Ferrari Power Unit Hardware & Software

[henry]

I don’t know if this has been posted before, but it’s a great read on the strategies that could be employed on the PU’s. It was written just prior to the 2014 season. Much of its contents are very much applicable today.

https://core.ac.uk/download/pdf/82194974.pdf

Thanks for that. It will be interesting to compare it with the 2013 paper by Prof. Lime beer et al,
“Optimal control of Formula One car energy recovery systems”, which supposedly was work for Ferrari.

I can’t now find a link to the original Limebeer paper, here’s a presentation about it https://www.city.ac.uk/__data/assets/pd ... 0-2014.pdf

Actually, my first link was good, but I gave the wrong link. So here is the original I wanted to do.
http://hybridfordonscentrum.se/wp-conte ... n_ver1.pdf
Both are very good reads.

Thanks again. When I get a quiet moment I’ll give them a serious read.

[henry]

Also @henry, “optimal control of formula one car energy recovery system:- D.J.N. Limebeer, G. Perantoni and A.V. Rao.”.
Also “McCabism:- optimal control theory and FERRARI’s turbo-electric hybrid”.

that will be the Limebeer paper I linked to 3 posts above yours. What information would you like to share with me from it?

I’ve seen the McCabism paper but I don’t have a copy.

[saviour stivala]

Also @henry, “optimal control of formula one car energy recovery system:- D.J.N. Limebeer, G. Perantoni and A.V. Rao.”.
Also “McCabism:- optimal control theory and FERRARI’s turbo-electric hybrid”.

that will be the Limebeer paper I linked to 3 posts above yours. What information would you like to share with me from it?

I’ve seen the McCabism paper but I don’t have a copy.

as I understand it Limebeer made a second paper later on based on SPA circuit.

[henry]

Also @henry, “optimal control of formula one car energy recovery system:- D.J.N. Limebeer, G. Perantoni and A.V. Rao.”.
Also “McCabism:- optimal control theory and FERRARI’s turbo-electric hybrid”.

that will be the Limebeer paper I linked to 3 posts above yours. What information would you like to share with me from it?

I’ve seen the McCabism paper but I don’t have a copy.

as I understand it Limebeer made a second paper later on based on SPA circuit.

If he did it’s not one I’m familiar with. Any idea how, if at all, it differed in modelling/methodology?

[Muniix]

In pre-season I read an article about Ferrari working on a new cylinder head concept which they wanted to bring any time in season. Obviously it's gonna be on spec 3 then.

Additive Manufacturing used in head, direct AM of the head. This was mentioned in the AM industry news recently with gains in cooling efficiency and weight reduction, honeycomb structuring used to improve cooling galleries with constant diameter reducing coolant volume and improving flow.
As for fuel chemistry with TJI Ferrari now likely have greater knowledge than Mercedes with Bill Attard it's inventor working at FCA while Elisa Toulson, his wife has been performing model development and experimental validation on TJI as associate professor at Michigan State University the two Australians who developed TJI such that it worked with liquid gasoline in prechamber. Bill also worked in the Bishop rotary valve development team for a few years. Now there's two Innovations who's features are screaming to be put together, so complimentary and symbiotic, the next engine regulations should encourage it.
The greater knowledge combined with new fuel chemistry the split battery could allow greater use of energy paths especially MGU-h, maybe different valve timing, turbine bypass.
Reducing the negative effects of the ES deployment being limited by the test
ES.stateOfCharge
Enforcing the 4MJ low to high window limit during race.

It's crazy to think about the fact that hydrogen and oxygen forming bonds with itself and carbon creates all this heat.

Regardless, Ferrari is taking the right approach, the model development and validation is key to getting all the gains out of the combustion concept. The problem is most F1 teams can't afford the hardware needed to study fuel kinetics properly, the large petrochemical companies sponsoring the cars CAN afford it. They also have the brains to not just develop the fuel but the combustion modeling as well, so that their partner F1 team can get the most out of the combustion process.

Shell is as big a petrochemical company as it gets, it's almost self evident Ferrari would make big gains with their fuel. Petronas is big, especially in Asia, however, I don't know if they compare to Shell. BP Castrol is also enormous so I don't know why they're lagging so far behind, same for Exxon Mobil. Maybe their relationship is not as close knit as Ferrari's is with Shell or Mercedes with Petronas.

As for combustion efficiency and chemical kinetics, those three atoms are doing a lot of chemistry. They are referred to as 'the' major combustion reaction, with NTC, the negative temperature coefficient - reactions rates are different.

Last month at the 2018 Princeton CEFRC combustion summer courses and conference we were discussing this reaction with professor William Greene and other researchers.

Research of the NTC low-t reactions using laser diagnostic and imaging of reacting species at Sydney University clean combustion laboratory. Joint volume-velocity shadowgraph imaging of direct injection spray and TJI hot jets.

With the development and analysis of adaptive neural network control of TJI engine operating in multiple combustion regimes.

Using autonomous driving platforms for their ultra high performance, AI, image & signal processing capability proving to be the ideal next generation ECU with 10's of teraflops, Minimising the sensor count with using cylinder pressure sensor and ion sensing indicating radicals.

[Muniix]

Fully agree, the turbo/MGU-H maximum RPM situation is the same as that of the ICE, hardly anybody ever try and approach the maximum RPM. But that doesn’t mean that those maximum RPM’S are not mandated. The actual and factual specification is as follows:. Energy recover system ERS:
Max exhaust turbine RPM = 125K.
Max MGU-H RPM = 125k.
Max power MGU-H = unlimited.
Max energy recovery per lap MGU-H = unlimited.
Max energy deployment per lap = unlimited.
Max energy storage per lap = 4MJ.
Max MGU-K RPM = 50K.
Max power MGU-K = 120KW/161BHP.
Max MGU-K energy recovery per lap = 2MJ.
Max MGU-K energy deployment per lap = 4MJ (33.33 seconds full power).
The other day henry was saying that with the turbo exhaust waste gates open the exhaust gasses are still able to assist turbine rotation. I invite him have a look at Andy Cowell graphically explaining the way exhaust gasses bypasses the turbine (Mercedes 2016 F1 exhaust explained).

IMO bolded parts are not true!

It seems from regulations that turbo speed is unlimited.

Battery effective capacity is limited to 4MJ, but you can charge it as many times you want/can - unlimited.

2MJ is the K>ES limit, but K>H is unlimited.

Also, 4MJ is limit on ES>K deployment, but not on H>K which is unlimited.

Also, having a bypass around turbo doesn't mean a portion of gasses doesn't go trough the turbine...

Also there is a test in the ES controller when in race mode/on track that enforcing the 4MJ state of charge window, a software test needed to verify conformance of dual bank ES.

[Muniix]

New article from AMuS
Shortened:

GPS measurements show that Ferrari has been gaining an exorbitant amount of time on the straights since two races. Calculations speak of a gain of up to 30 kilowatts (40hp).
Since the Austrian GP, Renault and Mercedes have again been measuring suspicious performance data via GPS. From a speed of 150 km/h the gap opens. Until then, the speed graphs are parallel. Then suddenly the red line for Ferrari moves away from the curves that stand for Mercedes and Renault.
If this phenomenon were explained by engine power alone, this would make a difference of 30 kilowatts. The competition claims to have determined that Ferrari recalls an energy of 4.4 megajoules per lap. 4.0 megajoules are permitted. [this is wrong].
The MGU-K will therefore feed 150kw (204 hp) instead of the permitted 120 kilowatts (163 hp) into the system. [wrong too imo]
At Silverstone, Mercedes noted with concern that Ferrari had significantly less de-rating in the race.

The assumption that Ferrari would rather use the electric power in the second part of the straight than in the first makes little sense. Most of the lap time is gained in the first part of the acceleration phase. Apparently Ferrari does not have to make sacrifices either at the beginning of a straight line or at the end.
Strangely enough, the power boost at Ferrari did not come with the introduction of the Spec2 engine at the GP Canada, but only two races later in Austria. That would speak for a software change, since the new hardware has already been in use for two races.

Meanwhile we hear that the Ferrari coup has nothing to do with the electric power from the MGU-K.
Does Ferrari use the MGU-H smarter than the competition? Usually this costs the combustion engine too much power because of the exhaust back pressure. That wouldn't explain 30 kilowatts. But perhaps Ferrari has found a way to reduce the power loss of the combustion engine.

Successful operation of the TJI operating a HCCI like combustion regime in main chamber followed by injection of further fuel for greater Imep at Melbourne University, this giving near 60% TE for the JCCI the secondary combustion starting as piston descends so full expansion, high turbulence from compression, Turbulent Jets and then injection entrainment, Imep is very high as is Bmep/Imep ratio, blow down and exhaust enthalpy for turbine would allow high MGU-h -> MGU-k while operating very high thermal efficiency. That level of performance and efficiency is going to kick everyone's butt.

[subcritical71]

Successful operation of the TJI in a main chamby HCCI mode followed by injection of further fuel at Melbourne University, this giving near 60 TE for the JCCI the secondary combustion starting as piston descends so full expansion, high turbulence from compression, Turbulent Jets and then injection entrainment, Imep is very high as is Bmep/Imep ratio, blow down and exhaust enthalpy for turbine would allow high MGU-h -> MGU-k while operating very high thermal efficiency.

I'm thinking this right now: https://youtu.be/PXmEsP-kpN0

[henry]

150kph is pretty much the traction limit speed. To go faster than the others at that speed Ferrari must be making more power. If the issue were SOC the speeds would diverge at a higher speed where the teams with lower SOC would need to reduce consumption of the ES.

Having theorised that teams would be emphasising energy management to extend the length of time they can deploy electric supercharger mode this evidence suggests that is not how Ferrari is gaining an advantage. I look forward to learning more.

[godlameroso]

Additive Manufacturing used in head, direct AM of the head. This was mentioned in the AM industry news recently with gains in cooling efficiency and weight reduction, honeycomb structuring used to improve cooling galleries with constant diameter reducing coolant volume and improving flow.
As for fuel chemistry with TJI Ferrari now likely have greater knowledge than Mercedes with Bill Attard it's inventor working at FCA while Elisa Toulson, his wife has been performing model development and experimental validation on TJI as associate professor at Michigan State University the two Australians who developed TJI such that it worked with liquid gasoline in prechamber. Bill also worked in the Bishop rotary valve development team for a few years. Now there's two Innovations who's features are screaming to be put together, so complimentary and symbiotic, the next engine regulations should encourage it.
The greater knowledge combined with new fuel chemistry the split battery could allow greater use of energy paths especially MGU-h, maybe different valve timing, turbine bypass.
Reducing the negative effects of the ES deployment being limited by the test
ES.stateOfCharge
Enforcing the 4MJ low to high window limit during race.

It's crazy to think about the fact that hydrogen and oxygen forming bonds with itself and carbon creates all this heat.

Regardless, Ferrari is taking the right approach, the model development and validation is key to getting all the gains out of the combustion concept. The problem is most F1 teams can't afford the hardware needed to study fuel kinetics properly, the large petrochemical companies sponsoring the cars CAN afford it. They also have the brains to not just develop the fuel but the combustion modeling as well, so that their partner F1 team can get the most out of the combustion process.

Shell is as big a petrochemical company as it gets, it's almost self evident Ferrari would make big gains with their fuel. Petronas is big, especially in Asia, however, I don't know if they compare to Shell. BP Castrol is also enormous so I don't know why they're lagging so far behind, same for Exxon Mobil. Maybe their relationship is not as close knit as Ferrari's is with Shell or Mercedes with Petronas.

As for combustion efficiency and chemical kinetics, those three atoms are doing a lot of chemistry. They are referred to as 'the' major combustion reaction, with NTC, the negative temperature coefficient - reactions rates are different.

Last month at the 2018 Princeton CEFRC combustion summer courses and conference we were discussing this reaction with professor William Greene and other researchers.

Research of the NTC low-t reactions using laser diagnostic and imaging of reacting species at Sydney University clean combustion laboratory. Joint volume-velocity shadowgraph imaging of direct injection spray and TJI hot jets.

With the development and analysis of adaptive neural network control of TJI engine operating in multiple combustion regimes.

Using autonomous driving platforms for their ultra high performance, AI, image & signal processing capability proving to be the ideal next generation ECU with 10's of teraflops, Minimising the sensor count with using cylinder pressure sensor and ion sensing indicating radicals.

CEFRC is a great resource, do you know if they're going to publish this year's lectures on YouTube?

[naukkis]

With the waste gates situated before the turbine scroll any/if any, exhaust exhaust gasses enters the scroll the gasses energy inside the scrol is as good as dead.

So you think that when wastegate opens turbocharged engine boost goes to zero instantly? Recovered energy from exhaust gases from turbine are from exhaust gas speed, that speed part is just the same whether wastegate is open or closed. Wastegate releases exhaust back pressure, which is harmful for engine performance but which is beneficial for turbine to extract more energy from exhaust flow. So with wastegate open they can harvest less energy from exhaust gases, but that less is probably nearer to full harvest capability than zero harvesting.

[MtthsMlw]

Bit more from AMuS

The most recent power upgrade was launched in two phases. A moderate one in Montreal, the big one in Spielberg. Since the Austrian GP, Renault and Mercedes have again been measuring suspicious performance data via GPS. According to their calculations, the Italians have increased from one race to the next 28 kilowatts (38 hp). "You'll usually find that much power in two years."
From the point that is no longer determined by traction, the gap opens. Until then, the speed counts are parallel. Then suddenly the red line for Ferrari moves away from the curves that stand for Mercedes and Renault. With Renault already at 150, with Mercedes at 200 km/h. If this phenomenon were explained by engine power alone, this would make a difference of 28 kilowatts.

[zibby43]

Bit more from AMuS

The most recent power upgrade was launched in two phases. A moderate one in Montreal, the big one in Spielberg. Since the Austrian GP, Renault and Mercedes have again been measuring suspicious performance data via GPS. According to their calculations, the Italians have increased from one race to the next 28 kilowatts (38 hp). "You'll usually find that much power in two years."
From the point that is no longer determined by traction, the gap opens. Until then, the speed counts are parallel. Then suddenly the red line for Ferrari moves away from the curves that stand for Mercedes and Renault. With Renault already at 150, with Mercedes at 200 km/h. If this phenomenon were explained by engine power alone, this would make a difference of 28 kilowatts.

It's incredible.

Merc lost 5 tenths to Ferrari on the straights in Germany.

Merc. That's something I'm used to reading about a Renault-powered car.

https://www.motorsport.com/f1/news/merc ... 62491/?s=1

[saviour stivala]

Bit more from AMuS

The most recent power upgrade was launched in two phases. A moderate one in Montreal, the big one in Spielberg. Since the Austrian GP, Renault and Mercedes have again been measuring suspicious performance data via GPS. According to their calculations, the Italians have increased from one race to the next 28 kilowatts (38 hp). "You'll usually find that much power in two years."
From the point that is no longer determined by traction, the gap opens. Until then, the speed counts are parallel. Then suddenly the red line for Ferrari moves away from the curves that stand for Mercedes and Renault. With Renault already at 150, with Mercedes at 200 km/h. If this phenomenon were explained by engine power alone, this would make a difference of 28 kilowatts.

The big surprise is how long a time is being spend this time on suspicious gains by FERRARI before firing actual cheating possibilities.

[subcritical71]

So you think that when wastegate opens turbocharged engine boost goes to zero instantly? Recovered energy from exhaust gases from turbine are from exhaust gas speed, that speed part is just the same whether wastegate is open or closed. Wastegate releases exhaust back pressure, which is harmful for engine performance but which is beneficial for turbine to extract more energy from exhaust flow. So with wastegate open they can harvest less energy from exhaust gases, but that less is probably nearer to full harvest capability than zero harvesting.

I'd look at it a little differently. The rules specify the min/max of the exhaust and waste gate pipes.
Total cross sectional area of the exhaust pipe must be between 7,500 and 14,000mm2.
Total cross sectional area of the waste gate pipe(s) must be between 1,590 and 2,375mm2.

Let's go with worst case scenario, for this point, of a team using a 7,500mm2 exhaust outlet (minimum allowed) and a 2,375mm2 wastegate pipe (maximum allowed). How does opening the wastegate pipe even to fully open cause a 100% flow diversion through the wastegate pipe?

I'm sure some basic calculations knowing RPM and cylinder volume will support that the turbine will never see zero flow.