F1technical.net

From an engine tuning perspective what they're doing is sheer insanity.

A regular B18 Integra GSR engine(1.8l 4 pot) has a 140lph fuel pump which is the same flow rate these cars use. No matter what port fuel injectors you use, you're limited by the fuel pump, with the stock pump you can't run more than ~6 psi of boost to maintain a safe .8 lambda. At 10.7:1 CR you're making about 340bhp @ 7,800 rpm on a GT35 ebay turbo. These engines are using the same fuel pump but running 45 psi of boost with 14:1+ CR making over 900hp, that the engines aren't exploding from detonation or are running at all is incredible when compared to what the budget build kids are doing.

There's a slight difference between a few hundred dollars and hundreds of millions of dollars.

dren wrote: ↑

02 Apr 2018, 20:07

There's a slight difference between a few hundred dollars and hundreds of millions of dollars.

Obviously, I just thought it would be interesting to highlight the difference. As someone that does the fab work, that is really the biggest cost of doing a turbo kit. The manifolds, exhaust, dump pipe, intercooler piping, and in some cases the end tanks cost more than the hardware itself sometimes.

If they run constant constant fuel rate, varying the boost to maintain a constant lambda then when they shift up they need to very rapidly increase boost, by maybe 0.5 bar. There are a number of inertias to overcome to achieve this so I wonder if they increase boost as a shift approaches, running a little leaner, perhaps sacrificing power for driveability. This might be particularly useful if the upshift occurs in a corner when they could target having the same power immediately before and after the shift.

In the data of Ferrari at Singapore the time in 7th gear is less than 2 seconds. I have previously done some calculations on the acceleration rate for the TC MGU-H assembly. With a driving power of 150kw I found an acceleration rate of 6000 rpm/sec. I would guess that this is in the ballpark for a change of 0.5 bar. So managing lambda with boost, particularly through upshifts may well involve quite a few compromises.

Henry,
re Acceleration rate, is that per second per second? or just per second?

johnny comelately wrote: ↑

02 Apr 2018, 22:33

Henry,
re Acceleration rate, is that per second per second? or just per second?

It’s revs per minute per second.

henry wrote: ↑

02 Apr 2018, 22:49

johnny comelately wrote: ↑

02 Apr 2018, 22:33

Henry,
re Acceleration rate, is that per second per second? or just per second?

It’s revs per minute per second.

OK< thank you...schools out now

There's a rising fuel rate until 10,500 rpm, I don't think they run the same boost at all RPM's nor do I believe they run constant fuel rate, nor do I think MGU-H works all that much in motor mode. The MGU-H only works to kick start the turbo, ideal boost for a given engine speed could be achieved with as little as 30kW from the MGU-H, the exhaust energy does the bulk of the work. At peak engine torque there's easily 120kw for the turbine in the form of exhaust energy. Turbos are positive feedback devices, the more energy you put in the turbine, the more boost you make which gives you even more turbine energy.

Hmm clearly these cars have blow off valves, I wonder if the boost is sent to the atmosphere or re-introduced into the air box.

imo
shifting eg upshifts is done by timed defueling for c.20 milliseconds to detorque the dogs for the shift
the K is for c.20 milliseconds at full generation (demand) to contribute the best it can to slowing the PU to speed the shift

similarly for downshifts the K is at full motoring demand to contribute its bit to speed up the PU
the fuel is cut for milliseconds to detorque the dogs and then brought back in to increase the PU rpm
the fuel cutting and restoration timing pattern for downshifts will be different to the pattern for upshifts

the K response is about as quick as the ICE's
if the K was not energised the shifts would be slower as its inertia would be parasitic

Tommy Cookers wrote: ↑

02 Apr 2018, 23:43

imo
shifting eg upshifts is done by timed defueling for c.20 milliseconds to detorque the dogs for the shift
the K is for c.20 milliseconds at full generation (demand) to contribute the best it can to slowing the PU to speed the shift

similarly for downshifts the K is at full motoring demand to contribute its bit to speed up the PU
the fuel is cut for milliseconds to detorque the dogs and then brought back in to increase the PU rpm
the fuel cutting and restoration timing pattern for downshifts will be different to the pattern for upshifts

the K response is about as quick as the ICE's
if the K was not energised the shifts would be slower as its inertia would be parasitic

Maybe not total defuelling, could be half?? to keep inlet wettish. total defuelling can give some strange driving characteristics

"inlet wettish" Johnny? How so?

These are very lean run DFI machines, which are so 'dry' - you seldom see visible flame in the exhaust..

J.A.W. wrote: ↑

03 Apr 2018, 05:53

"inlet wettish" Johnny? How so?

These are very lean run DFI machines, which are so 'dry' - you seldom see visible flame in the exhaust..

Of course, you are right and I am forgetting we are talking DFI here, sorry.
BUT..i am suspicious that if it was a zero setting for the change with the ignition cut (would it be 4 revolutions at 20 milliseconds at 11Krpm)?) that may give hesitation?

J.A.W. wrote: ↑

03 Apr 2018, 05:53

"inlet wettish" Johnny? How so?

These are very lean run DFI machines, which are so 'dry' - you seldom see visible flame in the exhaust..

In port injection engines, it creates a problem (I seem to be stuck in the past )

I expect that since F1 combustion is now so synergistic between fuel & ignition events, both are involved..

Tommy Cookers wrote: ↑

02 Apr 2018, 15:53

I can't see how a substantial variation of lambda (eg 1.2 - 2.0 steady state) is useful

1.2 - 2.0 is just the error band for my guess!