Honda Power Unit Hardware & Software

[diffuser]

Partial load recovery is definitely preferable to superclipping. The latter actually slows the cars down while it's happening, the time gain in other parts of the track just outweigh the losses. The only downside of partial throttle recovery is increased fuel consumption (and apparently driveability at least for Honda).

Observation and info leaking out doesn't seem to support this. Apparently they gain more by preferentially super-clipping and then using that energy to accelerate at the next straight while saving some fuel, rather than running the ICE in the turning phase.
I definitely don't hear the engine running at high power in turns effect that Newey alluded to sometime last year.

That was an odd statement as the ICE crankshaft is directly linked to rear drive through whichever ratio is in use in the gearbox (with clutch fully closed) meaning you'd not hear an increase in rpm, but the torque request from ECU can easily be higher than required to give the asked for track speed, the MGUK being the beneficiary of that strategy.

Right, the RPM would have to equate to the selected gear and the vehicle's speed.

[mzso]

Partial load recovery is definitely preferable to superclipping. The latter actually slows the cars down while it's happening, the time gain in other parts of the track just outweigh the losses. The only downside of partial throttle recovery is increased fuel consumption (and apparently driveability at least for Honda).

Observation and info leaking out doesn't seem to support this. Apparently they gain more by preferentially super-clipping and then using that energy to accelerate at the next straight while saving some fuel, rather than running the ICE in the turning phase.
I definitely don't hear the engine running at high power in turns effect that Newey alluded to sometime last year.

That was an odd statement as the ICE crankshaft is directly linked to rear drive through whichever ratio is in use in the gearbox (with clutch fully closed) meaning you'd not hear an increase in rpm, but the torque request from ECU can easily be higher than required to give the asked for track speed, the MGUK being the beneficiary of that strategy.

So? More fuel, more power, more sound.
I didn't say even say higher rpm. But, if they were really serious about recharging during the turning phase they would always be in a lower gear then what we come to expect, so the PU could remain at higher RPM for better recovery.

[diffuser]

Observation and info leaking out doesn't seem to support this. Apparently they gain more by preferentially super-clipping and then using that energy to accelerate at the next straight while saving some fuel, rather than running the ICE in the turning phase.
I definitely don't hear the engine running at high power in turns effect that Newey alluded to sometime last year.

That was an odd statement as the ICE crankshaft is directly linked to rear drive through whichever ratio is in use in the gearbox (with clutch fully closed) meaning you'd not hear an increase in rpm, but the torque request from ECU can easily be higher than required to give the asked for track speed, the MGUK being the beneficiary of that strategy.

So? More fuel, more power, more sound.
I didn't say even say higher rpm. But, if they were really serious about recharging during the turning phase they would always be in a lower gear then what we come to expect, so the PU could remain at higher RPM for better recovery.

It's always muffled by the turbine. Doudt you'd hear the difference with all the other noise going on.

[GhostF1]

I think the wording of Newey's quote was maybe a little weird. Made it sound like everyone would be coasting at 11,000rpm through corners when in reality he likely was trying to imply they'd be downshifting brutally into lower gears and ICE would be running at high torque demand against a braking MGU-K, which they are.

For those who can't really picture what's going on exactly;
Get in your car, accelerate, but then also press and hold the brake while still holding the accelerator down (somewhere safe obviously).
That is essentially exactly what is happening (but replace brakes and heat generated with electric motor and battery charge generated). The engine torque demand is obviously still high, but the MGU-K is applying extreme resistance so the RPM drops and slows the car, and that MGUK resistance against the ICE with high torque demand trying to rotate it, generates battery charge.
This would also have a "brake boosting" effect to keep the turbo spooled.

Easiest way to visualise!

The tuning of this so there isn't sudden jerks forward or varied braking force during gear changes would have to be extremely precise. Honda implying their engine does not provide linear torque while being braked is their issue and if you have a difficult or fragile gearbox, you're double screwed.

There's a video somewhere online of VCARB on track at Imola before the season started, it was the first taste of the RBPT engine. And there's a moment coming into T1 where you can see Liam's head jerking forward and back constantly while he slows down and goes through the gears (symptom of the tuning not finished yet). Very indicative of what the engineers need to perfect for these regs. I'll see if I can find it and highlight it.

[PlatinumZealot]

"This partial regeneration also occurs during acceleration. The internal combustion engine must generate stable torque to counteract the braking force of the MGU-K; otherwise, the driver cannot safely press the accelerator. The main cause of this lies in the internal combustion engine’s lack of power and its still-immature combustion settings."

This part - regenerating during acceleration - doesn't make sense.

Why don't you think it makes sense?

I'm curious.

[PlatinumZealot]

I think the wording of Newey's quote was maybe a little weird. Made it sound like everyone would be coasting at 11,000rpm through corners when in reality he likely was trying to imply they'd be downshifting brutally into lower gears and ICE would be running at high torque demand against a braking MGU-K, which they are.

For those who can't really picture what's going on exactly;
Get in your car, accelerate, but then also press and hold the brake while still holding the accelerator down (somewhere safe obviously).
That is essentially exactly what is happening (but replace brakes and heat generated with electric motor and battery charge generated). The engine torque demand is obviously still high, but the MGU-K is applying extreme resistance so the RPM drops and slows the car, and that MGUK resistance against the ICE with high torque demand trying to rotate it, generates battery charge.
This would also have a "brake boosting" effect to keep the turbo spooled.

Easiest way to visualise!

The tuning of this so there isn't sudden jerks forward or varied braking force during gear changes would have to be extremely precise. Honda implying their engine does not provide linear torque while being braked is their issue and if you have a difficult or fragile gearbox, you're double screwed.

There's a video somewhere online of VCARB on track at Imola before the season started, it was the first taste of the RBPT engine. And there's a moment coming into T1 where you can see Liam's head jerking forward and back constantly while he slows down and goes through the gears (symptom of the tuning not finished yet). Very indicative of what the engineers need to perfect for these regs. I'll see if I can find it and highlight it.

It's not linear torque they are refering to. I'm not sure what linear even means here but most of the Engines will have whatever torque shape they need to go fastest. I'm not sure anyone has ever dynoed these engines and publish the torque curve. Even though we know the fuel limit is linear.

They are refering to stable torque though. As in torque variation over miliseconds for a given load. Think of drawing a curve and it looks smooth from a far but up close it is very squiggly and unstable. They can solve this by making the combustion stronger and more stable.

[GhostF1]

I think the wording of Newey's quote was maybe a little weird. Made it sound like everyone would be coasting at 11,000rpm through corners when in reality he likely was trying to imply they'd be downshifting brutally into lower gears and ICE would be running at high torque demand against a braking MGU-K, which they are.

For those who can't really picture what's going on exactly;
Get in your car, accelerate, but then also press and hold the brake while still holding the accelerator down (somewhere safe obviously).
That is essentially exactly what is happening (but replace brakes and heat generated with electric motor and battery charge generated). The engine torque demand is obviously still high, but the MGU-K is applying extreme resistance so the RPM drops and slows the car, and that MGUK resistance against the ICE with high torque demand trying to rotate it, generates battery charge.
This would also have a "brake boosting" effect to keep the turbo spooled.

Easiest way to visualise!

The tuning of this so there isn't sudden jerks forward or varied braking force during gear changes would have to be extremely precise. Honda implying their engine does not provide linear torque while being braked is their issue and if you have a difficult or fragile gearbox, you're double screwed.

There's a video somewhere online of VCARB on track at Imola before the season started, it was the first taste of the RBPT engine. And there's a moment coming into T1 where you can see Liam's head jerking forward and back constantly while he slows down and goes through the gears (symptom of the tuning not finished yet). Very indicative of what the engineers need to perfect for these regs. I'll see if I can find it and highlight it.

It's not linear torque they are refering to. I'm not sure what linear even means here but most of the Engines will have whatever torque shape they need to go fastest. I'm not sure anyone has ever dynoed these engines and publish the torque curve. Even though we know the fuel limit is linear.

They are refering to stable torque though. As in torque variation over miliseconds for a given load. Think of drawing a curve and it looks smooth from a far but up close it is very squiggly and unstable. They can solve this by making the combustion stronger and more stable.

You're absolutely right and a poor word choice on my part!

[Farnborough]

The word "Reciprocal" would be more appropriate in regard to torque response.

The instability of torque delivery and predictability in response to "call" command from both throttle and ecu overlay giving the driver unaccounted for torque delivery at very critical - moment by moment - throttle modulation.

I think it was attributed to Alonso in that he said the response when changing to 1st gear in Monaco would likely put him in the barrier. Lacking confidence, and pure track space to use it reliably. This an absolutely critical need for a driver to repeatedly place the chassis as directed.

Their problem, reading between public statement "lines" appears to be that the linearity ( correct here) relates to just how each singular combustion event propagates, without which the ecu programming cannot deliver the required response.

That, as already noted from different sources, speaks to fuel characteristics and the combustion "landscape" they are applying. Not generating an assured predictable burn for each and every combustion event, preventing logical and assured basis to which they can apply algorithm in overall control.

[mzso]

I think the wording of Newey's quote was maybe a little weird. Made it sound like everyone would be coasting at 11,000rpm through corners when in reality he likely was trying to imply they'd be downshifting brutally into lower gears and ICE would be running at high torque demand against a braking MGU-K, which they are.

Yes, but evidently overwhelmingly they do it in the straights, via superclipping.

[Tommy Cookers]

.... torque response.
The instability of torque delivery and predictability in response to "call" command from both throttle and ecu overlay giving the driver unaccounted for torque delivery at very critical - moment by moment - throttle modulation.
... Alonso ,,, said the response when changing to 1st gear in Monaco would likely put him in the barrier. Lacking confidence, and pure track space to use it reliably. This an absolutely critical need ....
... how each singular combustion event propagates, without which the ecu programming cannot deliver the required response.

yes ...
if the ICE response to severe transient torque demands is inconsistent - this is failure

the gearshift is anyway a severe transient
ie the PU rpm ideally accelerates or decelerates by 14% in the very few milliseconds the clutch is open
(it can actually only make say 7% - this gives about 75% of the ideal impulse relief)

the MG (& PU) acceleration/deceleration response is predictable and reliable ....
unless there's a new and desperate-for-efficiency ICE strapped to the MG

with shifting (or in active control in the MG-to-crank gear train) the normal limits on MG torque and power don't apply
ie the MG displacement will be very close to the end of the synchronicity range ....
so any imperfection in the ICE transient behaviour may cause pole slipping (or skipping) .....
whereby the MG torque could collapse (regardless of what the ECU wants)
(the system act like gears made from strong rubber ..... above a certain torque it jumps a tooth)
and/or a cyclic effect ie 'torque stutter' might develop

the car responds to the change demanded by the accelerator pedal position ....
the ECU acts as an internal 'accelerator pedal' giving a demand to so-called control electronics CE that drive the MG
but the CE independently calculate MG actions from checks on MG rotational position made 1000 times per second ..
these actions (to maintain synchronicity) can over-rule the ECU


relative to the ICE the MG is now 4 times the size it was in 2025
why are people surprised that there is now more and more varied energy flow onboard the cars ?
why are people surprised that now the shift quality is worse ?