This explains why they early in the season were shifting at 12,5 k.....way to highMudflap wrote: ↑10 Dec 2017, 20:27I bought the autosport magazine and they have a pretty good report on what went on with Honda throughout the year. There are quite extensive Hasegawa quotes - I'll try to summarize the ones I find important.
In 2015 their H recovery was very poor, so for 2016 they increased the turbine size and as a result had to move the turbo up. This however increased the CG height so for 2017 they decided to go for the split turbo. As the compressor housing was now forward of the block, the oil tank had to be redesigned into a mercedes-like crescent shape. Unfortunately this caused all the issues they had in pre-season testing and took 2-3 days to fix.
Now the way I see it, McL would have provided Honda with the oil tank accelerations. Honda would then have tested the oil tank on a shaker rig (I think these are also known as rodeo rigs) to validate the design. I don't think it can be determined from the statements whose blame it was - either McL's for miscalculating acceleration (or not providing any at all) or Honda's for not testing the oil tank correctly.
On the issue of power Hasegawa explicitly states that top end power (at the start of the season) was virtually the same as 2016 while bottom end power was lower. He describes a significant torque hole between 9000-10000 rpm which combined with higher than expected driveline inertia created very strong oscillations during gearshifts. Essentially, the engine side inertia was very low compared to driveline inertia, causing it to bog down under upshifts. As a consequence the driveability was very poor due to low end torque and having to shift at non-optimal engine speeds.
I see Honda as being solely responsible for the lack of engine performance, however in my view McL should take all responsibility for the issue of driveline inertia. It is very unlikely that they had communicated correct driveline specifications and Honda had chosen to ignore them.
Hasegawa then goes on to say that the low speed torque issue was addressed by a change in inlet manifold design (introduced in Barcelona). I suppose this had to do with the poor performance noted in media when testing the complete engine rather than the single cylinder. They also mention altering clutch settings - these I suppose have to do with reducing the clutch preload to allow a bit of slip and alleviate the torque spike.
Finally Hasegawa attributes the MGUH bearing failures to a new 'oil blowing' used for 2017. This is not explained at all but I suspect that oil blowing is just an air line from the 'foam side' of the oil tank that allows the bearing cavities to be scavenged. Unlike the crankcase which can be scavenged without an air line due to blow-by, other cavities typically require a 'breathing' orifice with a restrictor that is adjusted to achieve the required cavity pressure.
These scavenge breathers have been used for a very long time in dry-sump racing engines and are very well understood. I fail to see what could have gone wrong - most likely the oil was not well separated in the tank and made its way into the air line, effectively preventing the bearing from being scavenged? Could the excess oil on the bearing and shaft increase dynamic loads and cause bearing failures ?
Anyway, according to the article, it took Honda quite a long time to understand what was happening and in the end they had to modify the oil tank to stop this 'oil blowing' and increase the size of the bearing.
Unfortunately the way I see it the MGUH failures were entirely Honda's doing.
Of course there are good words it's the start of a new partnership and that is what you do.carisi2k wrote: ↑10 Dec 2017, 22:09While honda made some miscalculations on the 2017 engine. Mclaren shouldn't escape blame either as their insane push to size 0 obviously had implications for Honda. Good words seem to be coming out of Honda, Torro Rosso and Red Bull about the 2018 Torro Rosso Honda package.
I agree about sustained g loads, however CFD simulations are a good alternative to physical testing. Fuel tank and sump slosh have been done to death in both automotive and motorsport - here's an example from Sauber:MrPotatoHead wrote: ↑10 Dec 2017, 21:38
This is a dead end. It's impossible to simulate the sustained g forces on any of the existing engine test beds. You can simulate instantaneous but not sustained. And the Honda engineers should know what the g forces are - this was not their first season in F1.
If the people on this forum knew the predicted cornering forces then what excuse do the engine manufacturers have for screwing it up?
CFD is indeed used for this. And should have been in this case as well.Mudflap wrote: ↑10 Dec 2017, 22:19I agree about sustained g loads, however CFD simulations are a good alternative to physical testing. Fuel tank and sump slosh have been done to death in both automotive and motorsport - here's an example from Sauber:MrPotatoHead wrote: ↑10 Dec 2017, 21:38
This is a dead end. It's impossible to simulate the sustained g forces on any of the existing engine test beds. You can simulate instantaneous but not sustained. And the Honda engineers should know what the g forces are - this was not their first season in F1.
If the people on this forum knew the predicted cornering forces then what excuse do the engine manufacturers have for screwing it up?
https://www.youtube.com/watch?v=-Iyh-S-LGN8
Keep in mind though that longitudinal and vertical g loads are also important and can't be easily predicted without good knowledge of the dynamic behaviour of the car - knowledge that only McL would have.
Although I haven't seen any engine test bed but I can imagine putting an engine on a rotational arm then centrifugal force simulate sustained g forces.MrPotatoHead wrote: ↑10 Dec 2017, 21:38
It's impossible to simulate the sustained g forces on any of the existing engine test beds. You can
I said I can't tell whether McL have provided the data or if the data was correct. Similarly we don't know if Honda used the data or misinterpreted it. I've not seen anything about this in the press.MrPotatoHead wrote: ↑10 Dec 2017, 22:23CFD is indeed used for this. And should have been in this case as well.Mudflap wrote: ↑10 Dec 2017, 22:19I agree about sustained g loads, however CFD simulations are a good alternative to physical testing. Fuel tank and sump slosh have been done to death in both automotive and motorsport - here's an example from Sauber:MrPotatoHead wrote: ↑10 Dec 2017, 21:38
This is a dead end. It's impossible to simulate the sustained g forces on any of the existing engine test beds. You can simulate instantaneous but not sustained. And the Honda engineers should know what the g forces are - this was not their first season in F1.
If the people on this forum knew the predicted cornering forces then what excuse do the engine manufacturers have for screwing it up?
https://www.youtube.com/watch?v=-Iyh-S-LGN8
Keep in mind though that longitudinal and vertical g loads are also important and can't be easily predicted without good knowledge of the dynamic behaviour of the car - knowledge that only McL would have.
So you don't think McLaren shared with them expected loading from the car based on the simulations they already had data from?
That's about as close as you can get but that wouldn't really be a true dynamic test.amho wrote: ↑10 Dec 2017, 22:42Although I haven't seen any engine test bed but I can imagine putting an engine on a rotational arm then centrifugal force simulate sustained g forces.MrPotatoHead wrote: ↑10 Dec 2017, 21:38
It's impossible to simulate the sustained g forces on any of the existing engine test beds. You can
I know, I wasn't saying you said that - just playing devils advocate.Mudflap wrote: ↑10 Dec 2017, 23:56
I said I can't tell whether McL have provided the data or if the data was correct. Similarly we don't know if Honda used the data or misinterpreted it. I've not seen anything about this in the press.
When the input of more than one party is required in the design of a component it gets very woolly.
You nailed it, the fall of the partnership started with Honda designing a oil tank that can not withstand a hard turn in F1 when the G force is 4 or 5 times higher than normal. After 3 years really?!?! And then they claimed it was an easy fix which was never actually fixed, instead they implemented a solution which was a compromise so they can have the car working ''properly''
A design flaw of a oil tank is a technical issue, which was just put into context that Honda R&D did not come up with a proper redesign just a counter measure to prevent failure during the course of the last season.... that's all.HondaRaceReplica wrote: ↑11 Dec 2017, 15:25It would be nice if all the Mclaren "Fanbois" left and we could retuen this thread to technical and positive talk about the Honda Power Unit...You Mclaren guys are already with Renault.....Get on with your lives pls...