2026 Hybrid Powerunits

[Farnborough]

Looking and on-boards it seems every PU has a lot of gear whine now, except Audi. I wonder if this suggests that they all moved to a similar layout to Honda's with the K in the front right beside the battery. Mercedes sounds particularly vicious on on-board footage.

From what we've seen publicly (and logical in the throughput/capacity hike) the MGU is substantially larger in size this year.

It appeared to be common to mount alongside lower left side crankcase area prior to these reg, but Honda seem to have moved it to projecting forward now, but drive from front of ICE crankshaft is very similar in orientation for either of these. Effectively "stepping" out sideways to run MGU axis in offset parallel to ICE crankshaft typically.

Something that is likely to increase the audible output is the shear load now passing through that gearset from crankshaft in pos & neg direction.
Competition gear sets ordinarily utilise straight cut gears, this for efficiency and load capacity. They make a big racket though
All method of noise and likely potential vibration reduction are dispensed with to reduce power loss when passing through.

Conventional gearset use skew/helical cut tooth form to mitigate the above, also needing to be physically larger, for the same load transferred. They do this by the geometry between teeth faces on opposing gear pairs "driving" into engagement with much more sliding together during takup and dissipation of the load as each tooth is used.
That puts more side/axial load into each component, needing more axial support from local "thrust" bearings, and more heat through that sliding action. These of course are simply a waste in F1 performance.

In effect, high highly loaded straight cut minimalised gear set design, when fully loaded, makes a complete "screamer" of a transmission system. That's regardless of MGU location.

[michl420]

I noticed something with the rear facing red lights in the wing endplates.

We are used to them flasing when harvest kicks in, that is still there.

But I noticed that often, not always, they would blink red a single time, at the beginning of the straights or even at the end of the corner just before a straight.
Is that to signal that electric deplowment has started? And why would the guy behind need to know that you will accelerate faster?

P.S. attempting to fix the thread name now that we have more data and less spculation.

I saw, and wondered the same.

Likewise unable to attach definitive reason to it, but thought was toward finishing "harvesting" with at that point full forward torque deployed ? Possibly as the driver ultimately hit full power request in throttle travel.


Perhaps to give some warning to following driver of impending rear ending potential, but couldn't see enough incident to give reasonable correlated comment.

1 blink is reduced max mguk powern (for batterie saving)
2 blinks is 0 mguk power (clipping)
steady blinks is superclipping

[Zynerji]

Steaming hot garbage.

[mzso]

Something that is likely to increase the audible output is the shear load now passing through that gearset from crankshaft in pos & neg direction.
Competition gear sets ordinarily utilise straight cut gears, this for efficiency and load capacity. They make a big racket though
All method of noise and likely potential vibration reduction are dispensed with to reduce power loss when passing through.

Well, I guess load increases the sound a lot then. For Mercedes it easily overpowers the ICE. (provided they didn't find a special position for the microphone atop the gearbox.
What do you think of this?:

It goes very silent when there's no load. There is only a stuttery sound, at much lower volume.
Is rotation by the ICE enough for this, or are they doing rev matching?
I would guess a geared electric motor would have substantial negative effect for a race engine if it's just dead-weight rotating. I don't think they'd want flywheels.

Conventional gearset use skew/helical cut tooth form to mitigate the above, also needing to be physically larger, for the same load transferred. They do this by the geometry between teeth faces on opposing gear pairs "driving" into engagement with much more sliding together during takup and dissipation of the load as each tooth is used.
That puts more side/axial load into each component, needing more axial support from local "thrust" bearings, and more heat through that sliding action. These of course are simply a waste in F1 performance.

In effect, high highly loaded straight cut minimalised gear set design, when fully loaded, makes a complete "screamer" of a transmission system. That's regardless of MGU location.

I would think that herring bone gears would combine the advantages of both styles. Maybe a minuscule increase in friction.

[Farnborough]

Something that is likely to increase the audible output is the shear load now passing through that gearset from crankshaft in pos & neg direction.
Competition gear sets ordinarily utilise straight cut gears, this for efficiency and load capacity. They make a big racket though
All method of noise and likely potential vibration reduction are dispensed with to reduce power loss when passing through.

Well, I guess load increases the sound a lot then. For Mercedes it easily overpowers the ICE. (provided they didn't find a special position for the microphone atop the gearbox.
What do you think of this?:

It goes very silent when there's no load. There is only a stuttery sound, at much lower volume.
Is rotation by the ICE enough for this, or are they doing rev matching?
I would guess a geared electric motor would have substantial negative effect for a race engine if it's just dead-weight rotating. I don't think they'd want flywheels.

Conventional gearset use skew/helical cut tooth form to mitigate the above, also needing to be physically larger, for the same load transferred. They do this by the geometry between teeth faces on opposing gear pairs "driving" into engagement with much more sliding together during takup and dissipation of the load as each tooth is used.
That puts more side/axial load into each component, needing more axial support from local "thrust" bearings, and more heat through that sliding action. These of course are simply a waste in F1 performance.

In effect, high highly loaded straight cut minimalised gear set design, when fully loaded, makes a complete "screamer" of a transmission system. That's regardless of MGU location.

I would think that herring bone gears would combine the advantages of both styles. Maybe a minuscule increase in friction.

The "stuttery" sound is pos and neg "drag" within the gear train, making just enough in that flick/flack effect to emit sound to us. Seems to be generated by crankshaft speed, with the generator intertia unable to follow exactly.

As far as I understand it, there's no " clutching" allowed, in F1 at least. And so, MGU rotation is inextricably linked to crank rpm. Effectively it's revolving at it's geared for speed, all of the time. Then "exited" by power OR regeneration as you hear it start to scream.

Most generators in automotive are this type (broad generalisation) and as you note, contribute to flywheel effect overall.
Many diesel road car application use a sprag clutch (one way drive) on generator pulley, that to cope with low crank rpm oscillations, the generator rotational mass can go faster than its belt drive to stop the crank derived oscillations "rattling" the generator.

If a test is made on a simple automotive alternator, spinning the rotor with electric drill, it spins very free when unloaded, but switching the field coil on when doing that to get power, then you really have to hold tight to it as torque reaction takes place.

Herringbone is very complex to use, in comparison to a stripped back and crudely simple straight cut gearset. I doubt they've not considered and evaluated everything on the table currently in design of these in place now.

[hollus]

1 blink is reduced max mguk powern (for batterie saving)

Any source for that?
It sounds unlikely precisely at the begining of traction zones.
Or would it be in the transition from 0 towards max electric deployment? And calling that "reduced". It felt to me like the blink happened in full traction zones, but come to think of it, only from slow corners, so... throttle modulation?
What a bizarre thing to "broadcast".

[karana]

1 blink is reduced max mguk powern (for batterie saving)

Any source for that?
It sounds unlikely precisely at the begining of traction zones.
Or would it be in the transition from 0 towards max electric deployment? And calling that "reduced". It felt to me like the blink happened in full traction zones, but come to think of it, only from slow corners, so... throttle modulation?
What a bizarre thing to "broadcast".

https://www.youtube.com/shorts/hWFvl57aVvg

[mzso]

1 blink is reduced max mguk powern (for batterie saving)

Any source for that?
It sounds unlikely precisely at the begining of traction zones.
Or would it be in the transition from 0 towards max electric deployment? And calling that "reduced". It felt to me like the blink happened in full traction zones, but come to think of it, only from slow corners, so... throttle modulation?
What a bizarre thing to "broadcast".

Well, whatever is it that happens at the start of the straight, if it's brief enough it will only be one blink.

[mzso]

The "stuttery" sound is pos and neg "drag" within the gear train, making just enough in that flick/flack effect to emit sound to us. Seems to be generated by crankshaft speed, with the generator intertia unable to follow exactly.

If a test is made on a simple automotive alternator, spinning the rotor with electric drill, it spins very free when unloaded, but switching the field coil on when doing that to get power, then you really have to hold tight to it as torque reaction takes place.

Well, I was thinking a multi 100 kW electric motor might by hefty as a flywheel. So maybe using electricity to match revs when not allowed to drive or regen is allowed? (Both in F1 and WEC)

Herringbone is very complex to use, in comparison to a stripped back and crudely simple straight cut gearset. I doubt they've not considered and evaluated everything on the table currently in design of these in place now.

What makes it complex to use? I mean you can't just slide the teeth into each other, but I can't think of anything else.

[hollus]

1 blink is reduced max mguk powern (for batterie saving)

Any source for that?
It sounds unlikely precisely at the begining of traction zones.
Or would it be in the transition from 0 towards max electric deployment? And calling that "reduced". It felt to me like the blink happened in full traction zones, but come to think of it, only from slow corners, so... throttle modulation?
What a bizarre thing to "broadcast".

https://www.youtube.com/shorts/hWFvl57aVvg

Thanks! Cheers.

[Farnborough]

The "stuttery" sound is pos and neg "drag" within the gear train, making just enough in that flick/flack effect to emit sound to us. Seems to be generated by crankshaft speed, with the generator intertia unable to follow exactly.

If a test is made on a simple automotive alternator, spinning the rotor with electric drill, it spins very free when unloaded, but switching the field coil on when doing that to get power, then you really have to hold tight to it as torque reaction takes place.

Well, I was thinking a multi 100 kW electric motor might by hefty as a flywheel. So maybe using electricity to match revs when not allowed to drive or regen is allowed? (Both in F1 and WEC)

Herringbone is very complex to use, in comparison to a stripped back and crudely simple straight cut gearset. I doubt they've not considered and evaluated everything on the table currently in design of these in place now.

What makes it complex to use? I mean you can't just slide the teeth into each other, but I can't think of anything else.

Firstly, in this application, just noise output and nobody cares if it does or doesn't. That's the overriding factor here, with ultimate funding available, if its of value in transmitting power with less loss, would have been utilised by now by teams.

Anything, Helix cut or diagonal will trade power loss for heat generation, there being two in this arrangement (herringbone) extending that loss (straight cut vs helical) to heat output that needs to be managed also.

They have to be extraordinarily precise, a straight cut set has much more latitude, with anything in the way of wear punished substantially, and likely drives failure risk more quickly if it moves in wrong direction. My impression is that's more problematic in raised speed range, as used here.

Gear sets are ordinarily made with backlash in preventing the opposing faces of each tooth from giving conflict, hence heat and wear failure. Giving clearance/backlash obviates that eventuality.
Even in the gearbox on these cars they place the highest torque/leverage geared pairs (1st &2nd etc) nearest to the bearing for support, as the potential for the gearshaft to bend under load gets those gears moving apart in extremis.

The attributes of a simple straight cut gear train are:- the narrowest, lightest gears possible. The least heat generative and most potential power transfer available while also giving the casing form factor the smallest potentially, then likely the most reliable for this small form of both gear and case size.
Negative, they're noisy, doesn't matter here.

[AR3-GP]

Why aren't the teams just charging the battery after the car stops in the grid box? This is permitted by the regulations:

https://i.postimg.cc/9QR6d88Q/image.png

Downside might be that you could overheat the ICE very easily without any airflow in the radiators.

In the pre-race show they explained that Antonelli had trouble with his start because the battery was too full, so the ICE didn't have a load on it (from MGU-K) to properly spool the turbo. It sounds like quite a complicated dance to get the start correct. Not using too much battery while trying to warm the tires and brakes, but no overcharging the battery so that you have room to load the ICE against the MGU-K during the 5 second for the best turbo spooling. They'll have to do bespoke calculations for every single circuit.

[venkyhere]

Why aren't the teams just charging the battery after the car stops in the grid box? This is permitted by the regulations:

https://i.postimg.cc/9QR6d88Q/image.png

Downside might be that you could overheat the ICE very easily without any airflow in the radiators.

In the pre-race show they explained that Antonelli had trouble with his start because the battery was too full, so the ICE didn't have a load on it (from MGU-K) to properly spool the turbo. It sounds like quite a complicated dance to get the start correct. Not using too much battery while trying to warm the tires and brakes, but no overcharging the battery so that you have room to load the ICE against the MGU-K during the 5 second for the best turbo spooling. They'll have to do bespoke calculations for every single circuit.

Can't they have a burn off resistor (which can be used to heat up the engine, even) for such cases where battery is full and revvs need to be high ? This is a problem that was beautifully solved by the MGU-H after the KERS era, think of the effort that went into it, and now just because Audi doesn't want it, we don't have it.
More and more data proving how stupid these regulations are. A step backwards, instead of forwards.

[mzso]

Can't they have a burn off resistor (which can be used to heat up the engine, even) for such cases where battery is full and revvs need to be high ?

Well carbon fiber is conductive. Why not pass current through the car? Saves on weight.
The added weight would be prohibitive I guess, even if it was allowed. Several kWs only in scarce and brief circumstances. Though I'm not sure what happens when someone wants to break with a full battery. Seems insane that he would loose almost all of the rear brakes. It would look like Hamilton with the upside down wing in FP1.

[Farnborough]

Why aren't the teams just charging the battery after the car stops in the grid box? This is permitted by the regulations:

https://i.postimg.cc/9QR6d88Q/image.png

Downside might be that you could overheat the ICE very easily without any airflow in the radiators.

In the pre-race show they explained that Antonelli had trouble with his start because the battery was too full, so the ICE didn't have a load on it (from MGU-K) to properly spool the turbo. It sounds like quite a complicated dance to get the start correct. Not using too much battery while trying to warm the tires and brakes, but no overcharging the battery so that you have room to load the ICE against the MGU-K during the 5 second for the best turbo spooling. They'll have to do bespoke calculations for every single circuit.

We had on here a discussion prior to first race about this, in which I forwarded this method "arrival on grid with % depleted battery to facilitate on load running as assistance to spooling turbo" and which does seem to be part of strategy we see now.

Exhaust ouput, volume/energy etc will always be maximized under as much torque load the ICE is subjected to. Thats always way above no load high rpm, and far more useful in this scenario.

Its the very èssence of turbo "lag" low torque until turbo comes on stream, low exhaust energy until subjected to full load. A catch 22 in that temporary phase of running, and now without any real means of effective gap filling by regulation in removing anything that would help ... H component, fuel in exhaust, variable nozzle turbo architecture, no K assistance either at lowest launch phase. I don't know what they were expecting from this scenario in setting these rules

There's inevitability about this, launch a very high traction mass, with naff all torque at lowest rpm when turbo is off boost, and with a tiny clutch that won't absorb much in the way of slipping as it'll just go up in smoke .... couldn't make this up really, what a crap shoot.

Ferrari the only one to initially address this by biasing their design to to fulfill this extremely constrained part of regulation. Whether it ultimately compromises them elsewhere we've yet to fully appreciate.