F1technical.net

My understanding is that most modern GDI systems operate with stratified charge at lighter loads and lower rpm and operate with homogeneous charge beyond that.

Re Mitsubishi : much lower tech
If we ignore technology level, Mercedes were there in WW2 on the me109

Tommy Cookers wrote:BTW if people Google V Ganesan Internal Combustion Engines
.... Googling dissociation should also get this book at the right pages
they should find Fig 3.6 (page 115 in my copy) showing real engine BTE benefits (from leaning) fall far short of fuel-air theory
and Prof Ganesan's statement .... 'the maximum efficiency is within the lean zone very near the stoichiometric ratio'

Thanks for the reference. Couldn't find the figure in the restricted copy I downloaded. Perhaps you could post an image?

There are however plenty of references to mixture for best efficiency e.g.
- Chapter 4.5 Dissociation including Fig 4.3 Effect of Dissociation on Power showing best efficiency at near 20:1
- Ch 4.8.2 Effect of Fuel-Air Ratio see Fig 4.1 Effect of Mixture Strength on Thermal Efficiency

On dissociation the maximum extent of dissociation occurs in the burnt gases of the chemically correct fuel-air mixture . . . but decreases with the leaner and richer mixtures.

Re AFR/GDI: I'm with Ringo, there's no answer in a text book but careful dyno work will tell you a lot, pretty quickly.

For reference material on combustion/GDI, I think Heywood is good on combustion fundamentals and Ricardo's papers on VOLCANO and HyBoost projects give good insight into practical optimisation of hybrid/GDI Powertrain, albeit in a road car context. Both google-able.

I don't think these guys learn much about mixtures on the dyno beyond getting the final trim sorted and perhaps seeing how far the envelope can be pushed in some cases.

What I mean to say is, they pretty much know what they would like to see - what mixtures they would like to be able to run before an engine even goes on the dyno. All courtesy of powerful computer simulations and lots of development with single cylinder engines (on a dyno of course).

Let me remind you what Ringo said.

ringo wrote:This discussion is not very novel, since we know a team will simply use the engine bench to play around with ratios to get the best result. I'll just say the engine maker will simply go with whatever air to fuel ratio suits their needs from the engine.

No - the maker would have very specific goals for AFR and the dyno work would be a struggle to develop an engine that will function and survive with that AFR.

Most posters clearly had not even considered the possibility of these cars running lambda 1.1 and above.

I have a feeling we may all be saying something similar in different ways.

WilliamsF1 wrote:

ringo wrote:Switching gears a bit.
The drivability of the different engines to me is very interesting. I haven't seen much discussion on that even though there is ample evidence that some units, like the ferrari have a poorer drivability.
What's going on with the mercedes power unit, that allows it to behave so well on acceleration out of a corner compared to the renault and ferrari?

Merc's secret for drivability control is probably the way the ERS-H works. FIA does not define the power rating of this MGU, Merc probably using a higher rated unit that does a very good job in spooling the turbo and avoid lag.

Other 2 engines are probably having a turbo lag and are using the ERS-K to compensate for this along with ERS-H which causes poor drivability when the turbo kicks in.

Isn't the mgu-h limited to 120 kw?

Cold Fussion wrote:

WilliamsF1 wrote:

ringo wrote:Switching gears a bit.
The drivability of the different engines to me is very interesting. I haven't seen much discussion on that even though there is ample evidence that some units, like the ferrari have a poorer drivability.
What's going on with the mercedes power unit, that allows it to behave so well on acceleration out of a corner compared to the renault and ferrari?

Merc's secret for drivability control is probably the way the ERS-H works. FIA does not define the power rating of this MGU, Merc probably using a higher rated unit that does a very good job in spooling the turbo and avoid lag.

Other 2 engines are probably having a turbo lag and are using the ERS-K to compensate for this along with ERS-H which causes poor drivability when the turbo kicks in.

Isn't the mgu-h limited to 120 kw?

No. The MGUK is limited to 120kW, but the MGUH is not restricted.

If the MGUH could produce more than 120kW the extra woudl have to be sent to the ES, as teh MGUK could only take 120kW directly.

gruntguru wrote:My understanding is that most modern GDI systems operate with stratified charge at lighter loads and lower rpm and operate with homogeneous charge beyond that.

Oops forgot to edit. It wouldn't make sense. Yes stratified at low load/rpm and homogeneous beyond.

gruntguru wrote:I don't think these guys learn much about mixtures on the dyno beyond getting the final trim sorted and perhaps seeing how far the envelope can be pushed in some cases.

What I mean to say is, they pretty much know what they would like to see - what mixtures they would like to be able to run before an engine even goes on the dyno. All courtesy of powerful computer simulations and lots of development with single cylinder engines (on a dyno of course).

Let me remind you what Ringo said.

ringo wrote:This discussion is not very novel, since we know a team will simply use the engine bench to play around with ratios to get the best result. I'll just say the engine maker will simply go with whatever air to fuel ratio suits their needs from the engine.

No - the maker would have very specific goals for AFR and the dyno work would be a struggle to develop an engine that will function and survive with that AFR.

Most posters clearly had not even considered the possibility of these cars running lambda 1.1 and above.

How is a simmulation going to tell you more than a real test?
I'm not clear on what you are saying.
What other goals will a a team be looking for if they can't check those goals on the real engine it'self. There are no specifc AFR goals.
The only goals the engine maker has is the end results of fuel consumption, power output, temperatures etc. Measured things, not simulated things or arbitrary "sweet spot" numbers.
You test until you get the results that you want, then you go back and see what kind of ratios gave you those results and what kind of settings. That's the scientific approach.

A simulation tells you more than a real test (virtual instrumentation can "see" things that real instruments can't) but is not as accurate or reliable as a real test. A simulation is also vastly less expensive, more flexible, faster turnaround. . . .

All of the usual design goals - power, flexibility, efficiency, reliability etc will be optimised in the virtual domain before the "real" engine is even designed - let alone built and ready for testing. The testing itself is - initially - all about validating the computer model. The AFR goals will be those shown on the simulations to give the best combination of the performance goals.

The "scientific" approach you speak of is still applied in the "real" domain on test engines but with a much more limited scope.

Mercedes MGU-H can direct feed the power and bypass the battery.So more power the whole lap.

Not just MB. They all can.

gruntguru wrote:Not just MB. They all can.

It seems that MB can do it all thru every lap tho due to their oversized turbine.

djos wrote:

gruntguru wrote:Not just MB. They all can.

It seems that MB can do it all thru every lap tho due to their oversized turbine.

Supposed oversized turbine.

I would suggest that their control strategy is better, more sorted and that is why they can do it.