The red bull man said that one of his cars had its time disqualified for breaching the maximum 100 kg/h fuel flow rules by 0.02 grams and that had happened because the car had a tow and went into the speed limiter. he did not say for how long a time the fuel flow breach happened. going into the speed limiter means the engine was rotating at over the maximum power speed imposed inderectly by the maximum fuel flow rules. Those 0.02 grams combusted over and above what is allowed having been burned/combusted, they had first went through the FFS and than they had been injected into the combustion chambers by the fuel injectors.The FFS policing the fuel flow had dedected all that by sampling fuel flow through it at the rate of 4000 times a second.
Given that the 100kg/hr is an absolute limit the teams must operate below that. I imagine that they must run dyno tests to check that with the dynamics of the whole system they keep, just, below the limit.
This may be an explanation for Baku. Perhaps Honda carefully calibrated up to N RPM but a double tow got them to N+ at which the uncalibrated flow rate was just over the 100kg/hr.
This could be the basis of a tuning methodology. Hook up the FFS and take feedback from that to establish maximum flow rate at the injectors over the operating range. They might also vary the componentry and run this protocol again until they maximise power. I think this is essence of “grey area”.
The change to calibrate the FFS using measured weights would expose this is if the tuning meant the actual flow at the injectors was, on average, above the 100kg/hr. Unless, of course, they managed flow elsewhere in the rev range.
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
The FFM makes measurements at intervals of about 0.455ms. At 11krpm, about 12 FFM measurements happen during any revolution. Smoothing out flow rate fluctuations would have indeed to take place in the HP section, as @gruntguru said, unless the max flow rate through the FFM was kept below the legal limit by a margin that takes into account significant spikes. The uncertainty declared by Sentronics is 0.25% of the reading with 95% confidence interval. There must be a tolerance set by FIA, based on these specs. Any relatively large spike would need to have a duration of less than 0.455ms and happen to go through the FFM between measurements not to be detected.
For the max flow rate limit, there is no need to reconstruct the whole flow waveform exactly, as it's only the crossing of the threshold that matters. Concerning the measured cumulative mass flow, it is eventually checked vs fuel weight difference anyway.
According to Sentronics datasheet, the FFM CAN channel runs at 1Mbps, with the standard 11-bit-identifier protocol. In each message, 51b should be taken by support info and up to 8B by data. Just to make up some numbers, between measuments at 2.2kHz, it should be possible to send 4 messages carrying a total of 24B of data quite comfortably. Also, the ECU software may flag anomalies or breaches only, and low priority broadcast is surely sufficient (it doesn't really matter if an infringement is communicated to race control 30s later). Besides, the FFM has built-in max/min log.
Is it correct to say that at 12000RPM a four stroke 6 cylinder ICE does 600 "complete injections"* per second?
If so and assuming that there is only one common rail, I expect that there are pikes of pressure and flow in the fuel rail at a frequency of 600Hz and the fuel flow measure part of them since it measure at 1900-2200 Hz.
Is it correct or the measures are averaged on longer periods to eliminate this fluctuations?
* for "complete injections" I mean the sum of all the micro injections makes for each cycle from each injector (I do not know if multiple injections are allowed at all in F1)
6 cylinders at 12000rpm = 600 combustions per second. Excluding the possible use of more than one injection per combustion, its 600 injections per second.
F1 ultrasonic fuel flow metering. there are two clauses under FIA technical list no. 45
The device shall be able to recover automatically and without action from an incorrect fluid composition (Air bubbles for example).
When unable to read flow the device must store the last correct data and count no flow until it is able to measure flow again.
Are these two clauses a potential loophole if a team could consistently cause a flow reading failure on demand e.g. by entraining gas by example causing pump cavitation (or other means) upstream of the sensor.
The last good reading is retained while flow rate is exceeded and then gas entrainment is stopped on demand when the flow is legal again.
Would this be possible? and during that fault condition does the FIA logging detect this failt condition of just a frozen last good reading and if it did detect the fault would a repeatedly faulty sensor condition be cause for exclusion?
Effectively blinding the sensor for short periods of time.
A high risk strategy perhaps, but if it can be done consistently (and they’d easily have the means to perfect it in the factory on a test bed)
I accept entrain gas may not be a good thing for Fuel injection, but clever injector inlet design may be able to remove entrained gas downstream of the sensor.
Last edited by Nbywater on 06 Mar 2020, 01:50, edited 3 times in total.
It is possible to apply destructive interference or sound cancellation to the singals inside the measuring tube? Similar to a bose noise canceling headphones? Sample the sounds, cancel the sound. Feed back a false message phased just enough to indicate a simulated fuem flow....
What I think is happening in the Ferrari Power Unit is their turbocharger design "flaw." Most turbos have very minimal oil leakage on the compressor side on the turbine housing due to the designs of its bearings and their rings. If this leaking oil can be used to aid combustion, there will be an increase in power while not altering the fuel flow parameters; as you have an alternative combustion agent which is engine oil.
Although the power gains will be minimal, if Ferrari has managed to "control" or design their turbo with an oil leak larger than their competitors, this oil will be fed to the engine via the intake manifold from the intercooler, resulting to an increase in power.
Technical Regulation:
7.5 Cooling systems :
The cooling systems of the power unit, including that of the air destined for combustion, must not intentionally make use of the latent heat of vaporisation of any fluid with the exception of fuel for the normal purpose of combustion in the engine as described in Article 5.10.3.
So the word in question is "intent," I think as long as this oil is not being metered then the intent can not be proven. If the design of the turbo makes the rings leak more than normal, this will be considered as a design problem rather than intention to add more "combustible agent."
