Engine Learning

All that has to do with the power train, gearbox, clutch, fuels and lubricants, etc. Generally the mechanical side of Formula One.
hardingfv32
hardingfv32
32
Joined: 03 Apr 2011, 19:42

Engine Learning

Post

Quote from Horner: "the engines have to go through a learning process as well to get to an optimum state".

What might he be talking about?

Brian

User avatar
godlameroso
309
Joined: 16 Jan 2010, 21:27
Location: Miami FL

Re: Engine Learning

Post

That the cars learn where to harvest and where to deploy in real time as the car turns laps, this in turn gives feedback to the engineers regarding what ICE power settings are safe, how much boost to run, how lean you can run etc. Humidity, ambient temperature, track temperature, and atmospheric pressure all affect how a power unit runs. So there's a lot of optimization going on during the run up to parc ferme.
Saishū kōnā

Greg Locock
Greg Locock
233
Joined: 30 Jun 2012, 00:48

Re: Engine Learning

Post

Adaptive tuning has been used on production cars since about 1996, from memory. That's why your gear changes and idle are a bit rough for the first few miles after a hard reset.

Brian Coat
Brian Coat
99
Joined: 16 Jun 2012, 18:42

Re: Engine Learning

Post

Yes, adaptive tecnology was certainly in use by automotive OEMs ECUs in late 1980s.

The way a modern slushbox uses adaptions and real time closed loop control to achieve a seamless shift is one of the unsung wonders of the automotive world. It has been one of the key enablers for those auto gearboxes with many (8+) gears - a technology that actually improves real world economy not just test cycle economy.

Back on F1, here's one:

Does anyone think it would also be a good idea to run adaptive impellor/housing clearances?

State-of-art GT practice is to cool the HP turbine housing using a (modulated) cooling air supply - it "shrinks", reducing clearances and improving efficiency.

Since the turbo only has one compressor stage, you could try it here too.

You could even have a "minor rub detection" algorithm based off shaft rpm to allow you to wind the housing down to the "kiss" point then back off?

Or maybe I should take a realism pill ... :lol:

Tommy Cookers
Tommy Cookers
621
Joined: 17 Feb 2012, 16:55

Re: Engine Learning

Post

Brian Coat wrote:Yes, adaptive tecnology was certainly in use by automotive OEMs ECUs in late 1980s.
The way a modern slushbox uses adaptions and real time closed loop control to achieve a seamless shift is one of the unsung wonders of the automotive world. It has been one of the key enablers for those auto gearboxes with many (8+) gears - a technology that actually improves real world economy not just test cycle economy.
unsung wonder indeed - eliminating 'shift hunting' usual in small automatics and facilitating/enabling the converter to lock in cruise

though if you drive steadily and slowly for long enough and then need full acceleration from near-rest it will adaptively choose too high a gear
and afaik some say their 8 speeds have never seen higher than 6th

another plus re automatics is that the engines are identical to manuals wrt power curve, not needing an automatic-specific one ?