Ferrari Power Unit Hardware & Software

[Big Tea]

Without wanting to go off topic, I have this strage bias/intuition from my electrical engineering education about what is more or less cheating... For example, consider three hypothesized tricks

1. Sampling alias tricks by controling the fuel pump to pulse the fuel synchronous with the fuel flow sensor strategy = less towards cheating
2. Introducing high freqency noise nearby the fuel flow sensor to disrupt the reading / raise the noise floor = more towards cheating
3. Porous/semi-porous/elastic metal fuel lines that accumulate fuel in some controlled pressure condition = more towards cheating

It's certainly not easy and I can't really explain why I feel differently about those strategies other than some immate feeling of 'cool' about them.

What do you suppose is the sampling frequncy of the sensor?

The injectors pulse width modulation could be around 200kHz or maybe more.

The mechanical pump is too slow compared so I am assuming the closing of the fuel injectors were used to synchronise the fuel flow in phase with the sampling rate.

This pathway is also brings many questions. Because there are six fuel injectors out of phase by whatever uneven firing phases comes with a 90 degree V6, all these pulses would superimpose as they reflect back to a single pipe to the fuel sensor. Not to mention the pulses have to survive going through the cam driven fuel HP pumps if the sensor is located before the fuel HP pumps which I doubt.

The only way I see using the injectors to do the trick is if the fuel sensor(s) was located after the HP fuel pump one on each engine bank. If it were before the fuel pump, it would be too much obstacles between injector and flow sensor.

What went through my mind at the time was- fuel fed to 'the system' and not used would have to be returned to some area where it could cool or 'de-froth' or there would be air in the injectors. I imagined (at the time) that there may have been a way to reverse flow this supplying more flow for the injector rail.

Electronically, the only way that comes to mind is not the frequency of the signal but wave shape or height.
Could they slip it through in the way 3 phase AC is displayed? the pulse is the same duration.

[63l8qrrfy6]

Interesting that Ferrari had to do an air refill on LEC.
If the valve pneumatic seals are gone that's a new ICE I think, unless the leak was outside the engine.

[NL_Fer]

Could one charge the pneumatic tank with a combustible gas?

[PlatinumZealot]

That's a bomb right there.
The last time a tank exploded wasn't pretty and that was only Nitrogen.

Say this combustible gas... Or even safer and thus better... "lubrication oil" for the pneumatic system.. "leaks" through the valve seals and is inducted in the combustion chamber.. That is a good idea you have there.

[Big Tea]

That's a bomb right there.
The last time a tank exploded wasn't pretty and that was only Nitrogen.

Say this combustible gas... Or even safer and thus better... "lubrication oil" for the pneumatic system.. "leaks" through the valve seals and is inducted in the combustion chamber.. That is a good idea you have there.

Does it have to be combustible? Can it not just have an extremely high expansion rate or be an oxidiser?

(whaaay, get my foil hat out of the box again )

[PlatinumZealot]

What went through my mind at the time was- fuel fed to 'the system' and not used would have to be returned to some area where it could cool or 'de-froth' or there would be air in the injectors. I imagined (at the time) that there may have been a way to reverse flow this supplying more flow for the injector rail.

Electronically, the only way that comes to mind is not the frequency of the signal but wave shape or height.
Could they slip it through in the way 3 phase AC is displayed? the pulse is the same duration.

To my understanding the signal reads the instantaneous velocity of the liquid particles.

When the injector opens the flow rate rises up like a linear ramp and the reverse occurs for closing. About 0.3 ms to open /close. I guess this opening time depends on the lift and springrate of the armature. It seems to take about 1ms delay of applying current to begin lift of the armature.

So the fuel flow would be pulses of this almost square wave form. The cylinder timings do not overlap so in theory you can send an instantaneous 100kg/hr to each injector. However those who work with flowmeters know that the flow meter can display the mean flow rate over whatever interval you want. So If you do infact inject 100kg/hr in 2 milliseconds of injector opening, what average will this show on the flow meter? The total cycle time over two revs is 10ms at 12,000 rpm. So the average is 10kg/hr per injector and you have six cylinders... So fuel flow meters sees 60kg/hr at that pulse width. You can increase the pulse width some more to hit the average 100kg/hr.

I suppose the fuel pressure is varied to get smaller injections.
Just playing with rough numbers here.

Anyway... So.. Is 2ms enough resolution to trick the flow meter? I hear these injectors can fire multiple times. The pistons are shaped so that they influence the fuel spray from the injectors. Obviously this only has effect close to TDC on the induction stroke. The injectors have some amount of time here to work. Too early and fuel will be sprayed out the exhaust, too late, you miss the piston top, and you get poor combustion. 5ms total time to work in (approx 360 degrees crank) how many times can an injector open and close here? Maybe the injectors have multiple armatures/ or staggered pintles? This could be Ferrari's "double anchor" innovation perhaps.

So.. What i am saying is that the injectors would have to be basically well close to the sampling frequency of the flow meter for the Aliasing to work... And so far its at most maybe 200 to 1000Hz... Waaay too slow to trick the flow meter...

This thesis leads to me to believe Ferrari was doing something else to trick the flow meter.

[Big Tea]

What went through my mind at the time was- fuel fed to 'the system' and not used would have to be returned to some area where it could cool or 'de-froth' or there would be air in the injectors. I imagined (at the time) that there may have been a way to reverse flow this supplying more flow for the injector rail.

Electronically, the only way that comes to mind is not the frequency of the signal but wave shape or height.
Could they slip it through in the way 3 phase AC is displayed? the pulse is the same duration.

To my understanding the signal reads the instantaneous velocity of the liquid particles.

When the injector opens the flow rate rises up like a linear ramp and the reverse occurs for closing. About 0.3 ms to open /close. I guess this opening time depends on the lift and springrate of the armature. It seems to take about 1ms delay of applying current to begin lift of the armature.

So the fuel flow would be pulses of this almost square wave form. The cylinder timings do not overlap so in theory you can send an instantaneous 100kg/hr to each injector. However those who work with flowmeters know that the flow meter can display the mean flow rate over whatever interval you want. So If you do infact inject 100kg/hr in 2 milliseconds of injector opening, what average will this show on the flow meter? The total cycle time over two revs is 10ms at 12,000 rpm. So the average is 10kg/hr per injector and you have six cylinders... So fuel flow meters sees 60kg/hr at that pulse width. You can increase the pulse width some more to hit the average 100kg/hr.

I suppose the fuel pressure is varied to get smaller injections.
Just playing with rough numbers here.

Anyway... So.. Is 2ms enough resolution to trick the flow meter? I hear these injectors can fire multiple times. The pistons are shaped so that they influence the fuel spray from the injectors. Obviously this only has effect close to TDC on the induction stroke. The injectors have some amount of time here to work. Too early and fuel will be sprayed out the exhaust, too late, you miss the piston top, and you get poor combustion. 5ms total time to work in (approx 360 degrees crank) how many times can an injector open and close here? Maybe the injectors have multiple armatures/ or staggered pintles? This could be Ferrari's "double anchor" innovation perhaps.

So.. What i am saying is that the injectors would have to be basically well close to the sampling frequency of the flow meter for the Aliasing to work... And so far its at most maybe 200 to 1000Hz... Waaay too slow to trick the flow meter...

This thesis leads to me to believe Ferrari was doing something else to trick the flow meter.

I recall reading a while back of injectors having more than one 'position' I think it was double stop or similar?
Could this be used to allow two intakes of fuel, one prior to the 'natural' one but at a point where flow is low due to the filing angles? I know the amount of fuel 'stored' in the injector would be small, but even a 5% increase would be 5% more fuel flow than anyone else.

Edit.
Just found it, it was double anchor, and the 'benefit' of it was not allowing too much fuel to be injected.

[dfegan358]

In terms of power unit development for 2021? How restricted are Ferrari now in making improvements?

Have the power unit regulations for 2022 been confirmed? Are they likely to be much different from 2021 regs?

[nzjrs]

viewtopic.php?f=4&t=28491 is the previous discussion of methods to trick the FFS

[ScrewCaptain27]

Ferrari is suffering from substantial clipping in Monza:

[MtthsMlw]

Ferrari is suffering from substantial clipping in Monza:
https://uploads.tapatalk-cdn.com/202009 ... a1f4ad.jpg

That's quite dramatic. Not jumping to conclusions (engine/SOC mode can't be ignored) but their H looks like it took a TD hit as well.
Let's see if that stays like this for Quali.

[aleks_ader]

Ferrari is suffering from substantial clipping in Monza:
https://uploads.tapatalk-cdn.com/202009 ... a1f4ad.jpg

That's quite dramatic. Not jumping to conclusions (engine/SOC mode can't be ignored) but their H looks like it took a TD hit as well.
Let's see if that stays like this for Quali.

Logically if got hit into ICE u got hit also into H department. Less MGUK to MGUH tanking for example (in both direction really-harvesting and deploying regime). Less speed bread less heat recovery. Its vicious circle. Especially if they in till 2019 compensate lack of thermal efficiency with higher fuel flow trick. ERS trick is less clear to me. Also BAN (or more strict) oil or even coolant leaking could reduce the theirs overall ICE efficientcy and also harvesting potential. SO total system efficiency is lacking behind. Hence there they using mapping tricks to hide the losses on straight as much they can.

[PlatinumZealot]

Ferrari is suffering from substantial clipping in Monza:
https://uploads.tapatalk-cdn.com/202009 ... a1f4ad.jpg

Lift and coasting it looks like.
Before the braking point Ferrari isn't just slower, it seems to hit a brick wall. Most I can think of is lift and coast.

[Big Tea]

Ferrari is suffering from substantial clipping in Monza:
https://uploads.tapatalk-cdn.com/202009 ... a1f4ad.jpg

Lift and coasting it looks like.
Before the braking point Ferrari isn't just slower, it seems to hit a brick wall. Most I can think of is lift and coast.

Are Ferrari stil running the dual battery?

[henry]

Ferrari is suffering from substantial clipping in Monza:
https://uploads.tapatalk-cdn.com/202009 ... a1f4ad.jpg

Lift and coasting it looks like.
Before the braking point Ferrari isn't just slower, it seems to hit a brick wall. Most I can think of is lift and coast.

I don’t think it’s lift and coast. It’s a quali simulation so no need to save fuel.

They may have turned off the ES, or running ICE only charging the ES from the H or in extreme they may be charging from H and K simultaneously.

They would result in decelerations of roughly 0.1g 0.2g 0.35g respectively.

Lift and coast would be around 1g