Ferrari Power Unit Hardware & Software

[gruntguru]

I wonder if Ferrari have found something extra if the electric S/C mode is the differentiator. The one unorthodox thing about the electric SC mode is that the turbine is still linked (whereas a 'normal' electric SC wouldn't have a turbine at all), and thus becomes a fan/vacuum pump for the exhaust. This would ostensibly remove any backpressure and maybe even create a more ideal vacuum for the combustion chamber?

Radial inflow turbines don't do that when motored. If anything they tend to flow in reverse to the normal direction.

Mixed flow (radial/axial) turbines have less tendency to oppose flow when motored.

Either way, it is almost certain that blowdown pulses are still being forced through the turbine with wastegates open, to capture some valuable energy during "electric supercharger" mode.

[bill shoe]

If you have a Merc and a Ferrari accelerating through 150 km and exceeding the traction limited speed, both cars will be deploying the maximum 120 kW to the K. If the Ferrari has an additional 40 hp available it has to be either a crankshaft power advantage or electric supercharger deployment (which also ends up at the crankshaft). Electric supercharger mode is worth less than 3 bar in BMEP terms (exhaust stroke pumping work reduction). 40 hp is about 1 bar BMEP at 11,000 rpm which is approximately the gain I would expect from electric supercharging.

So rumored Ferrari gain of 38 hp is roughly what they'd get if they could run electric supercharging on every straight during the race instead of just for one lap in qualy by sucking excess (unsustainable) power from ES.

A vague but reasonable hypothesis is that Ferrari is able to use the dual-battery setup to harvest more energy from the H regardless of the K situation.

During braking, the K obviously harvests kinetic energy from the car with the usual 120 kW power limit.

Can the H harvest significant energy during braking to subsequently drive the H on the next straight? This would allow electric supercharging after every significant braking zone. H harvesting is not limited to 120 kW or any other limit. Sounds good in theory. I think this is likely key to Ferrari's trick. But the big question-- How to harvest car's kinetic energy during braking with an electric device on your turbo-compressor?

[roon]

Can the H harvest significant energy during braking to subsequently drive the H on the next straight? This would allow electric supercharging after every significant braking zone. H harvesting is not limited to 120 kW or any other limit. Sounds good in theory. I think this is likely key to Ferrari's trick. But the big question-- How to harvest car's kinetic energy during braking with an electric device on your turbo-compressor?

That's precisely what is suggested with Honda's 'extra harvest.' Send any K regen during braking beyond 2MJ to the H, and then on to ES. It is speculated the H consumes ~60kW, but in this schema it would be sized to relay 120kW between the ES and the MGUK. This should help balance the braking and on-throttle energy requirements.

[bill shoe]

That's precisely what is suggested with Honda's 'extra harvest.' Send any K regen during braking beyond 2MJ to the H, and then on to ES. It is speculated the H consumes ~60kW, but in this schema it would be sized to relay 120kW between the ES and the MGUK. This should help balance the braking and on-throttle energy requirements.

But this still limits total kinetic energy harvest to the 120 kW power limit of the K. The Honda trick just gets around the K's 2MJ energy limit, but not around the K's 120 kW power limit. Likely every manufacturer understands this trick. But if you could somehow harvest kinetic energy with the H, then you are getting around the K's power limit and that would be special.

I think. I base this on gruntguru's definitive post p. 123: "the maximum POWER that can go between the MGU-K and everything else is limited. (to 120 KW)"

[chaoticflounder]

Can the H harvest significant energy during braking to subsequently drive the H on the next straight? This would allow electric supercharging after every significant braking zone. H harvesting is not limited to 120 kW or any other limit. Sounds good in theory. I think this is likely key to Ferrari's trick. But the big question-- How to harvest car's kinetic energy during braking with an electric device on your turbo-compressor?

I feel like this would be somewhat straight forward.

You would use the expansion of the gases as the cylinder forces them out on the exhaust stroke for useful work in the MGU-H.

The MGU-K is on the crankshaft, you could divert more energy from going into the brakes as heat and increase the crankshaft draw by requiring the 120 kW draw to the MGU-K and also requiring the crankshaft to compress it's working fluid, air, against a generating exhaust turbine in the MGU-H. Fairly similar to engine braking in a gasoline, or petrol, engine when you lift your foot off the throttle you are closing the throttle body and the engine is pulling a vacuum on the intake stroke in the four stroke cycle which is performing measurable work to slow the vehicle down. This concept would work on the exhaust stroke with compressing the exhaust gases against the turbine.

[wuzak]

I've heard different interview with Toto of Merc, and he said Merc and Ferrari are even on the straights up to the point where accel stops being traction-limited, and from then on Ferrari easily pulls away. Traction-limited stops at ~ 150 kph. This is a very different scenario than the cars separating at 250 kph like the Andy Cowell interview.

A difference that started at 150 kph would imply higher peak power. A difference that started at 250 kph would imply either less aero drag or a longer sustaining of existing peak power enabled by, um, well, mechanism unclear as has been discussed in this thread extensively.

Longer sustain of peak power could come from a couple of things.
More SOC to play with. So all the ways that harvest come in to play, plus any ability to maximise the use of the batteries.
Lower MGU-H power demand. Perhaps the turbine makes more power from blowdown. Perhaps the compressor is more efficient.
If they have found a way to get more from the MGU-H, that would likely affect both of these and allow longer deployment(s).
I don’t think we can discount chassis and aero either. If the customer teams don’t show this behaviour it’s possible that their deployment schedules are different, not because of default programming but because of behaviours learned during practise. And those behaviours will be dependent on chassis and aero performance.

If you have a Merc and a Ferrari accelerating through 150 km and exceeding the traction limited speed, both cars will be deploying the maximum 120 kW to the K. If the Ferrari has an additional 40 hp available it has to be either a crankshaft power advantage or electric supercharger deployment (which also ends up at the crankshaft). Electric supercharger mode is worth less than 3 bar in BMEP terms (exhaust stroke pumping work reduction). 40 hp is about 1 bar BMEP at 11,000 rpm which is approximately the gain I would expect from electric supercharging.

Is it possible that Ferrari are partially driving the turbo with the MGUH and exhasut gases, opening one wastegate to increase power at the crankshaft, but keeping the other closed so as to minimise the amount of power the MGUH needs to draw to drive the turbo?

To even this out the system would have to alternate which bank of three drives the turbo and the other vents through the wastegate.

The alternating of which bank does what has a precedent - the V10 in the Ferrari F1-2000 used a system where the throttles of one bank would advance ahead of the other, to improve driveability. This would alternate to keep wear in the engine even (from Peter Wright's Ferrari Formula 1).

[wuzak]

Another thought is that the posters here have described the wastegates as either fully open or fully closed. Is it possible that under some circumstances they open the wastegates partially?

[subcritical71]

Another thought is that the posters here have described the wastegates as either fully open or fully closed. Is it possible that under some circumstances they open the wastegates partially?

For some reason I was under the impression that it could be partially open, so your idea would work.

[roon]

But if you could somehow harvest kinetic energy with the H, then you are getting around the K's power limit and that would be special.

Engine braking.

[saviour stivala]

Another thought is that the posters here have described the wastegates as either fully open or fully closed. Is it possible that under some circumstances they open the wastegates partially?

For some reason I was under the impression that it could be partially open, so your idea would work.

Yes they do open partially and momentarily. they do so at start of H spinning-up the turbo to relieve start-up load on the H. sort of the same prosses when an air compressor sitting on a pressurized air tank re-starts. in normal race mode a leaking or open or partially open waste gate is considered as one of the most inefficient things. as I explained elsewhere, the modern day turbo + H design as used in F1 actually doesn't need either waste gate/s or pop-up valve, but they are incorporated into the design as a safety measure, and because of the use of waste gates as a safety measure along came the development of the free load mode, where the turbo is operated in electric supercharging mode at full fueling, with waste gates fully open and exhaust gases bypassing the turbine scroll, and with H and K sharing ES power. and ones again, the K and H is either motoring or generating, they cant do both things at the same time.

[saviour stivala]

Longer sustain of peak power could come from a couple of things.
More SOC to play with. So all the ways that harvest come in to play, plus any ability to maximise the use of the batteries.
Lower MGU-H power demand. Perhaps the turbine makes more power from blowdown. Perhaps the compressor is more efficient.
If they have found a way to get more from the MGU-H, that would likely affect both of these and allow longer deployment(s).
I don’t think we can discount chassis and aero either. If the customer teams don’t show this behaviour it’s possible that their deployment schedules are different, not because of default programming but because of behaviours learned during practise. And those behaviours will be dependent on chassis and aero performance.

If you have a Merc and a Ferrari accelerating through 150 km and exceeding the traction limited speed, both cars will be deploying the maximum 120 kW to the K. If the Ferrari has an additional 40 hp available it has to be either a crankshaft power advantage or electric supercharger deployment (which also ends up at the crankshaft). Electric supercharger mode is worth less than 3 bar in BMEP terms (exhaust stroke pumping work reduction). 40 hp is about 1 bar BMEP at 11,000 rpm which is approximately the gain I would expect from electric supercharging.

Is it possible that Ferrari are partially driving the turbo with the MGUH and exhasut gases, opening one wastegate to increase power at the crankshaft, but keeping the other closed so as to minimise the amount of power the MGUH needs to draw to drive the turbo?

To even this out the system would have to alternate which bank of three drives the turbo and the other vents through the wastegate.

The alternating of which bank does what has a precedent - the V10 in the Ferrari F1-2000 used a system where the throttles of one bank would advance ahead of the other, to improve driveability. This would alternate to keep wear in the engine even (from Peter Wright's Ferrari Formula 1).

Yes the FERRARI F1-2000 3L V10 used alternating throttles openings between banks on fast accelaration to minimize the Gyroscopic effect on the car. to this effect they were also constantly experimenting with verious firing orders.

[Tommy Cookers]

the K and the H each could in principle motor and generate simultaneously (the rules allow such a design configuration)

and in the conventional design presumably in use can each readily motor and generate pseudo-simultaneously
(ie 'interleaved' in time as the Honda 'extra harvest' technique is)
if such simultaneity-emulating behaviour is desired

[henry]

I made a very simple simulation. Two cars, same mass, same drag.

Car A 700kW accelerates From 150kph to 320kph in 5.44 secs over 389m

Car B 670kW accelerates from 150kph to 316kph in 5.50 secs over 389m

Even with a 30kW advantage it would take 8 such events to gain 0.5 seconds.

I also looked at 250kph as the lower speed, which it has been suggested is Andy Cowell’s number. As might be expected the time differences are very small.

I think to gain 0.5 seconds over 3 or 4 straights probably needs more than just a 30kW advantage.

My model is very crude, I’ve not allowed transmission loss for instance, but I think it’s in the right ballpark.

[henry]

For those contemplating using the MGU-H/turbo assembly as a kinetic energy store.

The Honda extra harvest numbers suggest that at 20hz and symmetrical 120kW in 120kW out (for 60kW average) the speed might fluctuate around 5000 rpm, approx 5%.

I used an inertia value of 6.4E-4 kgm2.

[SiLo]

Looking at the comparison between Merc and Ferrari (Bottas and Vettel) in Germany, it mostly looked like a a lot of time gained under breaking to be honest, rather than a whole chunk gained on the straight.