2021 Engine thread

[MrPotatoHead]

Indeed TC.

When you try to control the motor with a PWM like mentioned you inherently just drop the average torque seen at the output of the motor (Power) and so it would just seem like a smaller motor.

[roon]

Is it about filling in the torque curve, or filling in between combustion events? It could be that it smooths the power delivery of the ICE that it is attached to. The electric motor is used to fill in the gaps between combustion strokes. Thus a hybrid power unit can deliver constant torque like an electric motor.
In F1, the K is appropriately sized for this; 160hp being more than what a single cylinder can deliver (~140 hp). But is the K still selectively deployed? I don't think they're pulsing the K on and off for an entire lap.
But if they are, the smooth power output might help the longevity of the transmissions that we see. It may even influence chassis design. The 90 deg cylinder bank angle being better for installation into the chassis, but not ideal for a V6 engine. But it doesn't matter--torque fill could make the bank angle irrelevant.

roon is asking for K behaviour better than an extra V6 ICE slaved in ideal phase to the ICE ie to better an even-firing V12
doesn't electric torque filling of ICE cyclic torque need a much bigger electrical machine than a 160 hp MGU-K ?
ie a machine whose peak torque matches the in-cycle peak torque of a cylinder
the 'size' of the electrical machine and what drives it being closely related to peak torque

this difficulty is additional to those PHead has just mentioned

super responsive electric machines here are not as 'constant' as roon hopes - these characteristics are mutually exclusive
to produce by in-cycle torque filling a constant torque is particularly demanding
these (synchronous or reluctance) machines are in effect magnetic gears
we already need control miracles to prevent these 'gears' slipping given the ICE rpm/torque excursions of shifting etc

why beat ourselves to death for no benefit ? - given the load path to Earth is a low-pass 1 Hz mechanical filter

the K's dynamic response in-cycle will be much worse than the ICE's mechanically but not worse electromagnetically
ie it could produce torque pulses as rapid as the combustion torque pulses if (as is) there is a dominant velocity-related load
such torque pulses are an abnormal requirement and may be impossible electrically without some shortfall for various reasons

Thanks for the interesting response, TC. Even in an ideal even firing engine, there should some peaks-and-valleys in the output, resulting from the variation in force applied by any single cylinder unto the crankshaft i.e. the torque curve of a single cylinder, peaking at TDC and falling til 90*-beyond-TDC towards nil at BDC (just my intuition, I'm not exactly sure what the torque curve on the expansion stroke looks like).

If you evenly space these waveforms there will still be peaks-and-valleys present, no? Perhaps negligible, especially in high cylinder-count even-fire mills like you mentioned.

why beat ourselves to death for no benefit ? - given the load path to Earth is a low-pass 1 Hz mechanical filter

I can't keep up with your cleverness here. A reference to tires, transmission?

[DiogoBrand]

I don't know if this was discussed already. But I saw somewhere that for 2021 the FIA is considering an increase of the fuel flow limit to 120kg/h, so the engines can run at higher speeds to make the noise better. But that got me thinking:
If they want to make manufacturers run their engines at higher speeds, can this not be achieved simply by reducing displacement or creating a boost limit? Will 120kg/h not make engines too powerful?

[gruntguru]

why beat ourselves to death for no benefit ? - given the load path to Earth is a low-pass 1 Hz mechanical filter

I can't keep up with your cleverness here. A reference to tires, transmission?

A reference to everything I think. Inertias at the engine end, elasticity in the driveline (including tyres). Not sure if 1Hz would be the cutoff frequency for an F1 - given the extreme lightness of everything that moves.

Don't forget the fundamental excitation frequency you are suggesting they remove is 500 hz @ 10,000 engine rpm.

[roon]

Don't forget the fundamental excitation frequency you are suggesting they remove is 500 hz @ 10,000 engine rpm.

Yes, three times per rev. This being too high frequency for control? Or were you considering something else? At lower revs, not really an issue for F1, there would perhaps be more a noticable effect. Consider the uptake of pendulum mass dampers in production cars. Could their effect be simulated by electric motor action?

F1's K is geared something like 8:1 to the crankshaft. Does this aid motor control? Seems like it would.

[63l8qrrfy6]

Don't forget the fundamental excitation frequency you are suggesting they remove is 500 hz @ 10,000 engine rpm.

Yes, three times per rev. This being too high frequency for control? Or were you considering something else? At lower revs, not really an issue for F1, there would perhaps be more a noticable effect. Consider the uptake of pendulum mass dampers in production cars. Could their effect be simulated by electric motor action?

F1's K is geared something like 8:1 to the crankshaft. Does this aid motor control? Seems like it would.

1.5 times for uneven firing. The K is geared at about 3.3:1 so I think even with a single pole pair it could produce a sufficiently high frequency.

[Tommy Cookers]

the combustion torque pulses last about 1 millisec and are separated by about 1 millisec
filling the gaps with electrical torque pulses requires similar pulse duration - about 1 millisec
you cannot properly drive such a short current pulse through the armature because of its inductance
ie the electrical time constant will be maybe 2 or 3 millisec
and if motor acceleration within the pulse is required its inertia may additionally impede the performance
the electromechanical time constant (conventionally the response time under acceleration) being about 20 millisec

is the purpose of this torque fill to eliminate crankshaft torsional vibration at source ?
such vibration is also engendered by reciprocational inertia (depending on engine configuration)
and occurs simultaneously at multiple frequencies

we have an electrical machine with a relatively or very smooth torque output/input
and are apparently trying to control it to emulate the ICE's series of explosive torques
it may be possible to design a machine and drive that inherently emulates the ICE ie has very discontinuous ('explosive') outputs
this would be in many respects a very 'big' machine - and a much less efficient machine

[OO7]

I was of the opinion that torque filling was simply a method of using the MGU-K to smooth/fill the troughs in the torque curve.

[NL_Fer]

V6 = 3 combutions pulses per crank revolution. Also 3 mgu-k revoltion per crank revolution. Gearing is fixed, why would they need to puls the torque from the CE? They could built it in the mgu-k by design and synchronize when fitting the drive gear on the mgu-k shaft. But really is this all needed and would it not cause vibration on the mgu-k itself?

I don't know if this was discussed already. But I saw somewhere that for 2021 the FIA is considering an increase of the fuel flow limit to 120kg/h, so the engines can run at higher speeds to make the noise better. But that got me thinking:
If they want to make manufacturers run their engines at higher speeds, can this not be achieved simply by reducing displacement or creating a boost limit? Will 120kg/h not make engines too powerful?

120kg/h is to compensate for the losses by dumping the mgu-h. No more exhaust energy recovery, more fuel consumption needed. They still need to think of additional rules to increase rpm. For example increase fuel flow with rpm, reaching max flow only at max rpm. (120kg/h at 15000rpm). Limit boost, Turbo size or compression ratio.

[roon]

120kg/h is to compensate for the losses by dumping the mgu-h. No more exhaust energy recovery, more fuel consumption needed. They still need to think of additional rules to increase rpm. For example increase fuel flow with rpm, reaching max flow only at max rpm. (120kg/h at 15000rpm). Limit boost, Turbo size or compression ratio.

So it remains an efficiency formula with a different flow rate & curve. All the difficulties associated with combustion optimization remain. Merc continue to lead.

[gruntguru]

Don't forget the fundamental excitation frequency you are suggesting they remove is 500 hz @ 10,000 engine rpm.

Yes, three times per rev. This being too high frequency for control?

Impossible to control but more to the point - almost entirely removed by the low-pass filter (engine rotating inertia driving an elastic driveline) so the tyres don't feel this.

[gshevlin]

I find the provisional decision to scrap the current MGU-H, and the clear issues that most of the current-spec engine designers and builders have been having in creating an MGU-H that will last the expected design life, to be interesting.
Back in the 1950s the Wright Aircraft Company, looking to boost the power of the R-3350 radial engine, created a device known as the PRT (Power Recovery Turbine) which was a collection of 3 devices bolted to the exhaust header groups that converted the heat and energy of the exhaust gases back to energy applied to the crankshaft by means of fluid couplings. The PRT, when applied to an engine in high boost mode, was capable of returning in excess of 500 bhp to the engine, raising the maximum horsepower from 2700 up to 3300 for some R-3350 variants, using 150-grade fuel.
The challenge was, as with the MGU-H, durability. At low boost levels, the PRT assembly would just about survive, at high boost it tended to fail quite quickly (the ground engineers looking after R-3350 engines swiftly renamed the assembly to "Parts Recovery Turbine").

[NL_Fer]

120kg/h is to compensate for the losses by dumping the mgu-h. No more exhaust energy recovery, more fuel consumption needed. They still need to think of additional rules to increase rpm. For example increase fuel flow with rpm, reaching max flow only at max rpm. (120kg/h at 15000rpm). Limit boost, Turbo size or compression ratio.

So it remains an efficiency formula with a different flow rate & curve. All the difficulties associated with combustion optimization remain. Merc continue to lead.

Nog quite. Allot of Merc gains are made by the mgu-h. Yes it also needs superefficient combustion, but they need the mgu-h to make the energy usable at the wheels. A significant part of Mercs advantage will be lost. Why would Toto be lobbying for the mgu-h to remain?

[Tommy Cookers]

the Wright Turbo Compound had like all supercharged aircraft engines a low CR to allow very high boost for takeoff power
ie its CR was sub-optimal for typical (cruise) use and so its exhaust was more than optimally energetic
but F1's CR is not sub-optimal
we can speculate what efficiency (sustained power) F1 PUs gain by their compounding (freely, without extra fuel)
eg 8% sustained 'free' PU power gain needs 8% more fuel when without MGU-H

btw
at 200 mph a conventional engine's exhaust provides jet propulsion power equivalent to about 2% of crankshaft power
our heat dilution F1 engines have an outstandingly high relative massflow
so give 3 - 3.5% jet propulsion power at 200 mph ?
of course not much at 60 mph etc

[NL_Fer]

They should face the exhaust to the front side. The more exhaust gasses, the more drag