Honda Power Unit Hardware & Software

[gruntguru]

Mahle did initial testing on an Optical Test Stand and then moved straight to a test mule made from a real engine with a custom crank:http://i.imgur.com/B4sCKB0.jpg

That's a single cylinder engine too. Certainly doesn't help develop/wave-tune intake and exhaust plumbing.

Imagine the complexity of trying to do combustion R&D on a V6 turbo F1 engine? It would be exactly as (probably even more) cumbersome than I outlined in my first post. Repeat - there is NO WAY you would contemplate this.

There is no point running all six cylinders unless you have all the manifolding and turbo-machinery in place.

[MrPotatoHead]

Mahle did initial testing on an Optical Test Stand and then moved straight to a test mule made from a real engine with a custom crank:http://i.imgur.com/B4sCKB0.jpg

That's a single cylinder engine too. Certainly doesn't help develop/wave-tune intake and exhaust plumbing.

Imagine the complexity of trying to do combustion R&D on a V6 turbo F1 engine? It would be exactly as (probably even more) cumbersome than I outlined in my first post. Repeat - there is NO WAY you would contemplate this.

There is no point running all six cylinders unless you have all the manifolding and turbo-machinery in place.

You could most certainly run all six cylinders for R&D without having the turbo-machinery in place.
I don't know why you are arguing the point when I have seen it done this way.

The entire reason for Honda being where they are currently is from not testing all six cylinders early enough.

[Sasha]

Thanks for your view on things Wazari-san!
I have been critical myself of Honda at times (usually when Alonso gets another DNF :/) but you are absolutely right - most people on here have no idea what it takes to build a high end race engine let alone one of the current F1 engines. I have some involvement in F1 and other race engines myself so I see what most do not.
I know Honda will get things right given enough time... I just hope that it's not the eve of the engine formula changing again!

Btw the next time I am in Japan I would like to buy you a glass of sake or two! Reading your input on these forums has always been nice for me.

My permanent residence is in the US although I still maintain a residence in Tokyo. Whether in the US or Japan, I would never turn down a glass of good sake.

In terms of developing a PU from a single piston/cylinder, design/mock up/working model, it's the only economically and expediate way. To build 3,4 complete prototype PU's would consume even more time to develop. The powers at be want everything yesterday. From concept to actual mass produced road car is a 7 to 10 year process. Trying to cut down that time frame by a factor of 5 requires a lot of money, manpower and unfortunately short cuts which always don't work out the first time.

I will send you a PM. Perhaps we can meet for some sake sooner rather than later.

I do understand how single cylinder testing can cut down on development time but you can also build multi cylinder mule engines that you can use for full engine development testing.
In the grand scheme of things only the cylinder head "should" have been experimental.

And the cylinder head(combustion design) is the number 1 secret sauce to this formula,#2 secret sauce is ERS mapping.
So with that thought you would understand why they spent so much of their time testing concepts on single cylinder models.

[MrPotatoHead]

My permanent residence is in the US although I still maintain a residence in Tokyo. Whether in the US or Japan, I would never turn down a glass of good sake.

In terms of developing a PU from a single piston/cylinder, design/mock up/working model, it's the only economically and expediate way. To build 3,4 complete prototype PU's would consume even more time to develop. The powers at be want everything yesterday. From concept to actual mass produced road car is a 7 to 10 year process. Trying to cut down that time frame by a factor of 5 requires a lot of money, manpower and unfortunately short cuts which always don't work out the first time.

I will send you a PM. Perhaps we can meet for some sake sooner rather than later.

I do understand how single cylinder testing can cut down on development time but you can also build multi cylinder mule engines that you can use for full engine development testing.
In the grand scheme of things only the cylinder head "should" have been experimental.

And the cylinder head(combustion design) is the number 1 secret sauce to this formula,#2 secret sauce is ERS mapping.
So with that thought you would understand why they spent so much of their time testing concepts on single cylinder models.

Oh I fully understand why they would spend so much time on the single cylinder.
But they spent too much time on it. That is obvious.

[ringo]

Hey ringo!

Usually in automotive parts clearances are designed such that when you are up to operating temperature and pressure you have the correct clearances.
Piston to wall clearance is a good example of this.
Turbos typically are designed this way as well.
Now what Honda took into account with their design of the MGU-H I do not know for sure... Perhaps some one screwed up.

Understood. But on a bigger scale; maybe this can be considered clearance as well, the reason i called it a possible issue of design geometry, is because the clearances may well be fine for bearings and shaft,but there is a lot that can be done to manage the expected loads at different running conditions; as i will show with these crude drawings:

here we have the turbo housing with a fixed mounting (i'm not saying this is what honda is doing, but its just hypothetical for demonstration purposes). You can see where the hot turbo expands and pushes against the mounting causing the housing to move upward; and subsequent misalignment.


this other design has two slotted mountings now, but when the turbo expands it pushes outward on the mountings evenly in both directions. this doesn't cause misalignment issues.

keep in mind i'm no expert on this. Just an example of a simple design aspect that addresses the problem on additional loads introduced at high temperatures.

For all we know, this engine could be a science project for some university over there in Japan. Maybe the designers arent considering everything or simply dont have lessons learned from previous experiences.

[MrPotatoHead]

Hey ringo!

Usually in automotive parts clearances are designed such that when you are up to operating temperature and pressure you have the correct clearances.
Piston to wall clearance is a good example of this.
Turbos typically are designed this way as well.
Now what Honda took into account with their design of the MGU-H I do not know for sure... Perhaps some one screwed up.

Understood. But on a bigger scale; maybe this can be considered clearance as well, the reason i called it a possible issue of design geometry, is because the clearances may well be fine for bearings and shaft,but there is a lot that can be done to manage the expected loads at different running conditions; as i will show with these crude drawings:

here we have the turbo housing with a fixed mounting (i'm not saying this is what honda is doing, but its just hypothetical for demonstration purposes). You can see where the hot turbo expands and pushes against the mounting causing the housing to move upward; and subsequent misalignment.
http://rs1010.pbsrc.com/albums/af226/du ... 0&fit=clip

this other design has two slotted mountings now, but when the turbo expands it pushes outward on the mountings evenly in both directions. this doesn't cause misalignment issues.
http://rs1010.pbsrc.com/albums/af226/du ... 0&fit=clip
keep in mind i'm no expert on this. Just an example of a simple design aspect that addresses the problem on additional loads introduced at high temperatures.

For all we know, this engine could be a science project for some university over there in Japan. Maybe the designers arent considering everything or simply dont have lessons learned from previous experiences.

You know the more I think about it if they did rigidly mount the MGU-H or have the MGU-H as part of the engine block then you are totally right in that this could be the cause of their problems if they did not account for thermal growth.
I've been assuming they would have done it right. Maybe I'm giving them too much credit and someone simply screwed up.

[Sasha]

I will send you a PM. Perhaps we can meet for some sake sooner rather than later.

I do understand how single cylinder testing can cut down on development time but you can also build multi cylinder mule engines that you can use for full engine development testing.
In the grand scheme of things only the cylinder head "should" have been experimental.

And the cylinder head(combustion design) is the number 1 secret sauce to this formula,#2 secret sauce is ERS mapping.
So with that thought you would understand why they spent so much of their time testing concepts on single cylinder models.

Oh I fully understand why they would spend so much time on the single cylinder.
But they spent too much time on it. That is obvious.

They just tried to do three years of R&D in one year.They didn't have the time MB had to get their PU correct.Honda needed that extra year but McLaren pushed them to race in 2015.With so many problems they are fixing more fires than doing development, it is eating up all their man hours.

[MrPotatoHead]

And the cylinder head(combustion design) is the number 1 secret sauce to this formula,#2 secret sauce is ERS mapping.
So with that thought you would understand why they spent so much of their time testing concepts on single cylinder models.

Oh I fully understand why they would spend so much time on the single cylinder.
But they spent too much time on it. That is obvious.

They just tried to do three years of R&D in one year.They didn't have the time MB had to get their PU correct.Honda needed that extra year but McLaren pushed them to race in 2015.With so many problems they are fixing more fires than doing development, it is eating up all their man hours.

Agreed.

[f1rules]

And the cylinder head(combustion design) is the number 1 secret sauce to this formula,#2 secret sauce is ERS mapping.
So with that thought you would understand why they spent so much of their time testing concepts on single cylinder models.

Oh I fully understand why they would spend so much time on the single cylinder.
But they spent too much time on it. That is obvious.

They just tried to do three years of R&D in one year.They didn't have the time MB had to get their PU correct.Honda needed that extra year but McLaren pushed them to race in 2015.With so many problems they are fixing more fires than doing development, it is eating up all their man hours.

I dont think giving honda 1 more year would have changed a thing, just see their history of simulation vs ontrack. Seems there is absolutely no correlation between what honda simulate and real life. To learn they are also under funded, and i was under the impression they threw absolute everything at it, the dissapointment with them is just unbeliveable, and seems mclaren thankfully think so also, why on earth enter with a team like mclaren if youre not willing to allocate the necessary resources, fail on all fronts unfortunately

[gruntguru]

Mahle did initial testing on an Optical Test Stand and then moved straight to a test mule made from a real engine with a custom crank:http://i.imgur.com/B4sCKB0.jpg

That's a single cylinder engine too. Certainly doesn't help develop/wave-tune intake and exhaust plumbing.
Imagine the complexity of trying to do combustion R&D on a V6 turbo F1 engine? It would be exactly as (probably even more) cumbersome than I outlined in my first post. Repeat - there is NO WAY you would contemplate this.
There is no point running all six cylinders unless you have all the manifolding and turbo-machinery in place.

You could most certainly run all six cylinders for R&D without having the turbo-machinery in place.

Yes you could, but the problems Honda experienced when they transitioned to 6 cyl testing were related to wave tuning and interaction of pulses in the intake and exhaust manifolds and the turbine housing (which itself forms a 2-1 manifold). Those issues could not have been analysed fully (and rectified) without the turbo-machinery in place.
.
.

I don't know why you are arguing the point when I have seen it done this way.

Because you posted the following piece of nonsense and still seem to be arguing it.

As an Engineer myself I very much agree with you. If it was me I would probably skip the single cylinder stage all together.

[MrPotatoHead]

I don't know why you are arguing the point when I have seen it done this way.

Because you posted the following piece of nonsense and still seem to be arguing it.

As an Engineer myself I very much agree with you. If it was me I would probably skip the single cylinder stage all together.

I fail to see how it is nonsense when it is how most race engines are developed.

[gruntguru]

Oh well - I can start by re-posting the following and we can repeat the last two pages ad-infinitum or . . you could just go back and read it all again?

Impossible to develop an engine like this in a reasonable timeframe without doing the combustion R&D on a single cylinder engine. To make incremental changes to a full V6 mule involves - for each change made:

6x component build
2 - 3x assembly/disassembly time
6x resources such as fuel, oil, dyno capacity
3 - 4x instrumentation
extra mapping time to optimise individual cylinders
etc
etc

[wuzak]

As an Engineer myself I very much agree with you. If it was me I would probably skip the single cylinder stage all together.

Oh I fully understand why they would spend so much time on the single cylinder.
But they spent too much time on it. That is obvious.

You're saying two different things here?

[wuzak]

I fail to see how it is nonsense when it is how most race engines are developed.

Most race car engines - does that include the multitude of race engines developed from production engines, or are you talking specifically about prototype racing engines?

[wuzak]

Typically single cylinder testing is done on a single cylinder unit. Not a full engine converted to a single cylinder.
Using the full engine mule enable quicker modular testing with anything from 1 to 6 cylinders in this case.
For instance you build a complete cylinder head and then when you are happy with 1 cylinder you can move on to 6 without starting from scratch.

Doesn't this mean you have to build two complete cylinder heads (for however many cylinders)?