Honda Power Unit Hardware & Software

[taperoo2k]

Given the limited amount of pre season testing, Honda needs to get the 2017 PU in the back of the McLaren on Day One. So they can discover and fix problems as quickly as they can. Retrofitting a 2016 PU into the 2017 McLaren would also likely divert resources away from other areas. Tbh I believe Honda will hit the ground running with a 2017 PU and not care if they encounter problems. Better to get them out of the way early, than face them at the end of Pre Season Testing, with a mad dash to come up with fixes before they introduce the race spec engine for the first couple of races of the season.

You also have to keep in mind that pre season testing is also a useful time for the McLaren and Honda Engineers who install, fix or remove the PU's at the tracks, to get acquainted with the new systems and packaging in a racing environment. They'll have been practising in the factories, but you can't really mimic the dynamics of a race track where you don't have immediate access to lots of spare parts or manufacturing tools.

[godlameroso]

About the only thing the same between 2016 and 2017 PU is the 90 degree V.
So why would you ask/state they would run the 2016 PU for the first practice?

And cylinder bore.

I don't know why they would or why they wouldn't, all we know is McLaren hasn't fired up the engine in the chassis yet, where a lot of others have. At least we haven't heard anything.

Does this mean the 2016 was a mere evolution of 2015?

[taperoo2k]

I don't think it means all that much that McLaren haven't as far as anyone knows, fired up the new Honda PU in the MCL32 yet. We'll either get a sleek video of the PU being fired up in the run up to the reveal of the MCL32, or the first time anyone outside of McLaren and Honda hears the new PU will be when they fire it up in the pit lane on day one of the first preseason test.
The 2016 PU was developed within the constraints of the token system, so it's reasonable to assume it's an evolution of the 2015 PU. It makes sense for certain parts of the PU to carry over into the 2017 PU, the token upgrades were likely used with 2017 in mind. It's the parts they've redesigned or new solutions Honda's come up with that interest me. But it will be a while before that sort of information becomes apparent.

[mclaren111]

I don't think it means all that much that McLaren haven't as far as anyone knows, fired up the new Honda PU in the MCL32 yet. We'll either get a sleek video of the PU being fired up in the run up to the reveal of the MCL32, or the first time anyone outside of McLaren and Honda hears the new PU will be when they fire it up in the pit lane on day one of the first preseason test.
The 2016 PU was developed within the constraints of the token system, so it's reasonable to assume it's an evolution of the 2015 PU. It makes sense for certain parts of the PU to carry over into the 2017 PU, the token upgrades were likely used with 2017 in mind. It's the parts they've redesigned or new solutions Honda's come up with that interest me. But it will be a while before that sort of information becomes apparent.

If I remember correctly, Hasegawa said in an interview last year that the Camshaft & Block was replaced (under reliability) with a view on the 2017 engine .

[damager21]

McLaren will start 2017 testing with the new Honda engine. If there are any major reliability issues, they will tune down the engine but there is no way they will test with a 2016 engine.

Given the abolition of token system, there will be a major redesign in the engine architecture and as much as McLaren needs on track data to refine aero, Honda needs on track data to iron out reliability issues and also plan for next upgrade patch.

McLaren will receive Honda engine next week so expect a teaser to be released soon

[fellowhoodlums]

McLaren will start 2017 testing with the new Honda engine. If there are any major reliability issues, they will tune down the engine but there is no way they will test with a 2016 engine.

Given the abolition of token system, there will be a major redesign in the engine architecture and as much as McLaren needs on track data to refine aero, Honda needs on track data to iron out reliability issues and also plan for next upgrade patch.

McLaren will receive Honda engine next week so expect a teaser to be released soon

Completely right. Daft to use 2016 engine, only way to fix reliability/improve is testing so it's aero/chassis/engine in pre-season tests.

Lots of people implying the engine is more important than last year because of the regulation changes on the tyres/aero side meaning power through corners but i'm not so sure. All about mechanical grip I think so the chassis is key, more important than small engine deficiencies.

[diffuser]

I don't think it means all that much that McLaren haven't as far as anyone knows, fired up the new Honda PU in the MCL32 yet. We'll either get a sleek video of the PU being fired up in the run up to the reveal of the MCL32, or the first time anyone outside of McLaren and Honda hears the new PU will be when they fire it up in the pit lane on day one of the first preseason test.
The 2016 PU was developed within the constraints of the token system, so it's reasonable to assume it's an evolution of the 2015 PU. It makes sense for certain parts of the PU to carry over into the 2017 PU, the token upgrades were likely used with 2017 in mind. It's the parts they've redesigned or new solutions Honda's come up with that interest me. But it will be a while before that sort of information becomes apparent.

I agree.

That was a pub stunt. It was symbolic to "we're almost ready to race!"
They have a large number of PUs just like that one running on dynos for maybe Millions of hours? What difference would sticking it in a unfinished car and starting it?

[Joseki]

Judging by the past seasons I presume Honda will run detuned at about 80% in D1 to confirm reliability and then they'll start to "unlock potential". If I remember correctly they used full power only after China in 2016. This year they'll hopefully use all the power from Q2 in Melbourne.

[GoranF1]

Judging by the past seasons I presume Honda will run detuned at about 80% in D1 to confirm reliability and then they'll start to "unlock potential". If I remember correctly they used full power only after China in 2016. This year they'll hopefully use all the power from Q2 in Melbourne.

Judging by the past seasons they will run it at 100% from day one, but we will try to convince ourselfs and everyone its detuned to 80%.

[Brian Coat]

When Honda posted their telemetry data, Wazari pointed out that it is doctored.

We still found the MGU picture interesting, including the doctoring, but the speed data was just a load of straight lines - no good for understanding the engine.

The TV speed data is low resolution and is sometimes a bit out of phase vs time, revs etc., so also not much good except to get the gearing, like Hollus does.

I am still interested in the Honda-Merc gap at the end of the 2016 season, so I made an audio analysis of Hamilton (pole) and Alonso (Q2) in qualifying at Abu Dhabi.

Using Hollus's gearing data you can get high resolution speed data which correlates well to the trap data.

Away from wheel spin etc. the data shows that Hamilton can accelerate ~7% faster.

Does this imply a power deficit of 60-70hp, if aero is similar?

Does that sound right or did I screw something up?

[godlameroso]

I'd say it's right, but not over a whole lap. There are parts where the Honda can keep up, and parts where it can't.

https://youtu.be/2SvC6Ep4DAw
Look at the Brazil comparison, even in sector 1, despite being slower down the pit straight, the Honda keeps up well down the straight after the Senna Esses, but falls far behind in sector 2, and is 10kph down at the finish line. So it means there's a deployment difference. It's hard to tell how much difference the chassis makes, although the Mercedes does look more planted on that downhill left hand turn. They clearly have more acceleration, luckily Juzh was nice enough to try to synchronize the low res TV telemetry to what the cars were actually doing.

[FW17]

When Honda posted their telemetry data, Wazari pointed out that it is doctored.

We still found the MGU picture interesting, including the doctoring, but the speed data was just a load of straight lines - no good for understanding the engine.

The TV speed data is low resolution and is sometimes a bit out of phase vs time, revs etc., so also not much good except to get the gearing, like Hollus does.

I am still interested in the Honda-Merc gap at the end of the 2016 season, so I made an audio analysis of Hamilton (pole) and Alonso (Q2) in qualifying at Abu Dhabi.

Using Hollus's gearing data you can get high resolution speed data which correlates well to the trap data.

Away from wheel spin etc. the data shows that Hamilton can accelerate ~7% faster.

Does this imply a power deficit of 60-70hp, if aero is similar?

Does that sound right or did I screw something up?

That does not take into account the Mclaren rear end issue which does not allow for the driver to put his foot down

[Brian Coat]

Thanks for the comments.

The concerns raised are valid and I did what I could to take them into account.

I tried to take into account deployment constraints by using an early straight on a qualifying lap - full deployment should be available and the data suggests this is the case. Deployment curtailment is detectable in the data as a deviation from the 'normal' / deployed-state speed vs. time curve.

I tried to take into account wheelspin / mechanical grip issues by leaving out data early in the straight, where wheelspin is indicated in the rpm trace.

Notwithstanding these attempts at mitigation, I recognise that these techniques are prone to innaccuracy.

If we compare Merc-Mac at the start of this season, using the same method it might be interesting.

[Brian Coat]

They have a large number of PUs just like that one running on dynos for maybe Millions of hours?

Er maybe not millions ...!

[hemichromis]

I don't think it means all that much that McLaren haven't as far as anyone knows, fired up the new Honda PU in the MCL32 yet. We'll either get a sleek video of the PU being fired up in the run up to the reveal of the MCL32, or the first time anyone outside of McLaren and Honda hears the new PU will be when they fire it up in the pit lane on day one of the first preseason test.
The 2016 PU was developed within the constraints of the token system, so it's reasonable to assume it's an evolution of the 2015 PU. It makes sense for certain parts of the PU to carry over into the 2017 PU, the token upgrades were likely used with 2017 in mind. It's the parts they've redesigned or new solutions Honda's come up with that interest me. But it will be a while before that sort of information becomes apparent.

I agree.

That was a pub stunt. It was symbolic to "we're almost ready to race!"
They have a large number of PUs just like that one running on dynos for maybe Millions of hours? What difference would sticking it in a unfinished car and starting it?

200+ years of running? should be pretty reliable then!