Honda Power Unit Hardware & Software

All that has to do with the power train, gearbox, clutch, fuels and lubricants, etc. Generally the mechanical side of Formula One.
MrPotatoHead
55
User avatar
Joined: Thu Apr 20, 2017 6:03 pm
Location: All over.

Re: Honda Power Unit

Post by MrPotatoHead » Tue Jan 02, 2018 4:13 pm

Craigy is 100% right on the fact that without some nice animations it's very hard to describe the energy flow over time in just words. It is incredibly complex for sure.

Also - you have to take it with a grain of salt when you hear someone like Ferrari say "50 seconds per lap" because without knowing the duty cycle and thus power output for those 50 seconds it's impossible to know the energy total actually released.

henry
150
User avatar
Joined: Mon Feb 23, 2004 7:49 pm
Location: England

Re: Honda Power Unit

Post by henry » Tue Jan 02, 2018 7:30 pm

Craigy’s posts are very thought provoking.

They reminded me that with the advent of this formula I tried to understand the power requirements for the MGU-H to defeat lag. I think the numbers are also relevant to the “flywheel” discussion

Rotrex advertise an electrically driven centrifugal supercharger, C38-61, which has similar sort of flow/pressure profiles to the F1 numbers. It has a rotational inertia of 9.0E-3 kgm2. I doubled that to allow for the turbine and MGU-H itself giving 1.8E-2 to be accelerated.

In the driving part of the cycle this inertia will be driven by the MGU-K, 120 kw, plus the excess power from the turbine, that is the power in excess of that to drive the compressor. Obviously the latter is very variable, I’ll use 30kw for calcs. So the driving phase is 150 kw. Assuming 20 hz and symmetrical drive and recovery, and a starting speed of 110,000 rpm the speed will go up only 180 rpm. So the fluctuations will be relatively small.

Assuming symmetry the recovery will be 150 x .025 = 3.75 kJ. This gives 75 kJ/sec.

This would mean the MGU-H outputting 150kw in this mode vs the much lower, maybe 60 kw, in sustain mode.

Craigy made the good point that the cycle does not need to be symmetrical. If the MGU-H could generate at 300kw the output per cycle would double.
Fortune favours the prepared; she has no favourites and takes no sides.
Truth is confirmed by inspection and delay; falsehood by haste and uncertainty : Tacitus

AJI
32
Joined: Tue Dec 22, 2015 8:08 am

Re: Honda Power Unit

Post by AJI » Tue Jan 02, 2018 9:25 pm

Craigy wrote:
Tue Jan 02, 2018 12:48 pm
hurril wrote:
Tue Jan 02, 2018 6:40 am
I don't understand the recent thoughts of going completely haywire with transfers from the "k" to the "h"; where is that energy coming from do you think? It is not unlimited in the sense that there's an over-abundance in the source; it's unlimited in the very literal sense: you are free to transfer whichever amount you can produce. You cannot produce an unlimited amount of energy.

I think flywheel is a bad term when applied to the "h"; the amount of potential energy recoverable is probably way less than that which can be tapped into by "braking" against force/ torque applied either by the turbine or by this newfound theory of switched motoring.

But remember that this "k"-driven switched motoring is driven by physical fuel so it is an accounting trick more than one relating to efficiency. Min-maxing if you will.
What I'm trying to describe isn't actually that easy to describe without an animation because by the nature of the topic we're talking about energy flow directions changing around. I feel like I'm doing a bad job of explaining it - sorry for that.

The idea that the MGU-H as a flywheel provides a route for the MGU-K to dump whatever energy is needed into the ES, within the limitations described above in my earlier post. Yes, this will ultimately be from fuel and the IC, of course. It's not generating energy, it's just permitting a workaround to get more into the ES on a lap by lap basis.
It's an energy transfer route which wasn't/isn't (to me) immediately obvious on the flow diagram that form the technical regulations.

In terms of accountancy for energy, if you could take energy from the end of every straight and deploy it at the start of the next straight, that isn't just accountancy with energy. It's a larger area under the speed graph, and a faster laptime. 500KJ at the end of the Kemmel isn't as helpful as 500KJ on the run down to Pouhon, in terms of laptime.

People mention "steady state" and that made me think that the flow on the diagram wasn't changing from K to H and back, but thanks to Honda's lovely end-of-2017 information we know that's a poor assumption: this is being done.

Regarding the terminology of what I've called the "flywheel" transfer technique - please bear in mind that it's not like we're spinning the MGU-H up once per lap to huge speed and then deploying it all on a straight or something like that, totally ignoring its role in the turbomachinery of the ICE. It wouldn't actually store very much power anyway unless it was really heavy. The MGU-H would instead be spun up and down many times per second, storing and releasing a smaller amount of energy each time, and doing so very frequently. It would essentially just be noise on the graph of the required ICE compressor speed - for example could be plus or minus less than 1% (~1000rpm either side) of what you'd otherwise want just for ICE usage, from the range of 0 to 125K RPM for the MGU-H (working range is likely 80K-120K overall). It's not like the compressor is directly attached to the crankshaft or something: the driver won't feel it spinning up and down by that much. I doubt that it'd even show up at the intake valves because the whole air intake system is full of compressed air that would absorb pulses of higher and lower pressure like a spring.

I offer the opinion that not only is this flywheel effect something that needs to be designed-in, the optimal way to get it working (staying at an optimal 4KG mass for the MGU-H) would be to make the disk (flywheel diameter) bigger than needed for the compressor or turbine needs. I mention this because it's been publicised that Honda moved their compressor to make it larger, and this is a potential reason for that, besides the needs of the ICE's compressor.
In Hondas case that would potentially require that the compressor be moved from the Vee of the engine to the front or the back in order for the diameter to grow so that the flywheel energy storage potential is larger to facilitate transfers per pulse.

Here are Some very rough flywheel storage figures - say we are trying to move energy from the K to the ES via the H, and we're doing that for 50s per lap (somewhere like Spa, which is 57% full throttle by time and 70% full throttle by distance on a 110s race lap). At no point are we trying to recover energy to the ES from the ERS-K via ERS-H route while the driver is at 100% throttle.

A few stated assumptions: 50s duration, flywheel of 4Kg (the minimum in the rules), 100mm diameter (seems sensible enough for the turbo sizing alone), 40 pulses per second, 2000rpm rev change (of the MGU-H) during a pulse. Energy per pulse depends on the mass distribution on the MGU-H (mass in the centre has less inertia than the edge), but lets say that it's something like 180J because the majority of the mass is on the outer ring of the MGU-H, due to some clever designer putting weight out there.
In this case, we have 50s per lap * 40 events per second * 180J per event = only an additional unmetered 360KJ per lap. That's only 3s of full deployment and hardly worth the trouble. You could go for a larger RPM change, at cost of the ICE seeing more and more variable levels of manifold pressure.

If we adjust the shape of the flywheel (make it 200mm diameter rather than 100mm above) and leave everything else the same:
Then in this case we have 50s per lap * 40 events per second * 700J per event = 1.4MJ per lap. Quite worthwhile. Perhaps worth moving the flywheel disk out of the Vee of the engine so make it larger without being heavier, permitting more flywheel inertial storage without a weight penalty.

In terms of the effect on the ICE of doing all of this, it would see some variations in manifold pressure and in exhaust pressure due to the speeding up and slowing down of the MGU-H, compressor and turbine, but the variations would be so fast that it really wouldn't make the ICE feel different to drive than if the system wasn't in place.

In terms of mapping the ES deployment/charge, the big plus point of this system is to work around the 2MJ transfer limit from MGU-K to ES. You'd essentially have this system pulling power from the IC via the MGU-K, dumping it into the MGU-H and on to the ES whenever the driver isn't requesting 100% from the PU - it could be recovering any time on part throttle or zero throttle, as required. You would still use the ERS-K to ES direct route for the 2MJ permitted per lap of course, but the rest of the energy would be most effectively recovered using the flywheel method.

In terms of redeploying ES energy, there's a flow limit of 4MJ from the ES to the MGU-K using the direct route per lap, but again this could be worked around because there's nothing stopping the route from the ES to the K via the H in flywheel-routed "deployment" mode.

The 4MJ SoC ES limits don't appear to stop the car from actually receiving and deploying more than 4MJ per lap because of how that rule is worded.
It appears that the single biggest deployment event you can have within a lap is 4MJ (taking the ES from "full" to "empty"), but so long as you discharge and recharge you can cumulatively have more than 4 MJ entering and leaving the ES in a given lap. The ES to K *direct* route is limited to 4MJ out per lap, and 2MJ back to ES. There's no limit on ES to MGU-H transfers or rate, remember.

The route via the flywheel system suggested would raise those 2MJ and 4MJ limits significantly, potentially to the point where the MGU-K is deploying 120kW the whole time the PU is at 100% output.

Can I make a hypothetical scenario to explain this? (might not be that hypothetical, as it may correlate to Honda's first pass at the PU a few years ago):

If you were limited to (for example) 2MJ of energy into the ES per lap because you'd designed a system where only the MGU-K charged the ES, then you'd be limited to 2MJ of deployment on the next lap. You'd hear drivers complain a lot about a lack of deployment, because your 2MJ is only enough for 16.6s of deployment round a given lap. Does this sound familiar?

You would need to figure out how to add to the energy input into the ES, over the 2MJ you'd started off with.
You're potentially already at the 2MJ direct limit from the MGU-K from braking and from charging towards the end of straights, so you need to add whatever you could get from MGU-H using compounding. This is essentially limited by what the ICE's operating parameters are: the higher the amount of energy remaining in the exhaust gases, the worse the basic ICE performance was at driving the rear wheels directly, and the more fuel-hungry the ICE was.

If instead of this, you designed a system to pull energy from the K via the H into the ES, you have the ability to charge the ES at an abritrary amount per lap, and deploy 4MJ directly via the MGU-K on every lap as you please.

You can do that deployment wherever you want (start of straights, of course, but actually on quite a few circuits, most of the lap because you have 33s to play with).
You also have the ability to supplement this direct ERS-K drive route with energy taken from the ES to the MGU-K, then dumped in small pulses into the MGU-K using the "unmetered" ES->H->K route.
Your 120kW deployment is in fact made of pulses of power direct from the ES to the MGU-K, interleaved with pulses of power from the MGU-H to the MGU-K.
Your deployment per lap consists of 4MJ direct from ES to K plus the additional energy using the unmetered ES to MGU-H to MGU-K route, within the constraint that the ES State of Charge can't go above 4MJ or below 0MJ from their starting values.

Your deployment strategy now revolves around maximising the factors you control: ES SoC and MGU-H sizing, along with where on the lap to charge and where to deploy, using each system.
The more power the MGU-H has been designed to transfer, the physically bigger it will be
There's a compromise in terms of volume versus weight versus power rating.
You likely size the MGU-H to cope with the harshest duty-cycle scenarios on the calendar (Spa?).

Question: What would people prefer we call this technique if not "flywheel"?
I absolutely love where this topic is going. Thanks Craigy and godlameroso for kicking it off, but as this is the Honda PU thread and this discussion applies to all manufacturers (and teams for that matter) can I please ask one of the mods to move it to its own thread?
There's so much more to discuss.

Craigy
69
User avatar
Joined: Tue Nov 10, 2009 9:20 am
Location: UK

Re: Honda Power Unit

Post by Craigy » Tue Jan 02, 2018 9:28 pm

AJI wrote:
Tue Jan 02, 2018 9:25 pm
I absolutely love where this topic is going. Thanks Craigy and godlameroso for kicking it off, but as this is the Honda PU thread and this discussion applies to all manufacturers (and teams for that matter) can I please ask one of the mods to move it to its own thread?
There's so much more to discuss.
Thank you all for the kind words. I agree this should probably be in the main engine formula thread.

*reports own post*

PlatinumZealot
326
User avatar
Joined: Thu Jun 12, 2008 2:45 am

Re: Honda Power Unit

Post by PlatinumZealot » Wed Jan 03, 2018 2:07 am

hurril wrote:
Tue Jan 02, 2018 6:40 am
I don't understand the recent thoughts of going completely haywire with transfers from the "k" to the "h"; where is that energy coming from do you think? It is not unlimited in the sense that there's an over-abundance in the source; it's unlimited in the very literal sense: you are free to transfer whichever amount you can produce. You cannot produce an unlimited amount of energy.

I think flywheel is a bad term when applied to the "h"; the amount of potential energy recoverable is probably way less than that which can be tapped into by "braking" against force/ torque applied either by the turbine or by this newfound theory of switched motoring.

But remember that this "k"-driven switched motoring is driven by physical fuel so it is an accounting trick more than one relating to efficiency. Min-maxing if you will.
Mguk energy is harvested from the rear axle primarily during braking but less so durnig coasting at top speed. Of course some tracks give less harvesting than others. And in the same way, some tracks are less sensitive to ERS deployment... Such as monaco.

You missunderstand when we say flywheel energy. We are not talking about energy from the turbine wheel here. We sre talking about residual kinetic energy from using the MGUK to spin the MGUH in 20 to 40 Hz cycles.

The K drive motoring is not intentionally driven by burning fuel. That would be inefficient. It is primarily driven by braking. The only time fuel energy may be used to the K would be for like cornering events and high speed coasting. But this is just a benefficial side effect of using wasted energy from the driver controlling the car with the throttle in corners and high speed coasting.

EDIT:
AJI said it better than me!
On the end of straights where alot of energy is burnt for less laptime you can dump energy from the ICE into the energy store using the same mguh flywheel loophole for use at another part of the track where you need that acceleration to make a big improvement in laptime. The idea is not to burn fuel wantonly. But there are places like end of straights and in some corners where fuel is wasted anway.. So better to use it for "extra harvest"
"The true champions are also great men. They are capable of making difficult decisions, of admitting their mistakes and of pushing harder than before when they get up from a fall."

- Ferrari chairman Sergio Marchionne

Webber2011
41
Joined: Tue Jan 25, 2011 12:01 am
Location: Australia NSW

Re: Honda Power Unit

Post by Webber2011 » Wed Jan 03, 2018 3:01 am

Ok guys, so me being a bit of a dumbo with regards to some of the discussion about how all this energy is transferred between the different elements I have a few questions :lol:

They have most likely already been answered but not in a way I really understand.
( Or they maybe have but I've missed it in among a billion or so posts ? )

PZ, your last post raises a question I've been wondering about.
When a driver decides to "lift and coast" at high speed I imagine at times they are still using part throttle, as to use none at all would cause massive braking simply because of the drag.

So how much energy do they recover with a simple "lift and coast" as opposed to a hard braking event at the end of the same high speed straight ?

Is there some sort of equation like if you lift and coast at 300 metres, you gain the same as braking hard at 100 metres ?
Or is that too track dependent ?

I know this will differ from team to team, but is there some sort of rule of thumb ?

Thank you
Simo'

PlatinumZealot
326
User avatar
Joined: Thu Jun 12, 2008 2:45 am

Re: Honda Power Unit

Post by PlatinumZealot » Wed Jan 03, 2018 4:05 am

Ok. I admit it is not fresh in mind the vehicle dynamics to calculate the energy they can skim off from harvesting at end of straight and lift and coast.
But the gist of it that drag froce is proportional to velocity squared. And power is force times velocity, so power is proportional to velocity cubed.
Ok what this means.. Is that i can sacrifice some amount of power at end of straight and my velocity barely drops and also my lap time is negligably affected. This power is "invested" into the ERS to use at a more power sensitive part of the track.

I lift and coast when i drive to save fuel. But as you say the formula 1 guys do not fully come off the throttle before the true lift and coast. I presume this is where the MGUK adds additional electrical load and leeches some power from the ICE. The same for the MGUH to the turbine. It will not cause a significant speed drop at end of straight I think.. I would estimate you need 100hp to go 15kph faster in F1 but most of that has to do with accerating from low speed to high speed. At end of straight when acceleration has tapered off, the rate of change of horsepower with speed is much smaller.

Example... I chose to give up 5kph from 330 kph end of straight for three seconds. That roughly 35 bhp (26kW) engine power sent to mguk***. Energy wise It is 78kJ. Sounds small but it can be extended and done on all the straights. It would add up.
Laptime would only slow by 0.004 seconds on that particular end of straight. But it would be a bigger laptime gain if i use that energy at another critical part of the track to accelerate.

Say, this is only for the harvesting done before lift and coast. Lift and coast itself is a direct fuel saving measure i think.

EDIT
*** a conundrum.. I have i no idea what proportion of battery power is used down the straights! I assumed 800 hp from the ice alone. However you need battery power for initial acceleration! So i cannot confirm if ice alone is deployed at END of straigts typically, but common sense says you can only charge or discharge the battery at any one time. So therefore, very likely harvesting at end of straight means the ice alone is driving the flywheel and the battery is charging only.
Last edited by PlatinumZealot on Wed Jan 03, 2018 5:39 am, edited 1 time in total.
"The true champions are also great men. They are capable of making difficult decisions, of admitting their mistakes and of pushing harder than before when they get up from a fall."

- Ferrari chairman Sergio Marchionne

gruntguru
395
Joined: Sat Feb 21, 2009 6:43 am

Re: Honda Power Unit

Post by gruntguru » Wed Jan 03, 2018 4:40 am

Craigy wrote:
Tue Jan 02, 2018 12:48 pm
A few stated assumptions: 50s duration, flywheel of 4Kg (the minimum in the rules), 100mm diameter (seems sensible enough for the turbo sizing alone), 40 pulses per second, 2000rpm rev change (of the MGU-H) during a pulse. Energy per pulse depends on the mass distribution on the MGU-H (mass in the centre has less inertia than the edge), but lets say that it's something like 180J because the majority of the mass is on the outer ring of the MGU-H, due to some clever designer putting weight out there.
In this case, we have 50s per lap * 40 events per second * 180J per event = only an additional unmetered 360KJ per lap. That's only 3s of full deployment and hardly worth the trouble. You could go for a larger RPM change, at cost of the ICE seeing more and more variable levels of manifold pressure.
You have assumed a 2000 rpm delta but not stated the initial speed. If the initial speed is 120,000 the energy delta is 26 kJ (assuming 50 mm radius of gyration). If the initial speed is zero, the energy delta is 219 J which I assume is the scenario you used.

Although this looks more optimistic for your theory, I still don't buy it. The rules don't allow energy (beyond 4 MJ/lap) to be transferred from the ES to the 'K and this is nothing but a loophole which the FIA would close in a heartbeat once discovered. The unusual energy flow would be obvious in data reviewed by officials.
je suis charlie

PlatinumZealot
326
User avatar
Joined: Thu Jun 12, 2008 2:45 am

Re: Honda Power Unit

Post by PlatinumZealot » Wed Jan 03, 2018 5:43 am

gruntguru wrote:
Wed Jan 03, 2018 4:40 am
Craigy wrote:
Tue Jan 02, 2018 12:48 pm
A few stated assumptions: 50s duration, flywheel of 4Kg (the minimum in the rules), 100mm diameter (seems sensible enough for the turbo sizing alone), 40 pulses per second, 2000rpm rev change (of the MGU-H) during a pulse. Energy per pulse depends on the mass distribution on the MGU-H (mass in the centre has less inertia than the edge), but lets say that it's something like 180J because the majority of the mass is on the outer ring of the MGU-H, due to some clever designer putting weight out there.
In this case, we have 50s per lap * 40 events per second * 180J per event = only an additional unmetered 360KJ per lap. That's only 3s of full deployment and hardly worth the trouble. You could go for a larger RPM change, at cost of the ICE seeing more and more variable levels of manifold pressure.
You have assumed a 2000 rpm delta but not stated the initial speed. If the initial speed is 120,000 the energy delta is 26 kJ (assuming 50 mm radius of gyration). If the initial speed is zero, the energy delta is 219 J which I assume is the scenario you used.

Although this looks more optimistic for your theory, I still don't buy it. The rules don't allow energy (beyond 4 MJ/lap) to be transferred from the ES to the 'K and this is nothing but a loophole which the FIA would close in a heartbeat once discovered. The unusual energy flow would be obvious in data reviewed by officials.
Guru, take a look at the Honda graph in the last image. The data is there and it is obvious! Look at the rapid osciallting part of the mguh plot in yellow. That is the alternating generator/motor action.

https://i.imgur.com/k0cys2U.jpg
"The true champions are also great men. They are capable of making difficult decisions, of admitting their mistakes and of pushing harder than before when they get up from a fall."

- Ferrari chairman Sergio Marchionne

gruntguru
395
Joined: Sat Feb 21, 2009 6:43 am

Re: Honda Power Unit

Post by gruntguru » Wed Jan 03, 2018 6:08 am

Can't see it.
The MGUH current oscillates a lot during upshifting (not surprising) and there are downspikes during extended braking (electric turbine spooling not required during downshift throttle blips).

Is there something else you are looking at?
je suis charlie

wuzak
344
Joined: Tue Aug 30, 2011 2:26 am

Re: Honda Power Unit

Post by wuzak » Wed Jan 03, 2018 6:11 am

PlatinumZealot wrote:
Wed Jan 03, 2018 5:43 am
gruntguru wrote:
Wed Jan 03, 2018 4:40 am
Craigy wrote:
Tue Jan 02, 2018 12:48 pm
A few stated assumptions: 50s duration, flywheel of 4Kg (the minimum in the rules), 100mm diameter (seems sensible enough for the turbo sizing alone), 40 pulses per second, 2000rpm rev change (of the MGU-H) during a pulse. Energy per pulse depends on the mass distribution on the MGU-H (mass in the centre has less inertia than the edge), but lets say that it's something like 180J because the majority of the mass is on the outer ring of the MGU-H, due to some clever designer putting weight out there.
In this case, we have 50s per lap * 40 events per second * 180J per event = only an additional unmetered 360KJ per lap. That's only 3s of full deployment and hardly worth the trouble. You could go for a larger RPM change, at cost of the ICE seeing more and more variable levels of manifold pressure.
You have assumed a 2000 rpm delta but not stated the initial speed. If the initial speed is 120,000 the energy delta is 26 kJ (assuming 50 mm radius of gyration). If the initial speed is zero, the energy delta is 219 J which I assume is the scenario you used.

Although this looks more optimistic for your theory, I still don't buy it. The rules don't allow energy (beyond 4 MJ/lap) to be transferred from the ES to the 'K and this is nothing but a loophole which the FIA would close in a heartbeat once discovered. The unusual energy flow would be obvious in data reviewed by officials.
Guru, take a look at the Honda graph in the last image. The data is there and it is obvious! Look at the rapid osciallting part of the mguh plot in yellow. That is the alternating generator/motor action.

https://i.imgur.com/k0cys2U.jpg
Is it obvious?

Can you cut the relevant bit out and show the obvious bit?

gruntguru
395
Joined: Sat Feb 21, 2009 6:43 am

Re: Honda Power Unit

Post by gruntguru » Wed Jan 03, 2018 6:14 am

Interesting though that apart from braking the K is almost always motoring at maximum. I didn't think they had that much energy available.???
Also the H has an almost flat line when generating. It looks as if it has an electric limit lower than the surplus turbine power available.
je suis charlie

wuzak
344
Joined: Tue Aug 30, 2011 2:26 am

Re: Honda Power Unit

Post by wuzak » Wed Jan 03, 2018 6:35 am

The energy flow diagram shows the extra energy flowing from the MGUK to the MGUH and back to the ES.

But if they are dragging the MGUK to get that extra 1MJ to give to the MGUH, what is the MGUH doing?

hurril
42
Joined: Tue Oct 07, 2014 12:02 pm

Re: Honda Power Unit

Post by hurril » Wed Jan 03, 2018 8:01 am

Thank you very much, Craigy, for a brilliant post!

I agree with you all the way through and this is exactly as I understand what is suggested too. I can see what it is that some of you are calling flywheel(:ing) but there's no such thing required here. The spinning mass is enough but I'm not going to argue wording. Red herring, etc :)

The problem, as I see it, is that while the logic is sound, there's a skyhook in it. You say that braking the turbine only lowers the engine output and raises the fuel consumption. Well... what does the MGU-k do to provide the means of transferring energy in an "un-taxed" way through the MGU-h to the ES? It deletes 160kW worth of power that has been bought with fuel. Thus raising the specific fuel consumption.

This is not to say that this strategy is not viable nor that it isn't being used. Just that the value of it is probably a tad overstated because it is so sexy. (My inner nerd is drooling at this.)

My understanding is also that it is not that easy to even reach the allowed 2Mj per lap in "normal" MGU-k harvesting. There just isn't enough area under the time-under-breaking x 160kW graph.

(Again: what an excellent post!)

henry
150
User avatar
Joined: Mon Feb 23, 2004 7:49 pm
Location: England

Re: Honda Power Unit

Post by henry » Wed Jan 03, 2018 11:13 am

I think we need to differentiate between qualifying energy flows and race energy flows.

In qualifying fuel consumption isn’t really an issue so I imagine the following based on the rules

Energy positive
Start the lap with somewhat less than 4 mJ in the ES , some will have been used getting from the last corner to the start line end with 0

Recover 2 mJ from the K during braking and motoring against the K in sections not at WOT

Energy negative
Drive the beginning of straights in electric supercharger mode, ES drives K and H

Drive the next part of straights in self-sustain-plus , ES and H driving the K

Drive the final part of the straight in self-sustain, H driving the K

Putting some numbers to this

Start the lap with 3.5 mJ in the ES and add 2 making 5.5 to spend on a qualifying lap

In electric supercharge spend ES at 180 kw , assume 90 kw to drive the compressor with 30 from blow down and 120 to the K

In SS+ spend ES at 60 kw, assume the H provides 60

Spend some on lag control, I don’t know this so I’ll ignore it, just as I’ll ignore H to ES in part throttle, I would expect some but can’t begin to quantify.

Because of 4 mJ rule we have 1.5 mJ to spend driving the H, which gives 25 seconds (1.5 mJ at 60 kw)
In this mode the K consumes 120 kw for 25 seconds, 3 mj.
Leaving 17 seconds in SS+ for a total of 42 seconds at WOT.
The rest of the WOT would be in SS

If braking provides 1mJ to the ES from the K we would need to motor for 8 seconds against the K to provide the other 1 mJ.

If the “flywheel” exists any contribution it makes would extend the time in Electric suppercharger mode and reduce SS+ and reduce lap time.

It is also possible that at the end of the straight they switch to ICE only mode and divert the H to the ES allowing a little more time in supercharge.

In Race mode the numbers are very different. Consumption matters and so burning extra fuel to keep the ES charged becomes an issue to be managed.

Most of the time they start and end the lap with the same charge, 2 mJ.

If they get 1 mJ from braking they need to trade off power at the end of straights to increase power at the beginning,

If they switch to ICE only they get, say, 60 kw into the ES and if they go to ICE minus, I.e. dragging the K they get 180 kw, topping up the ES from the K in only 5 seconds.

If they do burn fuel above this the “flywheel” mode might be useful to build up the ES ready for some use of the overtake button

I’m afraid this gets too complex for me. With 4 power modes, 4 recovery options, track configuration, weather, tyres, aero setup, etc optimising all of the parameters for a race looks remarkably hard. Qualifying seems so much easier.
Fortune favours the prepared; she has no favourites and takes no sides.
Truth is confirmed by inspection and delay; falsehood by haste and uncertainty : Tacitus