2014-2020 Formula One 1.6l V6 turbo engine formula

[Abarth]

[[....]Finally, the GTX3076R turbocharger has been chosen because the smaller compressor was not going into surge at low engine speed while the smaller turbine made it possible to recover more energy.[...].

Using a standard turbocharger will result in too low power which could be harvested from the turbine.

In such an application, turbine design has not to be in line with compressor demand like in a stock TC, as you are going to divide turbine power into MGU-H and compressor, which is not the case in a usual arrangement.

[321apex]

the late Prof Gordon Blair was highly and specifically critical of such 'modelling' software
do people think this thesis represents what even Renault are doing ? (let alone Mercedes)

I believe Grodon Blair was a highly regarded expert in 2 stroke engines design and tuning working in Wisconsin.

Empirical engine modelling is only as good as test data available to calibrate it and validate it. Ricardo, AVL, FEV and few others have made their business in selling such modelling services among other "know hows" they possess.

AVL is especially well positioned to be able to gain access to engine test data to validate their empirical simulation tools. Of course they will never admit it, but the advanced dyno testing equipment they sell and service is used within high level racing industry and AVL as a supplier has confidential access to quite a bit of it. As such I would grant their analysis a huge benefit of any doubt I would ever have.

[Abarth]

That simulation was not done by AVL but by a software they sell.
With a stock TC.
Garbage input = Garbage output, was an old saying with software simulations....and it still holds true.

[mrluke]

From reading the article I can see very little if any mention of self sustaining mode. From the energy / boost flow diagram there appears to be no link between mgu-h and mgu-k.

[ringo]

Ringo figures sound correct to me. 200hp or 147Kw between 10k and 12k. Which turbine efficiency do you assume, Ringo?

Went with 85% which is typical i guess. I've held that at 85% for all calculations. My compressor is at 82%

The compressor of the study of that guy seems also very reasonable, averaging 102kW at the max airflow with max pressure ratios from 2 to 2.4 (which is 10000 to 12000rpm roughly).

If both are correct, one could recover 45kW between 10k and 12k, but model shows 12kw??

I would have expected something around 60Kw. But maybe there is a basic turbine limit I am overlooking?

must be related to the back pressure.

I find it intersting that they want the mguh behave as a motor all the way to 9000rpm.

[ringo]

As I said, part of this work was published in RCE Vol24 No5, p54-59; and I highly recommend you guys to check it out!

However, for the sake of this discussion, below are some figures from the article.

http://img.ly/system/uploads/008/242/05 ... al_ice.png
Figure 8 shows the obtained performance values of the engine without the recovered energy of the MGU-H, but it incorporates the assistance of the MGU-H to reach the fuel mass flow rate limit. The maximum power is 454.82kW @10,500rpm, while the maximum torque is 415.48Nm @7000rpm. It can be seen that the torque remains nearly constant from 5000 until 10,500rpm due to the MGU-H assistance.

@7000 rpm my simulation is around 416N.m at the flywheel. MGUH disengaged. It's not my peak value however, but i think that has to do with my "perfect" turbine and it's probably why these guys cut off below 5000 rpm. they may not know what is happening down there.

http://img.ly/system/uploads/008/242/05 ... l_mguh.png
The MGU-H power curve is illustrated in Figure 9. Before 9200rpm the MGU-H is acting as a motor while hereafter it is operating as a generator and recovering energy.

With the cars barely reving over 10,000 rpm i think it's a wrong assumption. Maybe if they expected the cars to rev to 15,000rpm then they may say motoring takes place up to 9,000. IMO it's only need to bring the turbine in operating speed range.

http://img.ly/system/uploads/008/242/05 ... 20mguk.png
Figure 10 displays the obtained performance with the addition of the MGU-K. The maximum power is 574.82kW @10,500rpm and maximum torque is 611.83Nm @5000rpm. It should be noted that the torque of the MGU-K @5000rpm had to be limited from 229.35 to 200Nm, because this is the maximum allowed torque according to the regulations. The power curve with MGU-K is a parallel line above the curve without MGU-K because it adds a fixed amount of 120kW.

My peak power is 457kW from the ICE flywheel, i suppose 120 more would bring it to 577 kW.


Not sure what these guys are doing, but it's seems awfully close to what i've been doing since this thread started, and i'm using excell as my simmulator. No professor or lab or program

I would much believe cosworth than these guys. I think they got their priorities wrong on the scope of operation of the MGUH.
Cosworth is an engine builder. These guys are just a bunch of students using some program as their main tool. They have the ICE part right, which i've agreed to; but so did Cosworth. However i would question their knowledge on the MGUH.

[ppj13]

Ringo figures sound correct to me. 200hp or 147Kw between 10k and 12k. Which turbine efficiency do you assume, Ringo?

Went with 85% which is typical i guess. I've held that at 85% for all calculations. My compressor is at 82%

The compressor of the study of that guy seems also very reasonable, averaging 102kW at the max airflow with max pressure ratios from 2 to 2.4 (which is 10000 to 12000rpm roughly).

If both are correct, one could recover 45kW between 10k and 12k, but model shows 12kw??

I would have expected something around 60Kw. But maybe there is a basic turbine limit I am overlooking?

must be related to the back pressure.

I find it intersting that they want the mguh behave as a motor all the way to 9000rpm.

I think this and the statement you make in the next post about mguh priorities are both right. They made the best PU with priority in maximum output, while one should design the PU with the priority in best efficiency. Their setup with the small turbine uses so much electric power (or, to be accurate, saves so little electric power), while giving a bit more peak power, perhaps.

And I cannot see many problems with backpressure. What's the deal? A little bit more EGR and lower VE, but that's okay because the forumla is not air restricted, it is fuel restricted. Probably the thing is temperature related?

[alexx_88]

There has been a lot of talk about using MGU-K as a generator to improve traction up to the limit where it could be considered a form of "practical traction control". However, since the SECU logs the torque demand of the driver and also the resulting torque at the wheel axle, won't anything resembling traction control be easily detected by the FIA? I mean, the logs would show something like:
T0 : Torque demand: 30% -> Torque at axle: X
T1: Torque demand: 35% -> Torque at axle: X - 10

Even without closed loop control, it doesn't seem like an impossible task to predict if a certain torque demand from the driver will determine traction loss or not and simply artificially lower it by moving the MGU-K into generator mode.

[Tommy Cookers]

your point is not quite clear to me .... but
ICE mapping allows eg with wheelspin the throttle plate to somewhat close automatically without any accelerator pedal movement

any real-world electromechanical application will be designed to have a similar inherent characteristic
ie reduced load causes increased rpm, but torque inherently falls with rpm rise, so acts substantially to damp rpm rise
as in a vacuum cleaner, a car starter motor, or a toy trainset
anything else would be dangerous and difficult

this aspect (and behaviour in general) is improveable with real or synthetic feedback
any type of responsive motor/generator and drive anyway needs internal feedback to work eg encoder (position) feedback
the ECU is not in the control loops of the mgu-k, it only generates demands

what TC rules ban is any use of wheel rpm difference ie spin
not a natural or mapped fall in torque (or even torque reversal) with feedback of a (rising) motor rpm
the rules also allow ignition cuts based on mapping

the mgu-k and the ICE are common and drive a common load
the mgu-k has a lesser capability than the ICE to act for banned TC (it's less responsive and less powerful)
but can legally help car control rather as ICE mapping does
to maximise this help it should have inherent 2-quadrant characteristic (no discontinuity between motoring and generating)
though any mapped equivalent is legal

to require the mgu-k to be car control-neutral would be a much bigger can of worms for the teams and the FIA

[321apex]

The rear axle is already instrumented for the purpose of modulation of braking/charging torque. As such at each entry into any corner under braking, the software can easily formulate definition of surface traction potential (coeff. of friction). This definition can be used by the software to command the PU for necessary torque exiting the corner.

There could be spark retardation or fuel enleanment strategy in engine mapping to cause the engine to limit torque output. Engine doesn't have to misfire to reduce torque. I won't get into the MGU-K control strategy of torque control since it is elementary engineering logic.

All this WITHOUT close loop.

In my view, this is how it's done. Since this "surface traction potential" is estimated each time, if the weather conditions change so does this "surface traction potential".

I will just add, that exiting corners, the ICE is usually off natural boost and MGU-H augmentation to spin turbine have to be excercised and may also be part of the software kit for static TC. The MGU-K has the instant torque, which coupled with momentarily sleepy ICE is enough to squirt the race out of corners.

[Abarth]

I think this rule will not come in handy if you want to have a rpm dependent loss of torque at constant pedal position:

5.5.5 At any given accelerator pedal position and above 4,000rpm, the driver torque demand map must not have a gradient of less than – (minus) 0.045Nm/rpm.

[Tommy Cookers]

torque falling by 45 Nm when rpm rises by 1000 ?
broadly, this is limiting generation as a control aid under traction
but requiring only that the combined PU acts similarly to 2013 mapped F1

I never thought otherwise or suggested that torque falling more steeply with rpm is practicable without software in the control loop
this would likely be ruled as TC and so illegal (the software functionality amounting to synthetic wheelspin sensing)

but this rate rule does not suggest to me that the mgu-k motor torque should not fall with rpm rise (accelerator fixed)
or that the generation quadrant could not help the response of the PU in following this rule
IMO some feedback (closed control loops) is necessary for this, and there is no rule against them
this should be good for car control

the rules seem directed at a rather crude system reliant on mgu-k torque measurement and 'natural' motor characteristics
nominally it would be torque-controlled (and partly torque-limited) up to some rpm, beyond this power-limited (falling torque/rpm)
ie as rpm rise the current will fall because the fixed limit of supply voltage is reached
it should work quite well, in part using mapped torque demand and feedback from the torque transducer
at these higher rpm this is a somewhat vague and unresponsive system
below these rpm it would realise most of the car contol benefits implied by the torque gradient rule

[321apex]

The rules seem to be written by people whose imagination goes beyond their technical understanding/comprehension and practical/objective means to measure or control such parameters. Specifically what method is there to determine whether there was 45 or 47 or 43 Nm torque rise ?

This is THE WORLD of transients, which happen very quickly. One may say telemetry will spot that, well at what sampling frequency? 50Hz, 200Hz, 1kHz ? What about noise, aliasing? It's a mess to break out any meaning from data even if all bring good intentions.

Torque isn't so simple to measure accurately and certainly isn't something that is readily instrumented on board a moving race machine. I think there is more of these. How do they police the position of CG in a race car?

I sympathize with Red Bull and their fuel flow measurement plight. However for political reasons, the FIA will reject their appeal in a futile effort to prove that morons were not formulating this years F1 regulations.

Just remember, that FIA may think that it makes the rules and it makes the interpretations, however there is a regular court which may override the FIA if a claim is strong enough and the entity is politically astute enough and has money. This will be something to see if FIA has to prove that a winning car had CG by 1 mm too low. LOL

[xpensive]

The rear axle is already instrumented for the purpose of modulation of braking/charging torque. As such at each entry into any corner under braking, the software can easily formulate definition of surface traction potential (coeff. of friction). This definition can be used by the software to command the PU for necessary torque exiting the corner.

There could be spark retardation or fuel enleanment strategy in engine mapping to cause the engine to limit torque output. Engine doesn't have to misfire to reduce torque. I won't get into the MGU-K control strategy of torque control since it is elementary engineering logic.

All this WITHOUT close loop.

In my view, this is how it's done. Since this "surface traction potential" is estimated each time, if the weather conditions change so does this "surface traction potential".

I will just add, that exiting corners, the ICE is usually off natural boost and MGU-H augmentation to spin turbine have to be excercised and may also be part of the software kit for static TC. The MGU-K has the instant torque, which coupled with momentarily sleepy ICE is enough to squirt the race out of corners.

I read the above with great fascination as "torque" is mention half a dozen times, when traction control is about
wheel-power control and actually not even that, it's about traction-force control. The engine torque is irrelevant.

But hey, whatever turns you on.

[irsq4]

I think this rule will not come in handy if you want to have a rpm dependent loss of torque at constant pedal position:

5.5.5 At any given accelerator pedal position and above 4,000rpm, the driver torque demand map must not have a gradient of less than – (minus) 0.045Nm/rpm.

Does that means this: http://postimg.org/image/mqmcnnf7v/