Ferrari Power Unit Hardware & Software

[henry]

As far as I know electrical power deployment can only be deployed by the “K” and not by the “PU”. Deploying 6-8 MJ per lap of electrical power, 4 MJ of which must be from the “ES” and the rest from the”H”, are conservative numbers compared to other numbers that have been banded about. However assuming that is correct, and on that lap the driver has managed to harvest by braking what is permitted, what happens on the next lap with a 50% charged “ES”?. If it is not qualifying, what sort of race strategy is that?. We had seen what happened with one of the red bulls when the driver said he had no deployment power left and he was told that he used it all.
I honestly don’t mean any disrespect to anybody, disagreeing does not mean that. I consider my knowledge as being able to just scratch the top of F1 matters, This being confirmed by being told that I risk being eaten up for breakfast by the more knowledgeable folks on here.

There are two very different ways the PUs and ES are used.

In qualification they start the qualification lap with 4MJ available in the ES, and they end the lap with zero. Using your assumptions they start at 8 and go down to 4.

In the race they try to start and end the lap with the same SOC. Let’s say 2MJ for the sake of this discussion.

In the case you quoted with Ricciardo a scenario might be he started the lap with 2MJ but ended it with 1MJ. He used a little more than he could recover. He was driving harder to try to catch or overtake. Now on the next lap(s) he needs to use less than he can recover to echarge the ES.

At Baku we saw Ricciardo spend several laps at a time charging his ES falling back from Verstappen a little, before launching an attack in which he intended to use most, if not all, of his ES to catch and pass. Once past he would have driven a little slower and defensively while the ES recharged.

You must now ask ourselves, if the ES can only be charged by braking how did Ricciardo brake much less than Verstappen without running off the track?

My answer is that he didn’t and that the MGU-H was used at the end of straights to send energy to the ES and not to the MGU-K.

Edit: fixed an auto correct although charming the ES is a nice idea

[Big Mangalhit]

[
Whilst it is insightful to think of them deploying the MGU-K all the time they are at WOT I don’t think that’s what they do. It is more laptime efficient to run electric supercharger at the beginning of the straight, switch to ICE plus MGU-K and finally switch off the MGU-K and charge the ES, they might even use the MGU-K as well, sort of lift and coast plus. This of course makes it difficult to estimate what the total,energy flow is.

I think this is the key of these hybrids. Power is much more useful on low speeds. A clio can drive pretty well between 0-100km/h with 50hp. But a Veyron needs an extra 100hp to increase a single km/h of its top speed at 416km/h just because drag scales with the square of speed.

With this in mind they want to deploy as much energy at the beginning of the straight (slower speeds) as long as they are not grip limited ofc. At then end of the straight the prefer to not have as much power but harvest has much energy possible for the next straight. They can do it obviously first with the MGU-H but I guess they can even drive the engine against the MGU-K and harvest through the K by inducing a bit of internal magnetic drag to the crankshaft (nothing on the rules against this I think).

Because all of this is legal they have very very complex maps to create the most efficient way to run in each track. We can see by the enormous number of cyclings between H harvasting/deplyoment that happen in a single lap on the Honda telemetry how complex the "perfect map" is (no wonder they would break so much hehe).

[Big Tea]

As far as I know electrical power deployment can only be deployed by the “K” and not by the “PU”. Deploying 6-8 MJ per lap of electrical power, 4 MJ of which must be from the “ES” and the rest from the”H”, are conservative numbers compared to other numbers that have been banded about. However assuming that is correct, and on that lap the driver has managed to harvest by braking what is permitted, what happens on the next lap with a 50% charged “ES”?. If it is not qualifying, what sort of race strategy is that?. We had seen what happened with one of the red bulls when the driver said he had no deployment power left and he was told that he used it all.
I honestly don’t mean any disrespect to anybody, disagreeing does not mean that. I consider my knowledge as being able to just scratch the top of F1 matters, This being confirmed by being told that I risk being eaten up for breakfast by the more knowledgeable folks on here.

There are two very different ways the PUs and ES are used.

In qualification they start the qualification lap with 4MJ available in the ES, and they end the lap with zero. Using your assumptions they start at 8 and go down to 4.

In the race they try to start and end the lap with the same SOC. Let’s say 2MJ for the sake of this discussion.

In the case you quoted with Ricciardo a scenario might be he started the lap with 2MJ but ended it with 1MJ. He used a little more than he could recover. He was driving harder to try to catch or overtake. Now on the next lap(s) he needs to use less than he can recover to echarge the ES.

At Baku we saw Ricciardo spend several laps at a time charging his ES falling back from Verstappen a little, before launching an attack in which he intended to use most, if not all, of his ES to catch and pass. Once past he would have driven a little slower and defensively while the ES recharged.

You must now ask ourselves, if the ES can only be charged by braking how did Ricciardo brake much less than Verstappen without running off the track?

My answer is that he didn’t and that the MGU-H was used at the end of straights to send energy to the ES and not to the MGU-K.

Edit: fixed an auto correct although charming the ES is a nice idea

So is the car able to harvest all the power from the rear wheels during 'retardation' (if regen only) or is it produced at a rate too high and disks have to be used too?

[saviour stivala]

Whatever they start with, depending the “ES” size capacity, it being 12-10 or 8 MJ they can only at least safely go down 4MJ.
In the race they start end the lap with the same state-of-charge, I agree that at least they try to.
In the case I quoted it was about Verstappen when defending and nothing to do with Ricciardo.
The MGU-H is always used to recharge ES when possible, without which the ES will end up only 50% charged.

[saviour stivala]

That's not what this picture says, and I hate bringing it up because I feel it's a dead horse. It very specifically says "Max 4MJ/lap from ES to MGU-K" it does not say "Max 4MJ/lap from ES to any MGU". In fact it very specifically says ES to and from MGU-H is unlimited. You cannot read subjective interpretations into the rules, you have to interpret them objectively. State of charge only refers to energy sent to the MGU-K per lap. State of charge = 4MJ/lap, only 4MJ/lap can be sent to the MGU-K it's very straight forward.
https://statics.sportskeeda.com/wp-cont ... 24x606.jpg

According to your mental model what happens if 500kJ is sent from the ES to the MGU-H during a lap to deal with turbo lag? How many Joules can be sent to the MGU-K during that lap?

If I understand your position it would be reduced to 3.5MJ. And if they were to use electric supercharger mode the K would get less still.

In electric supercharging mode "free load mode" both "K" and "H" will be sharing "ES" power.

[godlameroso]

That's not what this picture says, and I hate bringing it up because I feel it's a dead horse. It very specifically says "Max 4MJ/lap from ES to MGU-K" it does not say "Max 4MJ/lap from ES to any MGU". In fact it very specifically says ES to and from MGU-H is unlimited. You cannot read subjective interpretations into the rules, you have to interpret them objectively. State of charge only refers to energy sent to the MGU-K per lap. State of charge = 4MJ/lap, only 4MJ/lap can be sent to the MGU-K it's very straight forward.
https://statics.sportskeeda.com/wp-cont ... 24x606.jpg

According to your mental model what happens if 500kJ is sent from the ES to the MGU-H during a lap to deal with turbo lag? How many Joules can be sent to the MGU-K during that lap?

If I understand your position it would be reduced to 3.5MJ. And if they were to use electric supercharger mode the K would get less still.

The state of charge would still be 4MJ/lap if none was deployed to the MGU-K. If you deploy 4MJ to the turbo in one lap, your SOC is still 4MJ for that lap as none was sent to the K. Only energy sent to the K from the ES counts against the 4MJ/lap SOC at least the way I see it.

[henry]

That's not what this picture says, and I hate bringing it up because I feel it's a dead horse. It very specifically says "Max 4MJ/lap from ES to MGU-K" it does not say "Max 4MJ/lap from ES to any MGU". In fact it very specifically says ES to and from MGU-H is unlimited. You cannot read subjective interpretations into the rules, you have to interpret them objectively. State of charge only refers to energy sent to the MGU-K per lap. State of charge = 4MJ/lap, only 4MJ/lap can be sent to the MGU-K it's very straight forward.
https://statics.sportskeeda.com/wp-cont ... 24x606.jpg

According to your mental model what happens if 500kJ is sent from the ES to the MGU-H during a lap to deal with turbo lag? How many Joules can be sent to the MGU-K during that lap?

If I understand your position it would be reduced to 3.5MJ. And if they were to use electric supercharger mode the K would get less still.

The state of charge would still be 4MJ/lap if none was deployed to the MGU-K. If you deploy 4MJ to the turbo in one lap, your SOC is still 4MJ for that lap as none was sent to the K. Only energy sent to the K from the ES counts against the 4MJ/lap SOC at least the way I see it.

State of charge is a quantity, it’s units are MJ. 4MJ/lap is a discharge rate. There is no restriction on the discharge rate of the ES.

The SOC restriction is not per lap but “any time the car is on the track”. So using your way of seeing it and using the restriction 4MJ/“time on track” the ES would be emptied after the first 4MJ was sent to the MGU-K and that would be it.

It is unfortunate that they picked the same number for the SOC and MGU-K transfer rate. Suppose they had chosen 5MJ for the SOC limit it would, I think, make things more obvious since we could see that the ES could service both the H and K.

[henry]

.....

So is the car able to harvest all the power from the rear wheels during 'retardation' (if regen only) or is it produced at a rate too high and disks have to be used too?

They can only harvest a fraction of the total kinetic energy in the car.

A stop from 80m/s to 40m/s is a change of kinetic energy of 1.28MJ, they typically set brake bias around 55% rearward so if the max rear braking was at 120kW , and the front somewhat less, the stop time would be about 6 seconds. In reality it’s much much shorter.

[Tommy Cookers]

the electrical braking torque stays roughly constant as car speed falls
and iirc at about 120 kph would be enough to lock the axle (as DF is trivial at low speed)


regarding generation during upshifts (Mr Stivala) this is clearly visible in the 2016 Spain Honda telemetry (P844)
showing the K being used as hard as possible to help slowing the PU rpm in upshifts (otherwise the K is just added inertia)
of course in this very brief useage it is not a particularly productive generator
and some/much of the generated energy is trapped (for internal re-use) locally by the drive's onboard capacitive store
so will not emerge to the line and be visible to the line data logging

[Big Tea]

.....

So is the car able to harvest all the power from the rear wheels during 'retardation' (if regen only) or is it produced at a rate too high and disks have to be used too?

They can only harvest a fraction of the total kinetic energy in the car.

A stop from 80m/s to 40m/s is a change of kinetic energy of 1.28MJ, they typically set brake bias around 55% rearward so if the max rear braking was at 120kW , and the front somewhat less, the stop time would be about 6 seconds. In reality it’s much much shorter.

Thanks

[sosic2121]

Whilst it is insightful to think of them deploying the MGU-K all the time they are at WOT I don’t think that’s what they do. It is more laptime efficient to run electric supercharger at the beginning of the straight, switch to ICE plus MGU-K and finally switch off the MGU-K and charge the ES, they might even use the MGU-K as well, sort of lift and coast plus. This of course makes it difficult to estimate what the total,energy flow is.

I agree 100% with you about this. Could this be the reason for "new ferrari engine sound" on corner exit?

On the other hand, I'm not sure about "divide the battery to enhance cooling".
I can't think of any good reason to treat 1 battery as 2.
Could it be that small part of battery(let's say 500-1000KJ) are super capacitors with heavy duty cycles in order to reduce lithium battery duty cycles/lower charge and discharge rates/increase efficiency/reduce heating?

[henry]

the electrical braking torque stays roughly constant as car speed falls
and iirc at about 120 kph would be enough to lock the axle (as DF is trivial at low speed)

....

I imagined the MGU-K was constant power at least until it hit the 200Nm thesshold, which I think is around 6000 rpm ICE speed.

So torque and hence TE will increase with falling speed.

I’m not sure what the best downshift strategy is other than not overreving the ICE, and staying above the 200Nm threshold.

At low speeds I would think it would be best to minimise engine braking. I wonder if they do that?

[sosic2121]

Assuming an average full-throttle lap time of ~64s, a fully utilized MGUK (as regulated) will consume a minimum of approximately 8MJ/lap, plus whatever is used at part and zero throttle. The ES can only send 4MJ to the MGUK (as regulated), so the remaining 4+MJ must be supplied by the MGUH via the ES or ICE. Said another way:

ES fed MGUK (per lap energy use)
Deploy: ≤4MJ at ≤120kW
Regen: ≤2MJ at ≤120kW

MGUH (via ICE and/or ES) fed MGUK (per lap energy use)
Deploy: ≥4MJ at ≤120kW
Regen: ≥4MJ at ≤120kW

The ES cannot vary in charge more than 4MJ per lap, so in order to deliver >8MJ per lap it must charge and discharge multiple times per lap. Two or more ES divisions (chemical or virtual) will alter how this charge cycling is allotted. No comment on the actual peak capacity of the ES.

Edit: clarified message.

Your assumption of 8 MJ + per lap deployment has been surpassed elsewhere by another assumption of 9 MJ. Fascinating reading so far which pushes the numbers possibilities well past the “K” deployment numbers possibilities. The good thing about all this is that their seems to be no risk of anybody being eaten up for breakfast.

And my assumption was 10MJ+ deployment(K+H) per lap during Q runs.

[henry]

Whilst it is insightful to think of them deploying the MGU-K all the time they are at WOT I don’t think that’s what they do. It is more laptime efficient to run electric supercharger at the beginning of the straight, switch to ICE plus MGU-K and finally switch off the MGU-K and charge the ES, they might even use the MGU-K as well, sort of lift and coast plus. This of course makes it difficult to estimate what the total,energy flow is.

I agree 100% with you about this. Could this be the reason for "new ferrari engine sound" on corner exit?

On the other hand, I'm not sure about "divide the battery to enhance cooling".
I can't think of any good reason to treat 1 battery as 2.
Could it be that small part of battery(let's say 500-1000KJ) are super capacitors with heavy duty cycles in order to reduce lithium battery duty cycles/lower charge and discharge rates/increase efficiency/reduce heating?

The way the PU deployment strategies play out emphasises the importance of small, incremental improvements. If, for instance, Ferrari have managed to get their turbine to extract an extra kW from the exhaust gases, and made their compressor need 1kW less to drive then that’s an extra 2kW to the MGU-H hence more harvesting and longer deployments in electric supercharger mode, or self sustain modes.

None of this would increase max power. But I think it possible this is where Mercedes and Ferrari are fighting it out.

I mentioned the split battery because at the time the FIA were investigating Charlie Whitehouse made mention of “virtual” batteries. For whatever reason it would appear Ferrari have run a pair of batteries in the ES for some while.

[godlameroso]

According to your mental model what happens if 500kJ is sent from the ES to the MGU-H during a lap to deal with turbo lag? How many Joules can be sent to the MGU-K during that lap?

If I understand your position it would be reduced to 3.5MJ. And if they were to use electric supercharger mode the K would get less still.

The state of charge would still be 4MJ/lap if none was deployed to the MGU-K. If you deploy 4MJ to the turbo in one lap, your SOC is still 4MJ for that lap as none was sent to the K. Only energy sent to the K from the ES counts against the 4MJ/lap SOC at least the way I see it.

State of charge is a quantity, it’s units are MJ. 4MJ/lap is a discharge rate. There is no restriction on the discharge rate of the ES.

The SOC restriction is not per lap but “any time the car is on the track”. So using your way of seeing it and using the restriction 4MJ/“time on track” the ES would be emptied after the first 4MJ was sent to the MGU-K and that would be it.

It is unfortunate that they picked the same number for the SOC and MGU-K transfer rate. Suppose they had chosen 5MJ for the SOC limit it would, I think, make things more obvious since we could see that the ES could service both the H and K.

In the latest regulations in the energy flow chart it states "one electrical sensor is connected to measure all electrical energy into and out of the MGU-K."

It also states "one electrical sensor is connected to measure all electrical energy into and out of the energy store"

https://www.fia.com/regulation/category/110

Why would they need to do this if SOC is simply based on any energy being deployed by the ES? Wouldn't they just measure that instead of specifically measuring MGU-K electrical energy?

SOC as a 4MJ window any time the car is on track needs context. That 4MJ window resets itself every time the car crosses the timing light, it doesn't matter if you depoloy .01MJ or 4MJ to the MGU-K per lap. Other than the crankshaft, the MGU-K is the only device that can send power to the wheels. Just like the old V8 KERS rule where you could only deploy for 6 seconds per lap regardless of the battery's physical charge. The same applies here, only deployment to the MGU-K(the electrical device propelling the car) counts against your 4MJ/lap pool.

I can't for the life of me figure out how you assume I think the car would run out of energy after using 4MJ any time the car is on track. I obviously know that the MGU-K and MGU-H harvest energy to replenish the ES.

If 4MJ is sent to the MGU-K in one lap, you cannot send any more energy to the MGU-K directly via the ES for that lap, you have drained your alloted state of charge for that one lap. As soon as you cross the timing light you have another 4MJ, this is regardless of the physical state of charge of the battery. The two are separate.

I guess it's a Monday thing.