Ferrari Power Unit Hardware & Software

[turbof1]

The only mandated sensor is the IVT between the ESS and inverter. Any improvement in the efficiency of the inverter, motor and even HV cables is free power.

Even though there are no mandatory sensors between crank and MGUK I know that most teams choose to monitor the torque between the two for reliability purposes. My brother in law works at torquemeters (http://www.torquemeters.com/applications/automotive/) and at the start of the V6 era they were supplying all engine manufacturers.

So the FIA has no direct means of reading the MGU-K power output on the crank. I assume they can do so by calculations based on readings elsewhere? Because the mgu-h can deliver power through the mgu-k to the crank as well without going through the battery first.

They don’t measure MGU-K power directly using a torque sensor, they do measure it using a single electrical sensor connected to the CU-K. they allow a maximum power of 126.3 kW at this measurement point. (120 / 0.95)

Measuring power would appear to be much more straightforward than measuring the energy flow. In the latter case the FIA want to measure flow between ES and MGU-K and somehow have to ignore the flows to/from the MGU-H.

Maybe the regulations are incomplete but as it is documented (page 105 of the 2018 regulations, dec 2017) they only require two sensors to measure all of the electrical flows.

Oh dear, I actually missed that small note below on the energy flow chart .

There are more sensors however. 8.2.2 mentions this:

Code: Select all

8.2.2 All control sensors, actuators and FIA monitoring sensors will be specified and homologated by the FIA. Details of the homologation process may be found in the Appendix to the Technical Regulations. 

I'm currently unsure to which Appendix it refers, possible to one outside the public rule book.

[henry]

....

Maybe the regulations are incomplete but as it is documented (page 105 of the 2018 regulations, dec 2017) they only require two sensors to measure all of the electrical flows.

Oh dear, I actually missed that small note below on the energy flow chart .

There are more sensors however. 8.2.2 mentions this:

Code: Select all

8.2.2 All control sensors, actuators and FIA monitoring sensors will be specified and homologated by the FIA. Details of the homologation process may be found in the Appendix to the Technical Regulations. 

I'm currently unsure to which Appendix it refers, possible to one outside the public rule book.

The text concerning the control of energy management wasn’t in the original diagram, I’m not sure when it appeared.

They don’t make it easy for enthusiasts to understand their sport, but maybe they don’t expect people to be interested in how things work.

[Phil]

According to the article it is one battery but treated as two by the software. The fun part is figuring out what is there to gain from doing this.

What if the idea of coming up with such a complex solution is for the sole intention to bypass the FIA sensor to get above certain limits? At the end of the day, it’s all down to clever interpretation of the rules. Even if it lets you ride on an advantage for half the season or indeed a couple of years, it may well be worth it.

Oil burning is similar. Mercedes has probably been doing it since way back in 2014, but it’s only since 2017 that it has became a big topic and the FIA has started to make big progress in closing it down and policing it in a highly restrictive manner. Why not the clever interpretation of Ferraris battery?

[roon]

"Treats it as two batteries..." These sorts of cryptic statements of dubious accuracy are usually found in the Honda PU thread.

Can this be related to:

-state of charge/discharge as interpreted by the software
-frequency (AC in the battery)
-three phase (two negative terminals on the battery, emulated or otherwise)

If there is modulation of current, AC, in the battery, perhaps this could influence measurements. Peaks and troughs instead of flat DC.

Continuous high frequency transfer of charge from one half of the battery to the other (treating it "as two") might emulate the effect of an inverter. A battery which provides AC power directly. Such an arrangement might again influence measurement, since the SoC across the ES would vary many times per second. 50Hz? 500Hz? Higher?

The maximum, alternating SoC could then be greater than that found in a DC ES. How the average power output and storage density is then calculated could be at play here.

[PlatinumZealot]

"Treats it as two batteries..." These sorts of cryptic statements of dubious accuracy are usually found in the Honda PU thread.

Can this be related to:

-state of charge/discharge as interpreted by the software
-frequency (AC in the battery)
-three phase (two negative terminals on the battery, emulated or otherwise)

If there is modulation of current, AC, in the battery, perhaps this could influence measurements. Peaks and troughs instead of flat DC.

Continuous high frequency transfer of charge from one half of the battery to the other (treating it "as two") might emulate the effect of an inverter. A battery which provides AC power directly. Such an arrangement might again influence measurement, since the SoC across the ES would vary many times per second. 50Hz? 500Hz? Higher?

The maximum, alternating SoC could then be greater than that found in a DC ES. How the average power output and storage density is then calculated could be at play here.

Only way to have AC in a battery is frequent charge and discharge, and even then it would be very inefficient - a lot of energy would get wasted as heat. I don't see any advantage there.
Remember the battery is one... meaning the cells have one set of terminals where the voltage is measured.
Now circuits internal to the battery.. with their own power electronics.. I don't know if that is legal...

[kalinka]

"Treats it as two batteries..." These sorts of cryptic statements of dubious accuracy are usually found in the Honda PU thread.

Can this be related to:

-state of charge/discharge as interpreted by the software
-frequency (AC in the battery)
-three phase (two negative terminals on the battery, emulated or otherwise)

If there is modulation of current, AC, in the battery, perhaps this could influence measurements. Peaks and troughs instead of flat DC.

Continuous high frequency transfer of charge from one half of the battery to the other (treating it "as two") might emulate the effect of an inverter. A battery which provides AC power directly. Such an arrangement might again influence measurement, since the SoC across the ES would vary many times per second. 50Hz? 500Hz? Higher?

The maximum, alternating SoC could then be greater than that found in a DC ES. How the average power output and storage density is then calculated could be at play here.

Only way to have AC in a battery is frequent charge and discharge, and even then it would be very inefficient - a lot of energy would get wasted as heat. I don't see any advantage there.
Remember the battery is one... meaning the cells have one set of terminals where the voltage is measured.
Now circuits internal to the battery.. with their own power electronics.. I don't know if that is legal...

Yes, AC in a battery is not possible / very inefficient. Re: One set of terminals...That's highly unlikely. Batteries are made of smaller cells, and you have to have multiple terminals coming out of the battery in order to be able to balance the charge. Without balancing, such battery life/efficiency/reliability would be impossible IMO. These batteries are charged/discharged too many times with too much current to be able to go without balancing. So basically, even without any tricks, you have many connections to the ES. Discharge can be only 2 terminals, but not the charging terminals. When they say one battery treated as two, I think of a battery packed in a single container, but leads coming in/out making it two seen by the electronics. I'm not sure it's illegal, Physically, you have only one battery, but with more terminals.

[PlatinumZealot]

Nope. One set of connections where the voltage is taken from. Unlikely two voltage outputs doing there own thing are considered one battery.

[rscsr]

Nope. One set of connections where the voltage is taken from. Unlikely two voltage outputs doing there own thing are considered one battery.

I haven't found anything in the technical regulations that would require them to have only one battery (or only one set of connections). It just says there is one ES. The only limit would be that the chart says that there is one sensor, that monitors all energy flowing from and to the ES. But that is easily solvable by using a common ground, where the sensor is connected to and have 2 terminals for the MGU-K and MGU-H.

[J.A.W.]

...Oil burning is similar. Mercedes has probably been doing it since way back in 2014, but it’s only since 2017 that it has became a big topic and the FIA has started to make big progress in closing it down and policing it in a highly restrictive manner. Why not the clever interpretation of Ferraris battery?

Actually Phil, 'oil burning' - as a power enhancing technique, was raised in this forum.. over 2 years ago.. April 2016..

[SiLo]

My understanding was that it was actually two batteries but one unit. So two separate sets of cells inside the battery "box".

[turbof1]

My understanding was that it was actually two batteries but one unit. So two separate sets of cells inside the battery "box".

Charlie Whiting described it being virtual split in 2, not physically. My guess is there are a dozen or more cells inside the energy store and each one can be controlled individually, with CU-K and CU-H each handling a portion of the cells, creating 2 batteries that way.

Possible is that the system as a whole constantly monitors the state of each individual cell and if needed reallocates cells to the one "battery" to the other. If the system is constantly busy swapping cells and reconfiguring to optimal situation, we can quickly get a hugely complex thing going on.

Perhaps there are even cells left unused in the case that some break down? The rules actually don't dictate a maximum capacity for the ES, just that you can only pull 4MJ from it the MGU-K and the MGU-K can only put 2MJ on the ES (each lap). I think the ES is perfectly allowed to be 10MJ capacity if you are able to stay within the weight restraints. It has to weight at minimum 20kg and at max 25kg. If you reach your optimal capacity at let's say 15kg, you still need to add weight. That 5kg can be reserve capacity. The energy stores have to do half a season, so having cells breaking down will not be such an issue if you can just use the reserve capacity. It's not even a CoG issue because the PU has a fixed CoG in the regs.

[henry]

Nope. One set of connections where the voltage is taken from. Unlikely two voltage outputs doing there own thing are considered one battery.

I haven't found anything in the technical regulations that would require them to have only one battery (or only one set of connections). It just says there is one ES. The only limit would be that the chart says that there is one sensor, that monitors all energy flowing from and to the ES. But that is easily solvable by using a common ground, where the sensor is connected to and have 2 terminals for the MGU-K and MGU-H.

You are right. The term “battery” is not mentioned at all. The ES definition says:

1.27 Energy Store (ES) :
The part of ERS that stores energy, including its safety control electronics and a minimal housing.

The “safety control electronics” might include overtemperature, cell balancing etc. but also the way of implementing virtually the “two batteries”.

In your scenario of two non-ground connections do you think it might be possible to modify the phasing of the outputs, I doubt they are a nice smooth DC, such that the other monitored sensor, at the MGU-K might subtract energy it thinks, by its timing, was destined for the MGU-H.

I imagine the control electronics and software between the 3 active elements (I think they call it the DC-DC converter) is pretty complex.

I think that at issue is not the power at the MGU-K, which in your scenario is easily measured at common ground and is instantaneous, but the SOC and energy flows which are dependent on a time integral. Might Phase differences between ES, H & K be exploitable to confuse these measurements.

[bjpower]

is it possible they are using 2 different types of cells in the one battery.
one set designed for rapid charging and one set for rapid discharging?

the charging cells would trickle(for lack of a better term) charge the discharging cells.
It may be more efficient and more reliable

[SiLo]

i don't think so. Ideally you would want something that charges and discharges quickly.

[outsid3r]

Both charging and discharging creates heat in the batteries, which in turn reduces efficiency. Could it be that they use "battery A" (which is a set of cells from the single battery pack) for a lap and then switch to "battery B" for the next lap? So in essence alternating between cells and having some rest and cool down, while the others are being used?

Also, isn't the battery pack and control unit part of the "power unit" so should be equal to customer teams? Probably Ferrari is just trying to do something cleaver with software for life and/or heat management (as mentioned by others in this forum) but hardware is identical between all teams.