2025/2026 Hybrid Powerunit speculation

[mzso]

Or it's about watching cars pushed flat out...

You reckon the cars are still spectacular in qualifying even with the superclipping and sacrificing corners?

They weren't spectacular to begin with. Drivers driving around alone. Sometimes they botch the braking, or run off the circuit, or worse.

[hollus]

...the cars should not decide to take fast corners 40-50 kph lower than they can just because its optimal for laptime...

That is not happening. If the car is 50 km/h slower is because the driver demanded partial power to the wheels instead of full, or maybe he demended no power to the wheels instead of partial. The car will automatically decide to harves the difference between the driver's demand to the wheels and the ICE's max output, but it is the driver deciding to go that way.
The only exception (I think) is a corner that can be taken flat out, and even then, the driver has an override button.

This is the driver allowing the car to maximize lap time, not the car telling the driver how to maximize lap time.

[gearboxtrouble]

...the cars should not decide to take fast corners 40-50 kph lower than they can just because its optimal for laptime...

That is not happening. If the car is 50 km/h slower is because the driver demanded partial power to the wheels instead of full, or maybe he demended no power to the wheels instead of partial. The car will automatically decide to harves the difference between the driver's demand to the wheels and the ICE's max output, but it is the driver deciding to go that way.
The only exception (I think) is a corner that can be taken flat out, and even then, the driver has an override button.

This is the driver allowing the car to maximize lap time, not the car telling the driver how to maximize lap time.

Yeah thats my point. Super clipping means the car slows down in fast corners because that's the optimal harvesting strat even when the throttle is pinned. It also creates a feedback loop that incents the teams to run (and develop) less downforce than they'd normally do because they'd optimize their deployment on the straights. Mercedes (and to a lesser extent Red Bull) showed the best way to optimize deployment is to deploy all of it in the acceleration zones leading into the active aero zones. This means super clipping in fast corners and in straightline mode which are the only other times 100% throttle is asked for. The driver has no input into that decision to slow the car down in the fast corners. We're going to see cars slow similarly in Eau Rouge and Copse and it will be awful. The rules need to be adjusted to effectively end super clipping on any track.

[hollus]

The driver has no input into that decision to slow the car down in the fast corners.

But the driver has, AFAIK, an override button for that. If they were in the middle of an overtake in said fast corner, I guarantee they’d push it.
The driver CHOOSES not to do that in the interest of lap time, because lap time = future track position.

It looks and feels very different, but it is no different from tire saving, from starting a race underfueled, or even from slowing down for a corner, if you think about it.

[bananapeel23]

The driver has no input into that decision to slow the car down in the fast corners. We're going to see cars slow similarly in Eau Rouge and Copse and it will be awful. The rules need to be adjusted to effectively end super clipping on any track.

The adjustment we need is a reduction in peak output large enough to remove the incentive to clip through fast corners or at the end of straights.

We still want super clipping. It works great in constant speed or gradually slowing turns like hairpins or tightening esses or the turns 1-4 complex in Shanghai.

The way to do that is by making the cars rocket out of corners slightly slower. In other words by reducing peak deployment. If you can’t get back ip to top speed as quickly, the incentive to clip in Eau Rouge or the chicane in Melbourne disappears.

The drivers might still lift and coast a bit into braking zones, but they won’t be super clipping for several seconds.

This will make the cars significantly slower, but it would be a decent bandaid fix for now, until they can work out a long term solution. That long term solution could be more fuel and/or maybe loosening some of the cost-cutting engine restrictions like 16:1 compression ratio, low boost pressure and variable trumpets ban. An unrealistic option could be an anti-lag system in the form of an electric turbo (but not MGU-H because it would be prevented from harvesting and would resort to the wastegate for overpressure)

Realistically though, politics made these regs a mess because Mercedes was afraid of Audi and Audi was afraid of the MGU-H and wanted a nominal 50/50 power split. As a result we got a compromise that removed both and resulted in lower efficiency engines with super clipping being the bandaid fix. In reality the cars could probably have 3600 MJ/h of fuel flow on the same amount of race fuel if super clipping was banned. Combine that with front axle regen and an MGU-H that caps harvesting at a level achievable by Audi. The resulting cars would be much less energy starved while burning the same amount of fuel over a race distance and putting out some 1100 peak horsepower with no turbo lag. The lack of turbo lag would make deploying the full 350kW in traction zones impossible due to the immense torque, which would extend the effective life of the battery and make overtake less wasteful. Combined with active aero these cars would be the fastest ever while burning much less fuel and making fewer compromises. They would also look really twitchy and be a handful to drive due to the extreme torque, which is fun to watch. If you really wanted to make sure the cars aren’t energy starved, you could probably even make the MGU-K power dropoff go slightly into the negative above certain speeds (linear harvesting load starting where deployment hits zero, peaking at -100 kW at 350 km/h or so), since 650 HP ICEs with active aero might actually be TOO fast in a straight line. This way super clipping wouldn’t actually be an issue. Since it doesn’t slow you down but limits firther acceleration.

If Audi were so concerned about the MGU-H being an advantage for established teams, give them additional budget allowance for it and give the non-WEC teams some extra budget for the front MGU. The only downside would be that the cars would come in at maybe 780-790kg instead of 768, which would be a small price to pay.

[MJL]

Hi, it is still not clear to me, how driver inputs are mapped to engine and mgu-k power? Is the throttle pedal directly connected to ICE power demand? And ECU is controlling where to deploy or harvest? Can the ICE continue producing 350kW in a corner without driver on throttle and use mgu-k as a brake that will effectively control current speed? MGU-K logic must run independently of the throttle to be able to harvest during full throttle and in the braking zones.

[Badger]

Bearman

“It’s a bit ridiculous to be honest,”. “To have that much delta in a button and to lose that much on the next straight. It's also very non-linear, so what you gain on the straight where you use the boost is a quarter as much as what you lose on the next straight.

"So unless you basically complete the move at the start of the straight - as in, you exit the corner, you complete the move - and then you harvest, harvest, harvest, the next straight they're gonna get you back.

“That's not racing, that's Formula E.”

This is a major issue in how they've designed the "overtake" function. Instead of a power increase early on the straight it's a "clipping reduction" button late on the straight. That is a very inefficient way of using that extra energy, late in the straight when drag is at its highest point you don't get much speed for your joules if you know what I mean.

How to solve it? Limit the output of the MGU-K to 200 kW on all laps, then make "overtake" into a power boost button which can be deployed at your discretion when within a second of a car ahead. Say you can deploy 300 kW with "overtake" instead of 200 kW normally. Then you could press it at the start of the straight and you'd close right up on the car ahead. By the time you hit the braking zone the relative speed of the overtaking and defending cars should be pretty equal which should make for great side by side racing.

This would also reduce "super-clipping".

[AR3-GP]

I agree that the overtake function is poorly devised. It should be the opposite as you say. Right now overtake mode is a trap designed to create yo-yo scenes, not any meaningful overtake. That is probably by design since yo-yo racing is good for the short format highlight reels that they say the younger generations prefer.

[wuzak]

The driver has no input into that decision to slow the car down in the fast corners. We're going to see cars slow similarly in Eau Rouge and Copse and it will be awful. The rules need to be adjusted to effectively end super clipping on any track.

The adjustment we need is a reduction in peak output large enough to remove the incentive to clip through fast corners or at the end of straights.

Super-clipping is a function of the allowed energy recovery and the ability of the cars to recover energy by other means (braking, part-throttle charging).

This also applies to lift-and-coast.

In Australia they were super-clipping and/or lift-and coasting for at least 10s per lap. Accounting for around half the allowed energy recovery.

Other tracks will see less of this because they have more braking zones, especially heavy braking zones, and more part-throttle application.

[wuzak]

Some parameters may be changed after the Chinese GP and before the Japanese GP.

Probably will restrict MGUK deployment power and increase super-clipping power.

Furthe changes may take place after that - they will likely have 1 month free after Japan to do something.

Increasing ICE power would be an option. This would be achieved through fuel flow.

More power from the ICE would also help recovery under part throttle.

https://www.the-race.com/formula-1/f1-c ... und-three/

[FittingMechanics]

The driver has no input into that decision to slow the car down in the fast corners. We're going to see cars slow similarly in Eau Rouge and Copse and it will be awful. The rules need to be adjusted to effectively end super clipping on any track.

The adjustment we need is a reduction in peak output large enough to remove the incentive to clip through fast corners or at the end of straights.

We still want super clipping. It works great in constant speed or gradually slowing turns like hairpins or tightening esses or the turns 1-4 complex in Shanghai.

The way to do that is by making the cars rocket out of corners slightly slower. In other words by reducing peak deployment. If you can’t get back ip to top speed as quickly, the incentive to clip in Eau Rouge or the chicane in Melbourne disappears.

The drivers might still lift and coast a bit into braking zones, but they won’t be super clipping for several seconds.

This will make the cars significantly slower, but it would be a decent bandaid fix for now, until they can work out a long term solution. That long term solution could be more fuel and/or maybe loosening some of the cost-cutting engine restrictions like 16:1 compression ratio, low boost pressure and variable trumpets ban. An unrealistic option could be an anti-lag system in the form of an electric turbo (but not MGU-H because it would be prevented from harvesting and would resort to the wastegate for overpressure)

Wouldn't a better solution be to limit the energy they can harvest? If you can harvest 5MJ by normal braking events, then you probably won't superclip or lift and harvest at all. It would make them slower, but your suggestion would as well.

I don't agree that this reduction in power would remove incentive to clip in Eau Rouge. They will super clip or lift and coast at all corners where they need to slow down (so they would normally brake or lift), especially if after that corner there is a straight. If you reduce the power, they would still super clip to get to the apex speed and then use their reduced power on the next straight.

This might be a problem even on tracks with lot of braking events - if there are two big straights one after another, you will want to fill up your battery as much as you can between the straights. So even if the lap overall is good for energy harvesting, they may want to superclip or lift and harvest for this second straight because they can't "carry over" the energy from the twisty bit onto this second straight.

Bearman

“It’s a bit ridiculous to be honest,”. “To have that much delta in a button and to lose that much on the next straight. It's also very non-linear, so what you gain on the straight where you use the boost is a quarter as much as what you lose on the next straight.

"So unless you basically complete the move at the start of the straight - as in, you exit the corner, you complete the move - and then you harvest, harvest, harvest, the next straight they're gonna get you back.

“That's not racing, that's Formula E.”

This is a major issue in how they've designed the "overtake" function. Instead of a power increase early on the straight it's a "clipping reduction" button late on the straight. That is a very inefficient way of using that extra energy, late in the straight when drag is at its highest point you don't get much speed for your joules if you know what I mean.

How to solve it? Limit the output of the MGU-K to 200 kW on all laps, then make "overtake" into a power boost button which can be deployed at your discretion when within a second of a car ahead. Say you can deploy 300 kW with "overtake" instead of 200 kW normally. Then you could press it at the start of the straight and you'd close right up on the car ahead. By the time you hit the braking zone the relative speed of the overtaking and defending cars should be pretty equal which should make for great side by side racing.

This would also reduce "super-clipping".

I agree that Overtake mode with the "power increase" at the end of straight is a useless feature. I don't think they are forced to use energy at these speeds, their PU will stop deploying to conserve battery well before they reach these speeds.

Suggestion of 300 kW overtake with 200 kW normal mode is a bad idea (IMO). It would be much more gimmicky, it would be hard to properly balance (just like DRS). It's effectively just like DRS, artificial benefit to the car behind. It's exactly the things people usually complain about.

Overtake mode increasing allowed recovery energy is a good idea but it suffers that to achieve this higher recharge the cars will need to lift and harvest or super clip more (as they can't brake more). Reason why I like this increase is that it is not free performance for the car behind (like DRS or higher power would be), it comes with downside as well. To utilize higher recharge you need to be able to recharge more.

[FittingMechanics]

Super-clipping is a function of the allowed energy recovery and the ability of the cars to recover energy by other means (braking, part-throttle charging).

This also applies to lift-and-coast.

In Australia they were super-clipping and/or lift-and coasting for at least 10s per lap. Accounting for around half the allowed energy recovery.

Other tracks will see less of this because they have more braking zones, especially heavy braking zones, and more part-throttle application.

Exactly. I am surprised people are calling for reduced deployment power. That is the not going to help. Melbourne with 200 kW limit probably produces the same levels of super clip/lift and coast but lower laptimes.

If they are able to recover, for example 4 MJ through braking in Australia and FIA reduced the allowed recharge to 4 MJ it would mean there would be no super clipping or lift and harvest (or at least much less). They would generate their energy through braking events. Exception would be two straights tied together where it could be more beneficial to recharge extra before the second straight. If that is deemed to be a problem just ban these superclip/lift and harvest recharge modes.

[gearboxtrouble]

The driver has no input into that decision to slow the car down in the fast corners. We're going to see cars slow similarly in Eau Rouge and Copse and it will be awful. The rules need to be adjusted to effectively end super clipping on any track.

The adjustment we need is a reduction in peak output large enough to remove the incentive to clip through fast corners or at the end of straights.

Super-clipping is a function of the allowed energy recovery and the ability of the cars to recover energy by other means (braking, part-throttle charging).

This also applies to lift-and-coast.

In Australia they were super-clipping and/or lift-and coasting for at least 10s per lap. Accounting for around half the allowed energy recovery.

Other tracks will see less of this because they have more braking zones, especially heavy braking zones, and more part-throttle application.

The problem is that the math will almost always make it more optimal to super clip in fast corners because they combine periods of full throttle with being nearly impossible to pass on (in the race). The most optimal deployment strategy is to use all of it in the acceleration zones that lead on to the active aero zones. This will maximize the impact of straight mode and vmax quickly allowing you to superclip more at the ends of those zones. We saw Mercedes doing just this in Melbourne quali and I expect most teams will start doing this going forward. Taken to its logical conclusion why would teams even develop their cars to maximize cornering performance in these fast corners if they know they're going to be harvesting through them? That's not the sort of car characteristics anyone should want in F1.

[wuzak]

The adjustment we need is a reduction in peak output large enough to remove the incentive to clip through fast corners or at the end of straights.

Super-clipping is a function of the allowed energy recovery and the ability of the cars to recover energy by other means (braking, part-throttle charging).

This also applies to lift-and-coast.

In Australia they were super-clipping and/or lift-and coasting for at least 10s per lap. Accounting for around half the allowed energy recovery.

Other tracks will see less of this because they have more braking zones, especially heavy braking zones, and more part-throttle application.

The problem is that the math will almost always make it more optimal to super clip in fast corners because they combine periods of full throttle for with being nearly impossible to pass on (in the race). The most optimal deployment strategy is to use all of it in the acceleration zones that lead on to the active aero zones. This will maximize the impact of straight mode and vmax quickly allowing you to superclip more at the ends of those zones. We saw Mercedes doing just this in Melbourne quali and I expect most teams will start doing this going forward. Taken to its logical conclusion why would teams even develop their cars to maximize cornering performance in these fast corners if they know they're going to be harvesting through them? That's not the sort of car characteristics anyone should want in F1.

Super-clipping is at the end of flat-out sections.

They didn't "super-clip" through the fast corners in Melbourne, but did at the end of low drag sections.

Into T1, T3, T6, T9, T11.

In some tracks they will likely forego deployment in fast corners to save energy for elsehwere. This is plain old clipping (derating).

Some people point out that Russell wasn't even braking for T9 because his super-clipping had brought him down to corner entry speed, or even below.


C5.12.6 Unless the electrical DC power of the ERS-K is negative, and subject to Article B7.2.1, the driver maximum power demand must not be reduced at any greater than the rates defined below:
a. 50kW in any 1s period at Competitions where the FIA determines that the power limited
distance exceeds 3500m. The vehicle fundamentals used for the calculation of the power
limited distance may be found in the document FIA-F1-DOC-034.
b. 100kW in any 1s period at all other Competitions.
Furthermore, the total power reduction is limited to a maximum of 600kW and the resulting electrical DC power of the ERS-K must remain above −250kW.

I would change the underlined sections to 350kW (or whatever deployment is reduced to) "must not be negative".

This would, likely, need to be accompanied by an increase in fuel flow to increase ICE power.

The FIA should tell the PU manufacturers to start testing a possible increase in fuel flow of 10-20% now, just in case that's the way they want to go.

[bananapeel23]

The driver has no input into that decision to slow the car down in the fast corners. We're going to see cars slow similarly in Eau Rouge and Copse and it will be awful. The rules need to be adjusted to effectively end super clipping on any track.

The adjustment we need is a reduction in peak output large enough to remove the incentive to clip through fast corners or at the end of straights.

Super-clipping is a function of the allowed energy recovery and the ability of the cars to recover energy by other means (braking, part-throttle charging).

This also applies to lift-and-coast.

In Australia they were super-clipping and/or lift-and coasting for at least 10s per lap. Accounting for around half the allowed energy recovery.

Other tracks will see less of this because they have more braking zones, especially heavy braking zones, and more part-throttle application.

Detune the MGU-K enough and it simply isn't worth it to give up the speed for the acceleration boost out of the following corner, even if you don't reach the harvesting allowance.

In other news. I just tried to look at how much energy the MGU-H used to harvest. I'm astonished that they chose to cut it. I can't find any actual numbers, but it was apparently doing at least 2/3 of the energy harvesting in the last regulation set. Assuming the MGU-K was harvesting the maximum permitted 2MJ per lap, the MGU-H must've been harvesting at least another 4 MJ. Cut fuel flow to 70 kg/hr and it would still be harvesting ~2.8 MJ per lap.

That would have gone a long way in this regulation set, providing 1/3 of the maximum permitted regen. If cars are currently super clipping or doing LiCo for 15 seconds per lap in order to reach the full 8.5 MJ of harvesting, the MGU-H would cut that to just over 5 seconds. In reality it would likely do more than that since the MGU-H is most effective on energy starved tracks like Monza or Albert Park.