Ferrari Power Unit Hardware & Software

[godlameroso]

Most electric road cars operare ~450 volts. I honestly though they were a long way away from the 1KV limit.

True, but in road cars you don't need to care (at least that much) for space and cooling requirements compared to F1. So if the rules allow up to 1kV in F1 you will always try to be as close to this limit to reduce currents i.e cooling and weight due to increased cabling cross sections. Electrical insulation is not very onerous at this voltage levels compared to cooling/cable sizing.

That's really cool to know, other than maybe the inverters I see no reason to not push the voltage to the limit.

[kalinka]

...due to increased cabling cross sections...

Sure it's a typo , it should be "decreased cabling cross sections"

[Nonserviam85]

...due to increased cabling cross sections...

Sure it's a typo , it should be "decreased cabling cross sections"

You are right it is a typo and it doesnt read correct.

It should read “to reduce currents in order to reduce the cooling and cabling cross section requirements”

[MarcJ]

5.22 When the car is on the track a lap will be measured on each successive crossing of the timing
line

https://i.imgur.com/RbNhYwH.jpg

It is implied by the regulations that the amount of energy able to be sent to the MGU-K resets every lap. If you use 4MJ half way through the lap, you cannot deploy again from the ES for that lap, regardless of how much you put back in to the ES. However if the battery is depleted and you still have allowance in your 4MJ/lap pool, you again will not be able to deploy until the battery has enough charge.

The 4MJ/lap SOC(the virtual SOC) is not replenished by anything other than the timing light, only the physical ES SOC is replenished by the MGU-K(by 2MJ/lap) and the MGU-H, and exists independent of the virtual one. The virtual SOC as you say is limited to 4MJ any time the car is on track. The actual SOC of the ES is the actual energy available. This energy is consumed by both MGU-H and MGU-K and is replenished by both. The actual SOC of the ES is not subject to the 4MJ/lap rule because the ES also sends energy to the MGU-H which is unlimited by the regulations.

That lap measurement rule doesn't say anything about energy SoC.

Meaning?

The rule that defines the 4mj SoC limit of the ES "anytime while on track" is a rule that cannot be broken unless specifically defined by another rule which would refer to this rule.

And you operate cells in their lowest internal resistance range, 40-60% SoC to reduce heat losses, cooling and maximise power efficiency.

[MarcJ]

Most electric road cars operare ~450 volts. I honestly though they were a long way away from the 1KV limit.

True, but in road cars you don't need to care (at least that much) for space and cooling requirements compared to F1. So if the rules allow up to 1kV in F1 you will always try to be as close to this limit to reduce currents i.e cooling and weight due to increased cabling cross sections. Electrical insulation is not very onerous at this voltage levels compared to cooling/cable sizing.

That's really cool to know, other than maybe the inverters I see no reason to not push the voltage to the limit.

With cell voltage at 3.1-3.8 volts at load the number of cells in series will be huge and any failure of one say going open circuit, high resistance etc compromises the whole string, going parallel doubles mass.
At 46 grams per cell for small cells that quickly adds up.
You can't really use button cells the packaging overhead is too much, only used for cell testing during development.

[godlameroso]

That lap measurement rule doesn't say anything about energy SoC.

Meaning?

The rule that defines the 4mj SoC limit of the ES "anytime while on track" is a rule that cannot be broken unless specifically defined by another rule which would refer to this rule.

And you operate cells in their lowest internal resistance range, 40-60% SoC to reduce heat losses, cooling and maximise power efficiency.

I never said that 4mj SoC is ever violated, rather that the actual capcity of the battery exceeds this amount by at least 200%. The 4MJ SoC is a per lap quantity, meaning you are allowed to deploy 4MJ from the ES to the K every lap and no more. There is no energy or power restriction for motoring the MGU-H. Only ES to MGU-K is subject to the 4MJ per lap restriction.

You cannot send 2MJ to the K one lap and then 6MJ the next. Its always 4 per lap MAXIMUM. I don't understand why you think I disagree, you yourself even said the ES is kept at 40-60% SoC, which obviously means the ES can store more than 4MJ of energy.

[godlameroso]

True, but in road cars you don't need to care (at least that much) for space and cooling requirements compared to F1. So if the rules allow up to 1kV in F1 you will always try to be as close to this limit to reduce currents i.e cooling and weight due to increased cabling cross sections. Electrical insulation is not very onerous at this voltage levels compared to cooling/cable sizing.

That's really cool to know, other than maybe the inverters I see no reason to not push the voltage to the limit.

With cell voltage at 3.1-3.8 volts at load the number of cells in series will be huge and any failure of one say going open circuit, high resistance etc compromises the whole string, going parallel doubles mass.
At 46 grams per cell for small cells that quickly adds up.
You can't really use button cells the packaging overhead is too much, only used for cell testing during development.

ES is 25kg max weight total correct? So roughly ~500 cells?

[dren]

[henry]

You said the only SOC level was a 4MJ pool and that pool did not reset every lap. It was in the Ferrari power unit thread. I said there is a 4MJ limit from the ES to the MGU-K per lap, which resets on crossing the timing beam, and consists of one SOC. In addition to that, there is also the actual battery SOC which is independent of the SOC per lap.

Yes I recall that discussion. I’m afraid I don’t see any connection between the flow restriction, 4MJ ES to MGU-K and the state of charge.

The 4MJ SOC lasts for the duration of the race it is never reset, as far as I know. The proposal I have made is that because the ES is not 100% efficient the SOC that is measured drifts down through the physical SOC. The physical SOC is consumed to replace the energy that goes in but doesn’t come out.

I am happy for anyone to show my reasoning is faulty.

ES SoC is kept in the 40-60% range as much as possible, that's where the lowest internal resistance is, it will not drift down, the current voltage maps over SoC near 30% very unimpressive, voltage drop is large, heat is higher.

If we assume that the maximum physical SOC is 20MJ. Then you are suggesting they use the range 8MJ to 12MJ. In order to do that they will need to generate more than 4MJ to make 4MJ available for consumption. I don’t know what the efficiency of the ES is, but even at low c-rates they would lose 1% round trip and some literature I’ve seen suggests very poor efficiencies at high c-rates, maybe in the range 90 to 95%.

My suggestion is that they might operate in the range 20% to 80%, 4MJ to 16MJ. which would provide them with 8MJ to replace some of the energy lost by ES inefficiency during the course of the race.

It is possible that operating in the ranges above and below 40-60% is so inefficient as to reduce the utility of that strategy. I don’t think there is sufficient data available to assess that.

[gruntguru]

Yes I recall that discussion. I’m afraid I don’t see any connection between the flow restriction, 4MJ ES to MGU-K and the state of charge.

The 4MJ SOC lasts for the duration of the race it is never reset, as far as I know. The proposal I have made is that because the ES is not 100% efficient the SOC that is measured drifts down through the physical SOC. The physical SOC is consumed to replace the energy that goes in but doesn’t come out.

I am happy for anyone to show my reasoning is faulty.

ES SoC is kept in the 40-60% range as much as possible, that's where the lowest internal resistance is, it will not drift down, the current voltage maps over SoC near 30% very unimpressive, voltage drop is large, heat is higher.

If we assume that the maximum physical SOC is 20MJ. Then you are suggesting they use the range 8MJ to 12MJ. In order to do that they will need to generate more than 4MJ to make 4MJ available for consumption. I don’t know what the efficiency of the ES is, but even at low c-rates they would lose 1% round trip and some literature I’ve seen suggests very poor efficiencies at high c-rates, maybe in the range 90 to 95%.

My suggestion is that they might operate in the range 20% to 80%, 4MJ to 16MJ. which would provide them with 8MJ to replace some of the energy lost by ES inefficiency during the course of the race.

It is possible that operating in the ranges above and below 40-60% is so inefficient as to reduce the utility of that strategy. I don’t think there is sufficient data available to assess that.

A good analysis Henry.

Until you brought this to our attention, I had never considered that this is an obvious consequence of the FIA rule and measurement applying only to the "virtual SOC" not the actual SOC.

[gruntguru]

That's really cool to know, other than maybe the inverters I see no reason to not push the voltage to the limit.

With cell voltage at 3.1-3.8 volts at load the number of cells in series will be huge and any failure of one say going open circuit, high resistance etc compromises the whole string, going parallel doubles mass.
At 46 grams per cell for small cells that quickly adds up.
You can't really use button cells the packaging overhead is too much, only used for cell testing during development.

ES is 25kg max weight total correct? So roughly ~500 cells?

So perhaps 280 in series for 1000V with a second bank in parallel for a total of 560 cells?

[gruntguru]

With cell voltage at 3.1-3.8 volts at load the number of cells in series will be huge and any failure of one say going open circuit, high resistance etc compromises the whole string, . . .

No doubt the pack includes cell monitoring technology to effectively "short out" any defective cells and maintain the integrity of the string.

[MarcJ]

Meaning?

The rule that defines the 4mj SoC limit of the ES "anytime while on track" is a rule that cannot be broken unless specifically defined by another rule which would refer to this rule.

And you operate cells in their lowest internal resistance range, 40-60% SoC to reduce heat losses, cooling and maximise power efficiency.

I never said that 4mj SoC is ever violated, rather that the actual capcity of the battery exceeds this amount by at least 200%. The 4MJ SoC is a per lap quantity, meaning you are allowed to deploy 4MJ from the ES to the K every lap and no more. There is no energy or power restriction for motoring the MGU-H. Only ES to MGU-K is subject to the 4MJ per lap restriction.

You cannot send 2MJ to the K one lap and then 6MJ the next. Its always 4 per lap MAXIMUM. I don't understand why you think I disagree, you yourself even said the ES is kept at 40-60% SoC, which obviously means the ES can store more than 4MJ of energy.

Please explain where in the regulations it says 'per lap'.
Adherence to conventions in describing rules is important.

Contravention of ES SoC not exceed 4mj anytime while on track.

DoD depth of discharge will never go beyond 60% for heat and internal resistance reasons.
Similarly charging above a SoC of 80% isn't productive, charge discharge efficiency.

[godlameroso]

The rule that defines the 4mj SoC limit of the ES "anytime while on track" is a rule that cannot be broken unless specifically defined by another rule which would refer to this rule.

And you operate cells in their lowest internal resistance range, 40-60% SoC to reduce heat losses, cooling and maximise power efficiency.

I never said that 4mj SoC is ever violated, rather that the actual capcity of the battery exceeds this amount by at least 200%. The 4MJ SoC is a per lap quantity, meaning you are allowed to deploy 4MJ from the ES to the K every lap and no more. There is no energy or power restriction for motoring the MGU-H. Only ES to MGU-K is subject to the 4MJ per lap restriction.

You cannot send 2MJ to the K one lap and then 6MJ the next. Its always 4 per lap MAXIMUM. I don't understand why you think I disagree, you yourself even said the ES is kept at 40-60% SoC, which obviously means the ES can store more than 4MJ of energy.

Please explain where in the regulations it says 'per lap'.
Adherence to conventions in describing rules is important.

Contravention of ES SoC not exceed 4mj anytime while on track.

DoD depth of discharge will never go beyond 60% for heat and internal resistance reasons.
Similarly charging above a SoC of 80% isn't productive, charge discharge efficiency.

Really? How are you going to rationalize that it's not per lap with that staring you in the face? The picture clearly says Max 4MJ from ES to MGU-K per lap. The energy going to the MGU-H doesn't count against this 4MJ amount.

[gruntguru]

I never said that 4mj SoC is ever violated, rather that the actual capcity of the battery exceeds this amount by at least 200%. The 4MJ SoC is a per lap quantity, meaning you are allowed to deploy 4MJ from the ES to the K every lap and no more. There is no energy or power restriction for motoring the MGU-H. Only ES to MGU-K is subject to the 4MJ per lap restriction.

You cannot send 2MJ to the K one lap and then 6MJ the next. Its always 4 per lap MAXIMUM. I don't understand why you think I disagree, you yourself even said the ES is kept at 40-60% SoC, which obviously means the ES can store more than 4MJ of energy.

Please explain where in the regulations it says 'per lap'.
Adherence to conventions in describing rules is important.

Contravention of ES SoC not exceed 4mj anytime while on track.

DoD depth of discharge will never go beyond 60% for heat and internal resistance reasons.
Similarly charging above a SoC of 80% isn't productive, charge discharge efficiency.

https://i.imgur.com/RbNhYwH.jpg

Really? How are you going to rationalize that it's not per lap with that staring you in the face? The picture clearly says Max 4MJ from ES to MGU-K per lap. The energy going to the MGU-H doesn't count against this 4MJ amount.

The two regulations you have circled have nothing to do with SOC. People are contradicting you because you mention SOC and "per lap" in the same sentence.