Red Bull exceed fuel flow limit, Ricciardo DSQ at Australian GP

[Richard]

Oh dear we're back to "cheats"

Game over.

[turbof1]

I'll reopen the thread. However: if I see anybody bad mouthing, it'll be closed again. For good. So please behave. If you don't have anything usefull to post, don't post at all.

[Harsha]

I heard RBR changed MGU-k Drive shafts are changed here in Malaysia and also in Australia I'm not 100% sure about that
But my question is after watching the Malaysia GP where RBR was using a bit more fuel than merc engines and i know that an MGU-K failure or may be the lack of full functionality may eat more fuel than required
So it is the cause of high fuel rates for them and pushing them to use more than required to compensate the BHP deficit?

[alexx_88]

I've PMed turbof1 to re-open the thread and he was kind enough to do it. The reason for this is this article I've found in Racecar Engineering which talks precisely about a solution proposed on this thread: calculating the fuel flow based on the ECU data: http://www.racecar-engineering.com/arti ... ow-meters/

Enjoy the read and please, as turbof1 suggested, let's keep it civilized. Nobody gains/losses anything significant on a forum.

[Pierce89]

To jz11: You keep talking about a pulsating reverberating flow. Are you implying that the fuel sensor is in the high pressure part of the system? I would've thoght that surely the FFM is on the low pressure side, where there would be a whole lot less pulsation etc.

[Tim.Wright]

Its on the low pressure side. Otherwise it would need 2 ports in it to measure send and return flow rates from the fuel rail.

Trying to cheat the flow sensor by pulsing at high frequency is totally unrealistic in my opinion. Its risky because it reduces your noise margin, and it will be visible what you are trying to do since this frequency will still be visible in the flow sensor data, albeit attenuated.

[Pierce89]

Its on the low pressure side. Otherwise it would need 2 ports in it to measure send and return flow rates from the fuel rail.

Trying to cheat the flow sensor by pulsing at high frequency is totally unrealistic in my opinion. Its risky because it reduces your noise margin, and it will be visible what you are trying to do since this frequency will still be visible in the flow sensor data, albeit attenuated.

Thanks. I just wondered because jz11 kept saying the FFM was inadequate for measuring the pulsing flow in the fuel system, which made no sense to me because I thought surely the FFM was in the low pressure system rather than having to operate at 500 bar of the GDI systems.

[xpensive]

Why not just introduce a fuel-restrictor plate, like they do with air in Rally and Nascar, then monitor the pressure differential?

It means 100 kg/h, or 27.77 g/sec, needs a pressure differential of 75 Bar (1100 psi) to get through a 0.4 mm (0.0157") hole.

27.77 g @ 720 kg/m^3 density is a volumetric flow of 38.57 cc/sec

38.57 cc/sec through a 0.4 mm hole means a speed of 30.69 m/s

Δpf = 1.11 * Rho * v^2

1.11 * 720 * 30.69^2 = 7529 000 Pa, or 75.3 Bar, or 1092 psi

But that would obviously be way too simple?

[jagunx51]

http://www.racecar-engineering.com/news ... er-issues/

[djos]

http://www.racecar-engineering.com/news ... er-issues/

It's not such a big Deal in LMP1 simply because they have sensor redundancy - in F1 if the sensor is wrong and says you are using more fuel than you are, you're screwed .... as we have seen.

[jagunx51]

http://www.racecar-engineering.com/news ... er-issues/

It's not such a big Deal in LMP1 simply because they have sensor redundancy - in F1 if the sensor is wrong and says you are using more fuel than you are, you're screwed .... as we have seen.

the point is, the sensor is Gill's product.....so, i think what redbull experienced is justified with these cases
if it homologated, then it shouldn't happen right ?

http://www.racecar-engineering.com/news ... mologated/

[jz11]

Its on the low pressure side. Otherwise it would need 2 ports in it to measure send and return flow rates from the fuel rail.

Trying to cheat the flow sensor by pulsing at high frequency is totally unrealistic in my opinion. Its risky because it reduces your noise margin, and it will be visible what you are trying to do since this frequency will still be visible in the flow sensor data, albeit attenuated.

Thanks. I just wondered because jz11 kept saying the FFM was inadequate for measuring the pulsing flow in the fuel system, which made no sense to me because I thought surely the FFM was in the low pressure system rather than having to operate at 500 bar of the GDI systems.

having the pump on the low pressure side doesn't change much, since the high pressure pump is piston type, and it's pressure stroke is most likely tuned to coincide with a cylinder injection period, it will be source of one variable "noise" source on the low pressure side as the engine is going through rpm range, and there also may be a high pressure overflow valve, that returns some of the high pressure fluid back to input of the high pressure pump, that is also the same volume the FFM is supposedly sitting in, that pulse will be at somewhat different phase, but similar frequency to the pumps generated pulse, then there low pressure supply pump (might be multiple pumps actually) noise as well, low pressure overflow valve opening and closing, and on top of that, some noise from the high pressure will transfer to the low pressure side as vibration of the fuel encasing material and so on

and my theory is that not those individual pulses are the real problem, it is the combination of several of harmonics of those pulses that manifest as resonance of that volume that is causing erratic behavior of the sensor, and, since the flow isn't constant, and the incoming pulse frequency and amplitude is ever changing, you cannot create a logic inside the sensor to filter out reading affected by this noise*, thus getting inconsistent fuel flow measurements that are nowhere near accurate to the degree mentioned in the sensor datasheet, and all of those factors may be different for each team, and maybe car even

don't get me wrong, I'm not saying those sensors cannot be used here at all, its just that they should not be relied on as the main source of information to determine the actual fuel flow

* might even be better to not filter any readings at all, bump up base sampling rate to the max and include all the readings to calculate some average over extended (compared to the 0,2sec that they use now) period of time - 1sec or more even, this might help with the aliasing errors that were described here earlier - a band aid though, not the fix to the problem

[thomin]

According to AMuS, there's a new Technical Directive which says that from the Spanish GP onwards, manipulations on the fuel sensor will no longer be allowed.

Three teams (Red Bull, Toro Rosso and Lotus) are modifying their sensors due to packaging constraints and 95% of all problems have occurred with those teams. Apparently there were drillings into the winding which can get very close to the sensitive glass body of the actual sensor.

Another theory for why those three teams suffer from issues the most is that they all use fuel by Total.

[timbo]

Why not just introduce a fuel-restrictor plate, like they do with air in Rally and Nascar, then monitor the pressure differential?

It means 100 kg/h, or 27.77 g/sec, needs a pressure differential of 75 Bar (1100 psi) to get through a 0.4 mm (0.0157") hole.

27.77 g @ 720 kg/m^3 density is a volumetric flow of 38.57 cc/sec

38.57 cc/sec through a 0.4 mm hole means a speed of 30.69 m/s

Δpf = 1.11 * Rho * v^2

1.11 * 720 * 30.69^2 = 7529 000 Pa, or 75.3 Bar, or 1092 psi

But that would obviously be way too simple?

Is the dependence of the flow on the diameter quadratic? The problem might be that the tolerances are already low. Also, the variations of size due to temperature etc might contribute to the precision. It might turn out that there are no fewer problems IMO.

[andrewf1]

For anyone interested, there's a live-feed from the trial going on today here - http://ticker.auto-motor-und-sport.de/