Honda Power Unit Hardware & Software

[gruntguru]

The MGUH could probably generate 120kW as well with some re-tuning of the engine to sacrifice boost and crankshaft power. That way the ES could be recharged in about 17 sec with 2 MJ each from the 'K and the 'H.

[alexx_88]

@j.yank: After reading about it, your way of simulating and "reverse engineering" why Mclaren are having problems is very basic and not at all precise, at least that's my opinion.

You still didn't reply to me how did you establish it was case 'a', not case 'b':
a. MGU-K works at half the power because of its design limitation
b. MGU-K works at half the power because there isn't enough energy harvested for the whole lap, as the MGU-H doesn't harvest enough (i.e. 50% of MGU-H harvest over a lap means that the total available would be ~2.5MJ, not 4).

I'd start with different values for the total amount of energy available over a lap, put them into the simulator and see what data you come up with. You're making assumptions, without any evidence, that it's a MGU-K limitation.

[j.yank]

@j.yank: After reading about it, your way of simulating and "reverse engineering" why Mclaren are having problems is very basic and not at all precise, at least that's my opinion.

You still didn't reply to me how did you establish it was case 'a', not case 'b':
a. MGU-K works at half the power because of its design limitation
b. MGU-K works at half the power because there isn't enough energy harvested for the whole lap, as the MGU-H doesn't harvest enough (i.e. 50% of MGU-H harvest over a lap means that the total available would be ~2.5MJ, not 4).

I'd start with different values for the total amount of energy available over a lap, put them into the simulator and see what data you come up with. You're making assumptions, without any evidence, that it's a MGU-K limitation.

My "reverse engineering" cannot be precise at all and this is not the goal. It should give only very general idea were the problems can be.

The available energy is the leading parameter and this is calculated in the way that I already pointed. Note that even if this doesn't match the actual values given by stevesingo, the proportion between MGU-K and MGU-H is almost the same. When you have this then you can make several simulations that reflects different cases a, b, c, d for several cars. Then you see which case is closest to the proportional differences that we see in the reality between each of these cars.

[Andres125sx]

Comparing laptimes to work out some power numbers is flawed for a simple reason, this PUs do not provide same power constantly, it depend on the energy harvested, so an improvement in energy harvesting that allows using full power (ICE + ERS) longer will improve laptimes even if peak power is exactly the same.

Simulators like OptimuLap do not relay on the maximum power applied constantly during the entire lap. They are using the value of the maximum power as available power that is modified in each corner and straight of the circuit, depending of their specific characteristics. Yes, we do not know the exact moment when energy from MGU-H is applied and when from MGU-K but we know the aggregate sum per lap that put the ultimate limit over the available energy. This available energy is much, much more important than the peak power. This depends on the accuracy of the simulation software how this available energy will be distributed over the lap. Yes, the parameters of the chassis have huge impact but if you put uniform parameters for all cars except their power you can get at least oriented what is the weight of one or another parameter - in our case which one of the MGUs contributes more for the lap time deficiencies of McLaren. Don't try to be very precise, but don't ignore the general balances of the process.

How do you know the bolded part? You simply can´t, available energy is the main difference between Mercedes and Honda, Mercedes is able to harverst from mgu-h much much longer than Honda, so they can apply mgy-k much longer than Honda. What I mean is this is not a fixed parameter or constant, it is very different from team to team so you can´t know it

Since you can´t know how much available energy each car has, you can´t know how do they distribute them, so you don´t know what power they´re using in each straight and for how long, let alone doing any estimation about total peak power

Comparing the times from last year and this year on several circuits for different teams you can't find another explanation than Mercedes and Ferrari are running now in the range from 850 to 910 hp (with ERS).

Basically there´re too many unknowns (available energy during a lap, how it is distributed, real power of ICE...), you can do some assumption to get some results, but it will be strongly dependant on your assumption, so it will be more of an estimation than a real calculation.

That´s the reason I replied, your post states "you can´t find another explanation than they´re running now in the range from 850 to 910 hp", and I cannot agree with that since you´re doing assumptions to get those numbers, so if you change your assumptions (how much energy they have available for a lap AND how do they distribute it) your results would change a good ammount, so yes, I can´t find more explanations than that. For example they´re harvesting more energy than past season and/or more than you´re assuming.

[mikeerfol]

So, user muramasa from Autosport forum has posted more photos of that PU in Suzuka here

[tok-tokkie]

So it looks like a conventional radial flow turbine.
To me it seems quite small. Am interested if I am correct.

The MGU-K does not look much smaller than the ones we have seen. Possibly the diameter is less than the others.
I don't understand all the stuff to the right on the picture. I assume it is delivering power (& reaping it) from the clutch end of the engine (right of picture). What is between the actual MGU-K and the clutch bell housing?

[Blackout]

Nice looking wastegates... they're tiny and seem to reject 'waste' gases directly into the central axhaust exit, just after the turbine

[PlatinumZealot]

Only the compressor intake duct makes the Merc engine look a bit less compact than the Honda

Excuse me while I light my spliff...

https://www.youtube.com/watch?v=yLuHE-82o40

[PlatinumZealot]

Nice looking wastegates... they're tiny and seem to reject 'waste' gases directly into the central axhaust exit, just after the turbine

Yeah.. but the wastegates enter at a poor discharge angle into the exhaust pipe. A textbook mistake. Check out Williams or Torro Rosso for a good entry angle. Hmm.. they also have two wastegates which can be more of a headache that one single one as Renault had learned.

[diffuser]

someone turned on the light switch!

[Blackout]

I agree platinumZ


Are they using a similar system to this?

[ojlopez]

Seems like the exhaust is running kind of hot. The insulation looks all burned in this photo.

[RZS10]

Seems like the exhaust is running kind of hot. The insulation looks all burned in this photo.

That's from some showroom, probably just a reflection (check the other pics for the orange tint)

[Edax]

Seems like the exhaust is running kind of hot. The insulation looks all burned in this photo.

That's from some showroom, probably just a reflection (check the other pics for the orange tint)

Looks like typical tempering colors to me not reflection. Also they are visible in every angle.


So it has been a couple of hundreds of degrees there (apologies for the non IS units but it was the clearest pic I could find). Doesn't seem too surprising to me.


You have to be a bit careful though with the interpretation in terms of temperature. Since what you're seeing is basically the thickness of the oxide skin that is formed at high T (due thin film interference), the colors will shift in temperature with type of alloy used.

[PlatinumZealot]

I agree platinumZ
http://www.servimg.com/view/14795526/1826

Are they using a similar system to this?
http://www.autozine.org/technical_schoo ... udi_V8.jpg

They are actual photos of the variable trumpets. Yeah they work like that but instead they are only extend for a partial arc like a semi-circle or a macroni.