Honda Power Unit Hardware & Software

[godlameroso]

Pretty much. They need higher engine revs to create a higher volume of exhaust gas to drive the compressor and have good energy harvesting. The fact that they have a relatively undersized compressor means that it needs more rpm to create adequate amount of boost pressure. It's a pretty decent compromise, it's not like they're going to 13k rpm, they push to 12,500rpm max.

[godlameroso]

The ICE is really honda's problem. Turbo is soo... soo.. they can easily fix that to get a good qualifying, but since the race is where you score points, they sized the turbo for the race. Small power from the ICE means a small turbo for the race. So what we do know is that the ICE is the weakness. Combustion rather. It is too easy to fit a big ass turbo in dyno (or run the engine on compressed air) and fix the power issue (on the dyno to verify things).. so someone with half a brain know it's the ICE that the major weakness.

Not disagreeing with you, just look at it from another point of view. They designed their turbo based on the combustion technology they had in 2015 and they did not have enough tokens to do a complete power unit layout change. So they were stuck between a rock and a hard place. Once they introduced the combustion update this year, they didn't anticipate how lean they would be able to run and realized quite late that no matter how much they tried to improve the internal geometry of the compressor, that it simply would not be able to flow enough air to fully exploit how lean the new combustion concept allows. I'm sure they're well aware of how much of a step forward is waiting for them next year.

Hindsight is 20/20 as they say.

[gruntguru]

BMEP is calculated from actual output predominantly occuring 'per active rev' ie during compression and expansion strokes (that's why the calculation of BMEP from measured power differs according to whether the engine is a 2 stroke or a 4 stroke) but the notional '3 bar exhaust' pumping loss as described by gg occurs over a half rev ie 1 stroke only (exhaust) so a 35 bar BMEP engine with 3 bar exhaust pressure reduced to 1 bar will improve the BMEP by about 1 bar (averaged over the 2 strokes) about a 3% BMEP and power improvement

If that were the case, a change to the cylinder pressure during the power stroke of say 1 bar would only change the BMEP by half a bar (which is not the case). If you check the calculation of BMEP from brake output you will find a term "number of revolutions per power stroke" so a four stroke engine has double the BMEP of a two stroke at the same rpm and power.

well imo it seems as though it is the case ......
when I calculate the BP from your 35 bar BMEP and 5250 revolutions of a 1600 cc engine running at 10500 rpm I get 490 kW

BMEP in a 4 stroke is the steady fictitious pressure that fictitiously applies throughout the whole revolution of the engine ....
(but only to every other revolution)
and regardless of the fact that in real 4 stroke engines real pressures occur only for half revolutions
BMEP is not a steady fictitious pressure applying to half revolutions

exhaust pressure relief of 2 bar applies for half of each of the above 5250 revolutions (ie for a quarter of the total 10500 revolutions)
so cannot be directly compared to the 35 bar conventional BMEP (because this applies to the whole of the above 5250 revolutions)
exhaust pressure relief of 2 bar for half rev must be replaced by an equivalent 1 bar for the whole rev to calculate power benefit

only if we redefined BMEP as the equivalent fictitious pressure that would be need to be applied for half of the above 5250 revolutions ....
would the new BMEP (still developing the same power that the actual engine does, of course) be directly comparable with the 2 bar .....
this new BMEP would be 70 bar, and so the relief of exhaust scavenge power loss would be 2/70 ie 3% as I suggested before

it would be interesting to see a textbook expression for power loss on the exhaust stroke due to exhaust pressure

I bar average pressure during the exhaust stroke = 1 bar loss of BMEP
Work per cycle is the area enclosed by the loops in the pv diagram. Clockwise loops are positive work and anti-clockwise loops are negative (pumping) work.

The high pressure loop is formed by the compression stroke (-ve work created by positive pressure while reducing volume) and the power stroke (+ve work created by positive pressure while increasing volume). The high pressure loop is the only part analysed in the classical Otto cycle analysis.

The low pressure loop is formed by the exhaust stroke (-ve work created by positive pressure while reducing volume) and the intake stroke (-ve work created by negative pressure while increasing volume).
BMEP is work/cycle divided by displacement. Note that displacement is Vmax - Vmin on the pv diagram.

So if you increase the exhaust pressure by 1 bar, the pumping work will increase by 1bar x displacement and the BMEP will reduce by 1 bar x displacement/displacement = 1 bar.

Wikipedia link showing formulas for BMEP for 2 stroke and 4 stroke. https://en.wikipedia.org/wiki/Mean_effective_pressure

[Tommy Cookers]

@gg
with apologies, I now see that you are right on this matter and the power gain calculation

checking, I can see that the standard formulae treat the BMEP as acting over 1 swept volume ie for 1/2 rev in every 2 revs


MR MODERATOR - please can you cancel the down-mark I gave to gruntguru's post of 1037hrs 18 Sept ?

[PlatinumZealot]

The ICE is really honda's problem. Turbo is soo... soo.. they can easily fix that to get a good qualifying, but since the race is where you score points, they sized the turbo for the race. Small power from the ICE means a small turbo for the race. So what we do know is that the ICE is the weakness. Combustion rather. It is too easy to fit a big ass turbo in dyno (or run the engine on compressed air) and fix the power issue (on the dyno to verify things).. so someone with half a brain know it's the ICE that the major weakness.

Not disagreeing with you, just look at it from another point of view. They designed their turbo based on the combustion technology they had in 2015 and they did not have enough tokens to do a complete power unit layout change. So they were stuck between a rock and a hard place. Once they introduced the combustion update this year, they didn't anticipate how lean they would be able to run and realized quite late that no matter how much they tried to improve the internal geometry of the compressor, that it simply would not be able to flow enough air to fully exploit how lean the new combustion concept allows. I'm sure they're well aware of how much of a step forward is waiting for them next year.

Hindsight is 20/20 as they say.

Fair enough but the part in bold can't be right... The engineers are not that incompetent, all of that would be known in development. I am pretty sure the compressor flows enough air for what the engine was designed for and more.

[PlatinumZealot]

I bar average pressure during the exhaust stroke = 1 bar loss of BMEP
Work per cycle is the area enclosed by the loops in the pv diagram. Clockwise loops are positive work and anti-clockwise loops are negative (pumping) work.

The high pressure loop is formed by the compression stroke (-ve work created by positive pressure while reducing volume) and the power stroke (+ve work created by positive pressure while increasing volume). The high pressure loop is the only part analysed in the classical Otto cycle analysis.

The low pressure loop is formed by the exhaust stroke (-ve work created by positive pressure while reducing volume) and the intake stroke (-ve work created by negative pressure while increasing volume).
BMEP is work/cycle divided by displacement. Note that displacement is Vmax - Vmin on the pv diagram.

So if you increase the exhaust pressure by 1 bar, the pumping work will increase by 1bar x displacement and the BMEP will reduce by 1 bar x displacement/displacement = 1 bar.

Wikipedia link showing formulas for BMEP for 2 stroke and 4 stroke. https://en.wikipedia.org/wiki/Mean_effective_pressure


http://i.imgur.com/5ghwYMo.jpg

This is true.. thats why a more "open" or bigger turbine will give more horsepower as the flow goes up. Less power stealing back-pressure. The Mercedes turbine wheel is said to be very big and the waste-gates are the tiniest when you compare it to the others...

[daren_p]

The ICE is really honda's problem. Turbo is soo... soo.. they can easily fix that to get a good qualifying, but since the race is where you score points, they sized the turbo for the race. Small power from the ICE means a small turbo for the race. So what we do know is that the ICE is the weakness. Combustion rather. It is too easy to fit a big ass turbo in dyno (or run the engine on compressed air) and fix the power issue (on the dyno to verify things).. so someone with half a brain know it's the ICE that the major weakness.

Not disagreeing with you, just look at it from another point of view. They designed their turbo based on the combustion technology they had in 2015 and they did not have enough tokens to do a complete power unit layout change. So they were stuck between a rock and a hard place. Once they introduced the combustion update this year, they didn't anticipate how lean they would be able to run and realized quite late that no matter how much they tried to improve the internal geometry of the compressor, that it simply would not be able to flow enough air to fully exploit how lean the new combustion concept allows. I'm sure they're well aware of how much of a step forward is waiting for them next year.

Hindsight is 20/20 as they say.

Fair enough but the part in bold can't be right... The engineers are not that incompetent, all of that would be known in development. I am pretty sure the compressor flows enough air for what the engine was designed for and more.

The engineers likely did know they would be limited by turbo size, but because of the tokin limitations (& all the other areas that had to be improved from last year), they likely didn't have enough for the required layout change. Believe the turbo is probably about as big as they can make it in the current configuration, but is currently the limiting factor.

Merc's turbo was much larger to start with & each year they have upped its size as efficiency increases.

[PlatinumZealot]

No outsider has measured or even seen the Honda compressor... so let's not make too many assumptions...

[Sasha]

The one they started the season with was 30cm long(long and narrow,2nd gen. is wider)! But they been having problems with it.(stall problems and just doesn't follow the rules like a more normal design)

On paper it should of work but not too well in the real world.

So moving the Compressor out of the V is more a move to a more normal Compressor.(pancake style).Can't put that in the V.

[godlameroso]

It's an interesting philosophy, it's as if they took a conventional pancake compressor, and simply stretched it axially. I guess it goes back to my point that nothing in any engine will ever scale mathematically. Despite having a similar working volume of the traditional turbo, it just couldn't work the air the same as a conventional turbo. I have to commend them for taking it and developing it as far as they did, even surpassing last year's Ferrari engine by a noticeable margin.

Perhaps they learned interesting things about their hybrid compressor concept, and that's why they want to move the turbo to the back of the engine. Maybe they want to keep their stubby compressor, and freedom to size it, altering it's proportions is easier in the back, and more difficult in the front of the engine because the axial component means that more of the turbo juts out forward. The longer the shaft the harder and the more care needs to go in designing it. The inherent design of their axial hybrid turbo could work to their favor in hiding part of the compressor in the V. Could serve as a decent compromise to minimize the package size for the chassis side, but what do I know, I'm just some random nobody.

[Brian Coat]

I wonder if Honda's advanced cf compressor capability in another division played any role in emboldening them to go for something a bit too clever?

The attached HF118 data shows that even prior to the GE JV, the Honda team were none too shabby at advanced cf compressor design:

[Roostfactor]

"but what do I know, I'm just some random nobody." Definetely not a nobody! You have significant knowledge on ICE's, and most of us are speculating and hypothesizing what Honda really have going on. Oh to be a fly on the wall in Hondaland.

[gruntguru]

The attached HF118 data shows that even prior to the GE JV, the Honda team were none too shabby at advanced cf compressor design:

That's a handy chart Brian - thanks. Suggests that at PR = 3.5 the state of the art for efficiency (dashed line) is about 84% although that Honda outlier hints that better things are possible.

[Brian Coat]

All other things being equal, an F1 compressor will be less efficient due to its smaller scale. The HF118 is a tiddler but at 1700 lbf it will still kick out the equivalent of a couple of thousand hp?

But then again, Honda might have learned more through the partnership with GE, which this graph pre-dates.

[PlatinumZealot]

Their compressor is working as well as it should to provide enough air for what the combustion can support.
It would be too easy if the compressor was Honda's problem. We have seen them raise the airbox up to make the compressor bigger, and even Arai himself said in 2015 that the 2016 compressor would be close in size to the Mercedes one. So not sure why we are digging up old bones with regard to the compressor.