Some restrictor research

Good - Aki always needs lots of ballast!

Oh gosh, rather surprising that the curves are so different in shape! I assumed it could only be related to the shapes of the torque curves until I actually looked at your curves and the variations between them didn't look... I dunno, entirely correlated with the restrictor behaviour...

[Edit: would be interesting to note the AI up-shift RPM for each car, and perhaps see if it varies (a) with gear and (b) with restriction. I'd guess probably "yes" for (a) and "no" for (b)...]

Btw, the AC built-in powergraph looks somewhat different to the CM one for the 599XX Evo - see pic below. [Edit: maybe it's just that the built-in curve isn't using the correct max RPM value, unlike CM which probably reads it directly from the engine INI file.]

 

See, this is why the quadratic changes make sense to me as a top end restriction:


If you look at a nice linear power curve like the Zonda has, then limiting it to, say, 40% less than max HP cuts out a triangle of power - the change in area under the curve is what leads to lap times being quadratic, as the area of that triangle is 1/2*rpm*power



Then for the 458, If it's based on percentage of the max power, it could end up being something like this - so for the first range of restriction, it's acting over a fixed amount of rpm, meaning the overall lap time loss is ~3000*power, and you just get a linear change.

I drew these to both reduce the same amount of power at 40% + 20% restrictors, not to match the actual curves of the cars though.

 

I like that idea, @Stereo - sounds plausible in terms of what's going on.
I've finished a very half-baked check on the 599XX Evo case, with just two runs so far - 0% and 100% restrictor.
First comment: 100% is not quite enough restriction to change the optimal upshift RPM to something lower than the limiter (very nearly enough though).
The (I repeat, half-baked) power and torque curves for those two cases are shown in the plots below. [Edit: effective power in this case is my not-fully-corrected power-to-tarmac figure.]

 

Hmm, that looks like 33% power loss at 9000 rpm, about 10% at 4000... so I guess it is triangular, but extends a lot further down the power curve. Maybe to 0% at idle, or at 0rpm?

 

Mmm, could be. I need to muck around a little bit more with the data to let me divide one curve by the other - will do that later (and also attempt to extend the data to lower RPM for fun).

 

OK, have generated a basic pic with the fractional power (or torque) available at 50% and 100% restrictor vs. RPM.
It's awfully linear-looking
(Can't really get much lower in revs cos the auto-clutch kicks in.)