2025/2026 Hybrid Powerunit speculation

[FittingMechanics]

How difficult would it be to change the 50/50 split to a 60/40 immediately and then a 70/30 split after the summer brake ?

If you just reduce the MGU-K power then it is not hard at all. They can do this by introducing a software limit (I think FIA is allowed to do that already)

But, reduction in power would only mean they use the MGU-K for longer. Most of the complaints would still apply, they would still harvest in a similar way, they would still be able to yoyo, they would be even slower, etc.

Even if you increase ICE power, all of these still apply.

[wuzak]

How difficult would it be to change the 50/50 split to a 60/40 immediately and then a 70/30 split after the summer brake ?

If you just reduce the MGU-K power then it is not hard at all. They can do this by introducing a software limit (I think FIA is allowed to do that already)

But, reduction in power would only mean they use the MGU-K for longer. Most of the complaints would still apply, they would still harvest in a similar way, they would still be able to yoyo, they would be even slower, etc.

Even if you increase ICE power, all of these still apply.

If ICE power is increased using inccreased fuel flow across the rpm range, then there will be more opprotunity for part throttle recovery - more time where ICE power exceeds the driver's demands and potentially more difference between ICE power and demand.

[FittingMechanics]

How difficult would it be to change the 50/50 split to a 60/40 immediately and then a 70/30 split after the summer brake ?

If you just reduce the MGU-K power then it is not hard at all. They can do this by introducing a software limit (I think FIA is allowed to do that already)

But, reduction in power would only mean they use the MGU-K for longer. Most of the complaints would still apply, they would still harvest in a similar way, they would still be able to yoyo, they would be even slower, etc.

Even if you increase ICE power, all of these still apply.

If ICE power is increased using inccreased fuel flow across the rpm range, then there will be more opprotunity for part throttle recovery - more time where ICE power exceeds the driver's demands and potentially more difference between ICE power and demand.

I'm no expert but I think increasing fuel limit has massive ramifications for the design of the ICE. I don't think this is a realistic change mid season. Could be wrong though *shrugs*.

[bananapeel23]

How are teams modeling harvesting/deployment strategies on high energy tracks? It must be really difficult to figure out how to best deploy and harvest in Shanghai. Harvest too aggressively and you will have harvested the full 9MJ going into, or part way through the hairpin. That would cost you both braking power and acceleration out of the hairpin, at best costing you a lot of time, at worst causing lockups or cars going very wide.

Obviously you will want to harvest every little bit of the 9MJ available, but you also don't want to harvest or deploy in inefficient locations, but the 4 MJ battery must make modelling deployment and harvesting extremely difficult, while effective distribution of energy could be worth seconds per lap, even if both cars get the full 9 MJ.

Are teams brute forcing it in the simulator? Are they manually calculating where to use the energy? Are they training neural nets on simplified versions of the track to find candidates, then testing it in the sim? Does anyone know?

[Dr. Acula]

I'm no expert but I think increasing fuel limit has massive ramifications for the design of the ICE. I don't think this is a realistic change mid season. Could be wrong though *shrugs*.

Well, a higher energy flow also means you need more fuel to begin with. The question here is, if the fuel tanks they have fitted in the car now, can even hold significantly more fuel for the race. I don't think so, because why would you design the tank bigger than it absolutly has to be? And as the fuel tank is fitted to the chassis, basically that could mean that a new chassis would be required.

[Martin Keene]

How difficult would it be to change the 50/50 split to a 60/40 immediately and then a 70/30 split after the summer brake ?

If you just reduce the MGU-K power then it is not hard at all. They can do this by introducing a software limit (I think FIA is allowed to do that already)

But, reduction in power would only mean they use the MGU-K for longer. Most of the complaints would still apply, they would still harvest in a similar way, they would still be able to yoyo, they would be even slower, etc.

Even if you increase ICE power, all of these still apply.

If ICE power is increased using inccreased fuel flow across the rpm range, then there will be more opprotunity for part throttle recovery - more time where ICE power exceeds the driver's demands and potentially more difference between ICE power and demand.

Whilst in theory increasing fuel flow is easy, in reality, it is a lot harder. Teams will have based the fuel tank size around the fuel flow limit vs the longest race/fuel consumption, so they likely do not have a fuel tank big enough to support an increase in fuel flow. That is probably the easiest obstacle to overcome, more of a challenge is the duty cycle of the engine will have been defined around that fuel flow limit and then the engine designed to that duty cycle.

Increasing fuel flow will increase the duty cycle so will shorten engine life and will require some redesign of the engine. How much of a redesign, is anyone's guess.

[AR3-GP]

Obviously you will want to harvest every little bit of the 9MJ available, but you also don't want to harvest or deploy in inefficient locations, but the 4 MJ battery must make modelling deployment and harvesting extremely difficult, while effective distribution of energy could be worth seconds per lap, even if both cars get the full 9 MJ.

Are teams brute forcing it in the simulator? Are they manually calculating where to use the energy? Are they training neural nets on simplified versions of the track to find candidates, then testing it in the sim? Does anyone know?

Its a multivariable optimization problem. So they will probably try a few different standard approaches in the field ( Neural networks, genetic algorithm, steepest descent algorithm, etc.)

They are essentially just building a map of the distribution of super clipping, brake harvesting, partial throttle harvesting, and lico which produces the fastest laptime. There are probably multiple very close solutions.

[venkyhere]

Question to all experts :

We saw how in the Australian GP, almost everyone was complaining about 'no charge left' in the battery at the end of the formation lap, before the lights. I reckon this is because the Energy Management S/W is unable to distinguish between a race lap and a formation lap, because it has no way to distinguish A/B pedal input made in the formation lap v/s a race lap, so it will deploy and recharge as it pleases, reacting to the pedals.
Now, can we think that inlap into pits and outlap from pits are also similar (no way to differentiate from race lap) ?
What about safety car restarts ?

Before, we have always heard team radios where the driver is asked to switch 'modes' whenever there is a safety car or at the end of formation laps or in the middle of the pitlane. Is that not the case in 2026 ? Can't many separate algorithms exist, to distinguish between race/formation/pit-in/pit-out laps, and the driver can choose to switch with some buttons ? How about existence of different programs (or subroutines) for Quali and Race, atleast ? OR is it a case of 'learned data' evaporating if modes are switched ?

Does anyone have any idea (even crude) about the S/W architecture (any insider info) ? If they can make the S/W adapt to each track, I am puzzled what's the great difficulty in distinguishing a formation lap from a race lap. The formation lap draining battery is pretty stupid, IMHO ; has to be one of the 'low hanging fruits'.

Kindly educate me.

[AR3-GP]

Question to all experts :

We saw how in the Australian GP, almost everyone was complaining about 'no charge left' in the battery at the end of the formation lap, before the lights. I reckon this is because the Energy Management S/W is unable to distinguish between a race lap and a formation lap, because it has no way to distinguish A/B pedal input made in the formation lap v/s a race lap, so it will deploy and recharge as it pleases, reacting to the pedals.

IT's not this. It's just that warmup lap has massive energy consumption (wheel spin to heat rear tires is 100% unrecoverable), but little opportunity for energy recovery (they push the brake bias forwards to warm the front tires which means MGU-K doesn't get much work at the rear). Warmup lap is not at racing speed so the fuel flow rates are very limited at part throttle. It just makes it extremely easy to use up a lot of the battery, but have little way to recharge it near the end of the formation lap. When drivers who are doing burnouts just before their grid spots they simply can't recharge to 100%. Warmup lap procedure is the antithesis of battery top up.

[Farnborough]

I wonder if we'll see a pole sitter change warm lap strategy:-

Perhaps a full out race start, 1st few corners at power brake overlap aggressively, then into full bore style for remaining lap distance with max regeneration, big arrival at grid with huge braking regeneration, to line up and wait with more or less full battery.

Somebody would likely give it consideration, especially at a long run to 1st corner track layout. China/Mexico possibly.