is the prime mover
so in that sense the MG isn't 'more efficient' than the ICE

charging under acceleration makes sense when the wheels could otherwise spin at ICE-WOT power
and because that makes more power available for use whenever spending stored energy will be race-profitable
these 2 phases may even be contiguous ....
there's continuous variability (the MG isn't jumping from -350 kW to +350 kW)

the MG must be a stable system - so has 'baked in' behaviour that some could have regarded as a driver aid ....
a 350 kW 'elephant in the room' - it would be justified as in line with the ICE mapping limitations started decades ago

[saviour stivala]

While risking another downvote. Still to me the extra burning of fuel to recharge ES is not only comparable to robbing Peter to pay Paul, but having to add interests to Paul's payments.

this 'robbing Peter to pay Paul' is wrong

it would be right if ICE efficiency was the same (or better) at lower power than at higher power ...
but it isn't (engine efficiency falls 0% at idle)
without the fall in efficiency due to throttling the normal SI vehicle would be undriveable
(the economy advantage of the CI engine is in large part due to the air supply always being unthrottled)

yes the ICE control of the 2014-2025 F1 car is somewhat unknown (and there's cylinder cutting of course)
cartainly it has throttling and the above characteristic is anyway reinforced by ....
fuel rate rules dictating reducing fuel per cycle with rpm below 10500 (and for rpm above 10500 of course)

AMuS (Gruner) has quite recently written that the F1 cars use 30% of the fuel to generate electricity

The analogy of robing Peter to pay Paul and plus to that having to pay Paul interest rate was due to the 'original argument of ''running the engine 'UP' against the MGU during braking', This is burning fuel to charge ES for the new 2026 formula. This stuff which is not presently possible as the rules are, will have to involve drastic changes in 2026 to the rear brake-by-wire fail-safe system. Re Gruner claim of 'present' FI cars use 30% of fuel to generate electricity, At present any fuel burned to generate electricity is burned by the use of the MGU-H to generate electricity. All other regeneration is generated under braking.

[saviour stivala]

When ICE is charging the battery via MGU on straights, it's doing so in a regime where it provides near constant output and this regime can be made quite efficient. On the other hand, when the car is accelerating ICE inefficiently generates huge amount of excess power - which is also effectively wasted no matter what. MGU provides acceleration a lot more efficiently and this is where massive harvesting on straights makes sense in a 50-50 ICE-MGU architecture. Is this a correct assumption.

I would say that this is a perfect description for the MGU-H operation currently in use, but not sure that it is for MGU-K.
Maybe I’m just not able to ‘see it’, but if the K MGU & ICE are connected before the clutch system any charging under acceleration will have a negative impact on performance (and is a time when you expect the MGU to be in output mode).
I can ‘see’ that a new version of ‘clipping’ will be possible once the required speed is achieved on any straight, prior to the braking event (something similar in effect to lift & coast, but systematically induced).

Yes. "THIS" is a perfect description for the MGU-H (present system) but certainly not for the for the MGU-K. MGU-K (present system) only generates electricity (charge ES) under braking.