2025/2026 Hybrid Powerunit speculation

[wuzak]

Depends on what the goal is.

Maximum efficiency/power is the goal.

If the heat lost to coolant is reduced, there is more energy left to turn the crankshaft.

I do not fully understand the statement regarding heat loss. The coolant system in an internal combustion engine is designed to prevent overheating by maintaining the engine below a maximum operating temperature. While I understand that heat is a form of energy, it is unclear to me how residual heat within the cylinder contributes to power generation. In practice, the intake air is cooled after leaving the compressor and prior to entering the cylinder. Additionally, one of the motivations for using direct injection is its ability to maintain lower fuel temperatures. Lower temperatures of the intake charge and engine components increase charge density, improve compressibility, and enable greater expansion during combustion. Maybe you mean fuel that is burned that turns into heat is energy lost ?

There is a set amount of fual energy supplied to the engine (3,000MJ/h).

Some will go to driving the crankshaft.
Some will exit through the exhaust.
Some will exit through heat transfer to the coolant and oil.

You want to maximise the first and minimise the last.

You want to minimise the second too, but there is some energy recovery with the turbo, transferring the energy from the exhaust to compressing the intake air.

Lower heat loss through the coolant has the additional benefit of requiring smaller radiators, and similar for oil and oil coolers.

Cooling the intake air is also wasting energy, but there are limits to how hot the air can be without causing detonation.

[BassVirolla]

Maximum efficiency/power is the goal.

If the heat lost to coolant is reduced, there is more energy left to turn the crankshaft.

I do not fully understand the statement regarding heat loss. The coolant system in an internal combustion engine is designed to prevent overheating by maintaining the engine below a maximum operating temperature. While I understand that heat is a form of energy, it is unclear to me how residual heat within the cylinder contributes to power generation. In practice, the intake air is cooled after leaving the compressor and prior to entering the cylinder. Additionally, one of the motivations for using direct injection is its ability to maintain lower fuel temperatures. Lower temperatures of the intake charge and engine components increase charge density, improve compressibility, and enable greater expansion during combustion. Maybe you mean fuel that is burned that turns into heat is energy lost ?

There is a set amount of fual energy supplied to the engine (3,000MJ/h).

Some will go to driving the crankshaft.
Some will exit through the exhaust.
Some will exit through heat transfer to the coolant and oil.

You want to maximise the first and minimise the last.

You want to minimise the second too, but there is some energy recovery with the turbo, transferring the energy from the exhaust to compressing the intake air.

Lower heat loss through the coolant has the additional benefit of requiring smaller radiators, and similar for oil and oil coolers.

Cooling the intake air is also wasting energy, but there are limits to how hot the air can be without causing detonation.

Plus hotter air requires more energy to fit a determinated mass into a volume.

[vorticism]

Depends on what the goal is.

Maximum efficiency/power is the goal.

If the heat lost to coolant is reduced, there is more energy left to turn the crankshaft.

Yet aluminium is almost always chosen for head construction. Did thousands of engineers forget first principals? Or were there other goals beyond what you’re calling “the” goal?

The post I had replied to claimed that steel heads are more prone to knock. Fair enough. That’s a common understanding of ferrous alloy heads, that regions of their CCs get too hot and cause autoignition. I offered him solutions for that as it would apply to a Formula One engine under these regulations. Forge the head to achieve sufficiently thin walls (to manage convection latency), machine features, weld a cam carrier onto it (must be one monolithic component and weldments seem to be permitted on the head), machine again, etc. Dial in the coolant flow rate to achieve CC wall temperatures similar to Al. The goal would be to end up with a head that has similar mass and structural properties which provides a more familiar, Al-like CC surface temperature, with the added benefit of the durability of steel (fatigue, surviving detonation wave fronts). Heat rejection would be identical to an Al head. If you need to solve a real world problem (f.e. make a Fe-alloy head which exhibits Al CC temps within the same performance domain), you have to make design choices that will solve it. Chanting first principals axioms while hoping for the best won’t get you very far.

If the design intent from the outset is to use less cooling and allow hotter CC surface temps, then your insistence makes sense and I agree with you, but that wasn’t the intention of my post. A combustion concept which takes advantage of an intentionally hotter steel CC surface does open up interesting design paths f.e. potentially eschewing water coolant for oil or air. The thermal limits of inserts and “dismountable components” (spark plug, injector, pressure sensor, poppets) will need to be accommodated.

[wuzak]

Depends on what the goal is.

Maximum efficiency/power is the goal.

If the heat lost to coolant is reduced, there is more energy left to turn the crankshaft.

Yet aluminium is almost always chosen for head construction. Did thousands of engineers forget first principals? Or were there other goals beyond what you’re calling “the” goal?

"Thousands of engineers" had to work around low octane fuels and less advanced combustion technologies.

And work within a budget - not just for development, but for production.

The post I had replied to claimed that steel heads are more prone to knock. Fair enough. That’s a common understanding of ferrous alloy heads, that regions of their CCs get too hot and cause autoignition. I offered him solutions for that as it would apply to a Formula One engine under these regulations. Forge the head to achieve sufficiently thin walls (to manage convection latency), machine features, weld a cam carrier onto it (must be one monolithic component and weldments seem to be permitted on the head), machine again, etc. Dial in the coolant flow rate to achieve CC wall temperatures similar to Al. The goal would be to end up with a head that has similar mass and structural properties which provides a more familiar, Al-like CC surface temperature, with the added benefit of the durability of steel (fatigue, surviving detonation wave fronts). Heat rejection would be identical to an Al head. If you need to solve a real world problem (f.e. make a Fe-alloy head which exhibits Al CC temps within the same performance domain), you have to make design choices that will solve it. Chanting first principals axioms while hoping for the best won’t get you very far.

I would say that iron-alloy heads tended to be older designs. Not really past the 1960s and 1970s, or maybe the 1980s.

And tended to be cast iron, rather than forged and machined steel.

The same could be said of pistons.

They went to aluminium a long time ago, even when the block and head(s) were cast iron. And a lot of effort went into finding aluminium alloys and coatings that allowed aluminium pistons to run in aluminium blocks, without using steel sleeves. Though many (most?) manufacturers opted for the steel sleeve in aluminium block solution.

But the 2026 F1 rules require a one of three specific steel alloys for the piston.

It may be for durability, but it also could be for other reasons.

The rumours are that the steel cylinder head for the Ferrari would be 3d printed. That would enable them to optimise where the material is and reduce weight.

If the design intent from the outset is to use less cooling and allow hotter CC surface temps, then your insistence makes sense and I agree with you, but that wasn’t the intention of my post. A combustion concept which takes advantage of an intentionally hotter steel CC surface does open up interesting design paths f.e. potentially eschewing water coolant for oil or air. The thermal limits of inserts and “dismountable components” (spark plug, injector, pressure sensor, poppets) will need to be accommodated.

Less cooling = less drag.

They will opt for the least amount of cooling they can get away with regardless of engine type.

Sometimes they get it wrong, or the air temperature is higer than expected, and have to run lower power to compensate.

I doubt air or oil cooling are options, or desired. Air cooling would be difficult to achieve the required air flow around the heads.

The coolant system must be fitted with a pressure relief valve with a maximum setting of 3.75 bar gauge. That would give a boiling point of around 150°C for water.

That would, I think, exclude air cooling.

[Tommy Cookers]

having thin steel combustion chambers could liberate design from the constraints (eg 'CR-related) of ....
valve seats separate from the head
heads separate from the cylinders


Stewart Tresilian chose steel heads for a 'small & light' but very high power aero engine for R-R
didn't the late Bristol Hercules have (or need) steel heads ?
bronze heads had outstanding conductivity and hot-strength (but were heavy)

[gruntguru]

There is a set amount of fual energy supplied to the engine (3,000MJ/h).

Some will go to driving the crankshaft.
Some will exit through the exhaust.
Some will exit through heat transfer to the coolant and oil.

You want to maximise the first and minimise the last.

You want to minimise the second too, but there is some energy recovery with the turbo, transferring the energy from the exhaust to compressing the intake air.

Lower heat loss through the coolant has the additional benefit of requiring smaller radiators, and similar for oil and oil coolers.

Cooling the intake air is also wasting energy, but there are limits to how hot the air can be without causing detonation.

Plus hotter air requires more energy to fit a determined mass into a volume.

Yes, the compressor work increases but the extra work is relatively cheap exhaust energy. OTOH the higher intake pressure helps drive the crankshaft during the intake stroke - basically returning some exhaust heat energy into useful crankshaft energy.

[gruntguru]

"Drag Power" = aero drag
"Wheel Power" = power applied through the tyre (PU or Brakes)
"Accel Power" = "Wheel Power" - "Drag Power" This is what accelerates the car.

Nice analysis Boris. I decided to explore the same data in some different directions.
- I used the speed and acceleration values with the same mass (768 kg) to calculate the acceleration power
- Assumed a drag-limited top speed of 315 km/hr and power of 630 KW to calculate wheel power lost to drag

Some interesting numbers arising from the above.
Average positive power to wheels - 375 KW (60% of max possible 630 KW)
Average negative power to wheels (tot regen available) - 89 KW (7.3 MJ/lap)
Average negative power to wheels (regen available if limited to 150 KW) - 23 KW (1.9 MJ/lap)
Average negative power to wheels (regen available if limited to 350 KW) - 45 KW (3.7 MJ/lap)
(assumes 4 wheel regen)
Open to suggestions on other interesting outputs.

[wuzak]

Some interesting numbers arising from the above.
Average positive power to wheels - 375 KW (60% of max possible 630 KW)
Average negative power to wheels (tot regen available) - 89 KW (7.3 MJ/lap)
Average negative power to wheels (regen available if limited to 150 KW) - 23 KW (1.9 MJ/lap)
Average negative power to wheels (regen available if limited to 350 KW) - 45 KW (3.7 MJ/lap)
(assumes 4 wheel regen)
Open to suggestions on other interesting outputs.

What does the 89kW figure represent?

Braking and lift-and-coast?

[gruntguru]

Braking and regen only. It represents deceleration force at the tyre patch. There will still be deceleration (due to aero) even when this value is zero. Example at ~48s this value is zero (lower graph, purple curve) yet there is about 5,800 N of aero drag (470 KW) decelerating the car at >1G.

89 KW is the average over the lap. The largest deceleration power was 1,797 KW.

[wuzak]

Braking and regen only. It represents deceleration force at the tyre patch. There will still be deceleration (due to aero) even when this value is zero. Example at ~48s this value is zero (lower graph, purple curve) yet there is about 5,800 N of aero drag (470 KW) decelerating the car at >1G.

89 KW is the average over the lap. The largest deceleration power was 1,797 KW.

So, the power to slow the car, but not what is available to the MGUK?

[gruntguru]

No. I didn't attempt to separate energy types. Just worked out the total energy going in and out of the drivetrain. For example:
- the negative energy to the wheels could be any mix of - brakes, mguk, ICE braking
- the positive energy to the wheels could be any mix of - mguk, ICE

I only got as far as breaking down how much negative energy might be available for two values of MGU-K power limit.

Happy to hear any suggestions on how to apply this info to try and analyse the breakdown further for 2025 or 2026 drivetrains.

[venkyhere]

Happy to hear any suggestions on how to apply this info to try and analyse the breakdown further for 2025 or 2026 drivetrains.

Similar to the suggestion I made to @BorisTheBlade, the most interesting thing would be a plot/analysis of a race lap from 2025 (from a stint with laptimes having very low standard deviation) so that we can discount 'quali-like' full deployment of battery. And we can 'save' this lap analysis as reference for later, for the same race lap from 2026, and then compare and find out exactly what the differences are. I am most interested to know what the 'floor'/base-level charge that has to remain in the battery (unusable) over a target laptime stint, and how it's different b/w 2025 and 2026.

[gruntguru]

Some interesting numbers arising from the above.
Average positive power to wheels - 375 KW (60% of max possible 630 KW)
Average negative power to wheels (tot regen available) - 89 KW (7.3 MJ/lap)
Average negative power to wheels (regen available if limited to 150 KW) - 23 KW (1.9 MJ/lap)
Average negative power to wheels (regen available if limited to 350 KW) - 45 KW (3.7 MJ/lap)
(assumes 4 wheel regen)

If anywhere near accurate, these numbers give us some useful benchmarks.

1. If there is zero regen available, this particular lap needs 375 KW average output from the ICE. This means a 2026 ICE with 375 KW peak would have to run at max power for the whole lap, plus storage and retrieval of energy would need to be 100% efficient (impossible of course).
2. Available braking regen is only another 45 KW, assuming - sufficient traction at the rear, 100% round-trip efficiency, etc so clearly significantly less than 45 KW.

Combining 1. and 2. above suggests that the 2026 PU would struggle to provide the energy used on Lando's lap using the 2025 chassis - although it is not that far off.

The 2026 chassis will provide some benefits however.
Total energy for Lando's lap was 376 KW x 82.17 sec = 30.8 MJ
Of that total, 290 KW (23.8 MJ) went to overcoming aero drag leaving only 86 KW (7 MJ) to accelerate the car. Clearly, a small reduction in drag will make a big difference to the energy balance.

[gruntguru]

Happy to hear any suggestions on how to apply this info to try and analyse the breakdown further for 2025 or 2026 drivetrains.

Similar to the suggestion I made to @BorisTheBlade, the most interesting thing would be a plot/analysis of a race lap from 2025 (from a stint with laptimes having very low standard deviation) so that we can discount 'quali-like' full deployment of battery. And we can 'save' this lap analysis as reference for later, for the same race lap from 2026, and then compare and find out exactly what the differences are. I am most interested to know what the 'floor'/base-level charge that has to remain in the battery (unusable) over a target laptime stint, and how it's different b/w 2025 and 2026.

If someone wants to pick a lap from the same race (probably another Lando lap), I will run the numbers. I doubt a second or two of lap-time will change the numbers much - less than the uncertainty inherent in my assumptions and calculations.

[michl420]

I think we are writting about 2 different things. My intention was to predict the ICE (only) output (work, not peak power) next year in a race distance.
I use assumed peak power (600 and 400kw) and full trottle/lap (=~ ICE operates with full power).
In THEORY it must run 50 % longer to have the same output.
The practical answer to this will only bring the fuel consumption next year.