How much can the MGU-K recover per lap?

[henry]

...I think it’s impossible for a casual observer to estimate the balance of fuel loads, ERS strategies and tyre wear. We can only speculate on what might be possible and how various modes might be used.

Well, I don't think it's impossible for us to estimate. That's one of the reasons I started this thread, to get people's opinions on what might be achievable. I'll concede that it is very difficult for us to actually know what the teams are doing, but that statement can be applied to the vast majority of topics on this site.

We agree and disagree.

I think it’s quite reasonable for us to offer opinions on what techniques there might be and their effect. We can also put forward theories on how the various tools might interact. But understanding how the teams actually deploy their resources is, I think, only possible with much more data and real simulations.

Your suggestion that teams might want to use the K more after a safety car to use fuel is a case in point. They might do this, they might run richer, they might just opt to finish the race with more fuel. Which option they choose may depend on many factors, what the fuel load lap time penalty is, where they are in their stint, what the tyre deg is, which cars are around them and so on. This doesn’t negate topics that discuss these parameters and their effects but it does, IMHO, mean that we can only guess at what actually happens.

[wuzak]

I’m afraid teams will use the negative torque on throttle as a way of traction control.

Do they not do this already? IIRC, the K is excluded from use for the first 100m(i think?) of the race to prevent it being used as traction control.

Up to 100km/h from the start.

[Tommy Cookers]

regarding the extent of K generation outside the braking phase - surely they can have ? ......

overrun generation -
fuel cut off - throttle plate open on the straight just before braking, this so-called 'lift and coast' ? giving a net benefit to laptime
(it may be in conflict re simultaneous ES charging with H generation bleeding down H rpm)

powered generation -
on every upshift (it's necessary and was shown in the Honda data)
when laptime benefit outweigh sum of 'ES loop' energy loss and ICE efficiency gain when driver only wants partial torque
whenever the K generation is less than 2 MJ/lap and the H generation is not good

remember .....
the K can load 200 Nm crankshaft torque at all rpm so its effect is larger at lower ICE power/rpm/fuel rate
a dollop of energy applied to the wheels early helps laptime more than the same dollop applied later
the % fuel weight is less now than it ever was (in decades of non-refuelling) so it's not a good time to save on fuel weight ?

[AJI]

...but it does, IMHO, mean that we can only guess at what actually happens.

I appreciate that, but as a team can only be certain about what it is doing in this respect, they are in a similar situation to us in regards to their competition. ERS is a strategic guessing game with many variables, that's why I find this topic so interesting.

[AJI]

I’m afraid teams will use the negative torque on throttle as a way of traction control.

Do they not do this already? IIRC, the K is excluded from use for the first 100m(i think?) of the race to prevent it being used as traction control.

Up to 100km/h from the start.

Thanks for the clarification. So, we can pretty much agree that once the cars have hit 100km/h after the start they can use the K as a quasi traction control?

[AJI]

regarding the extent of K generation outside the braking phase - surely they can have ? ......

overrun generation -
fuel cut off - throttle plate open on the straight just before braking, this so-called 'lift and coast' ? giving a net benefit to laptime
(it may be in conflict re simultaneous ES charging with H generation bleeding down H rpm)

powered generation -
on every upshift (it's necessary and was shown in the Honda data)
when laptime benefit outweigh sum of 'ES loop' energy loss and ICE efficiency gain when driver only wants partial torque
whenever the K generation is less than 2 MJ/lap and the H generation is not good

remember .....
the K can load 200 Nm crankshaft torque at all rpm so its effect is larger at lower ICE power/rpm/fuel rate
a dollop of energy applied to the wheels early helps laptime more than the same dollop applied later
the % fuel weight is less now than it ever was (in decades of non-refuelling) so it's not a good time to save on fuel weight ?

Agreed on pretty much all of the above. The percentage of recovery for those specific events (plus braking events) is what I'm trying to work out.
You do bring up one interesting point. 200Nm at all RPM? Would it not fall away in the upper RPM?

[Tommy Cookers]

yes to that

[henry]

...but it does, IMHO, mean that we can only guess at what actually happens.

I appreciate that, but as a team can only be certain about what it is doing in this respect, they are in a similar situation to us in regards to their competition. ERS is a strategic guessing game with many variables, that's why I find this topic so interesting.

I think we’re agreeing.

TC’s contribution raises some interesting points.

Recovery during upshifts is likely to be small per shift but there are lots of them. Obviously that varies with track.

Given that this is likely to be a fixed attribute, the PU has to do it to make upshifts work, it means that at tracks where braking recovery is potentially 2 mJ , perhaps they would need to reduce the braking recovery.

I wonder how close they sail to the 2 mJ limit and what happens if they go above it?

[Tommy Cookers]

a bidirectional power flow drive would likely have a small onboard capacitive energy 'float' as necessary to function
the rules are probably written to allow this - there's no incentive to cheat as a battery is preferable as the PU ES

[godlameroso]

The unlimited nature of the MGU-H -> ES means technically you could send MGU-K regen to the H, and from the H to the ES as a way to bypass the 2MJ K -> ES limit. It would require using high voltages to minimize losses, more than the 400-480 volts most CE run on.

[AJI]

The unlimited nature of the MGU-H -> ES means technically you could send MGU-K regen to the H, and from the H to the ES as a way to bypass the 2MJ K -> ES limit. It would require using high voltages to minimize losses, more than the 400-480 volts most CE run on.

That’s an interesting workaround to the 2MJ K recovery limit. Very F1!
I’m not sure they’d have to do it for the ‘17+ cars though as I don’t believe they can recover anywhere near 2MJ under braking at most tracks, and the ‘over run’ and ‘push’ conditions surely only contribute to a small percentage of recovery. That, and the H can generate power all day long if they desire.
However, this is F1, and the K>H>ES idea is a nice ace to have up your sleeve if you are getting too close to the 2MJ limit and rear brakes temps are marginal?

[AJI]

I wonder how close they sail to the 2 mJ limit and what happens if they go above it?

I assume the CE/BBW systems prevent them from ever going above the limit (surely the penalty is a DQ?), but it must be a pretty clever piece of software because it has to manage the 0-120kW rear brake assitance which could be a real safety concern. godlameroso’s suggestion above could be one way to do it..?

[roon]

KERS on front axle only sounds great.
- Better harvesting potential
- Deployment possible at all speeds (not just higher speeds where the car is not traction limited)
- AWD for better acceleration at lower speeds (where cars are currently traction limited)
- AWD in corners for increased lateral adhesion.

isn't there really less harvesting potential from the front axle than from the rear ?
average braking recovery is with a weight+DF distribution c.50/50

On harvesting, can we say that the rear axle has more kinetic energy to harvest due to weight distribution, but front axle recovery would be easier for the driver to use? For the same reason one might bias braking force toward the front axle in a traditional hydraulic braking system.

For AWD to increase lateral grip, would the tire design need to be modified? Front tire width should increase in order to provide required adhesion while delivering drive torque. Rear tires could narrow somewhat due to elimination of electric drive on the rear axle.

[wuzak]

On harvesting, can we say that the rear axle has more kinetic energy to harvest due to weight distribution, but front axle recovery would be easier for the driver to use? For the same reason one might bias braking force toward the front axle in a traditional hydraulic braking system.

The fact is that normal braking has a front brake bias of around 60:40, so the front axle has the greater potential to recover energy.

The brake balance may be closer front to rear than before because of the MGUK on the rear axle, but that may be a braking performance trade-off for improved energy recovery.

If they went to 4 wheel energy recovery, with 120kW on the front axle and 80kW on the rear axle the recovery would match the nominal brake bias, may improve braking performance and reduce the BBW complexity.

[AJI]

...If they went to 4 wheel energy recovery, with 120kW on the front axle and 80kW on the rear axle the recovery would match the nominal brake bias, may improve braking performance and reduce the BBW complexity.

I like this. One thing I’d like to alter though is the amount of recovery. If it’s unlimited recovery and inlimited storage but compulsory all wheel recovery, then the manufacturers will use their own ideal front/rear bias and will also determine their own ideal ERS setup. Could lead to some interesting concepts?