Honda Power Unit Hardware & Software

[godlameroso]

A larger compressor, has a larger turbine, it produces more torque than a smaller compressor with a smaller corresponding turbine. So not only do you not need to spin the compressor to generate enough boost, similarly the turbine doesn't need to spin as fast, so you can use overdrive to get the MGU-H to spin to max allowed speed. This produces more torque on the mgu-H, but less torque when it comes to using the MGU-H to spool the turbo hence more charge can be drawn from it. Conversely if you have a very fast compressor/turbine, you have to use gear reduction to get the MGU-H to spin to max allowed speed, so torque is lost on regen but gained on spool up.

[trinidefender]

A larger compressor, has a larger turbine, it produces more torque than a smaller compressor with a smaller corresponding turbine. So not only do you not need to spin the compressor to generate enough boost, similarly the turbine doesn't need to spin as fast, so you can use overdrive to get the MGU-H to spin to max allowed speed. This produces more torque on the mgu-H, but less torque when it comes to using the MGU-H to spool the turbo hence more charge can be drawn from it. Conversely if you have a very fast compressor/turbine, you have to use gear reduction to get the MGU-H to spin to max allowed speed, so torque is lost on regen but gained on spool up.

Errrrrrrrrrr.....what the........

Learn about gearing. When you gear down something I.e. The output shaft has a slower speed than the input shaft then the torque will increase linearly (assuming no drivetrain loss). That means that you don't lose anything on regen or gain anything on spool up. Therefore a smaller turbine that spins faster but produces less torque on its shaft can then use gearing to reduce the rpm on the output shaft under the regulated maximum for the MGU-H but have a higher torque than what the turbine is showing.

A compressor and turbine are chosen individually. Yes there is some correspondence in size but changing the size of one does not mean you HAVE to change the size of others. A larger compressor inlet size (referring to centrifugal compressors here) will generally have its peak efficiency range further up on the massflow axis of a graph.

Let us look at how a centrifugal compressor compresses air. It spins it and the centrifugal forces force the air outward into a diffuser (this is where velocity gets traded for pressure). Let's say we have a compressor flowing X amount of air at Y pressure. Say we want to increase this pressure. We can do 1 of two things, increase the diameter of the compressor outlet which increases the velocity of the airflow going into the diffuser (which in turn increases the pressure increase as the air moves through it) or increase the compressors rpm which also increases the velocity of the airflow with the same end result.

It appears the Mercedes and other teams have chose a larger diameter compressor. Therefore at a lower rpm they can achieve the same PR (pressure ratio). Honda's design team chose to pursue he route of increasing the rpm of the compressor to try and achieve the same PR.

Part 1.

[trinidefender]

Part 2.

Turbines can generally either be impulse or reaction turbines. F1 turbocharger turbines are most likely some combination of the two.

Turbines also have their own efficiency range. When you increase a turbines corrected mass flow past its general efficiency range you will still get an increase of torque on the shaft to spin the compressor and MGU-H, however you will also get a large increase in back pressure. This back pressure is felt by the ICE which then has to use more force from its crank shaft pumping airflow out of the cylinders. This reduces the available torque going to the rear wheels. That is why back pressure is generally bad. Also lastly, Past a certain point the turbine will become choked. Trying to increase the flow through the turbine beyond this point to provide more torque to the shaft will increase back pressure by a large amount

Conversely, a smaller turbine will take a lower corrected mass flow to provide a usable level of torque to the compressor and/or the MGU-H. This is why a smaller turbine will generally spool up quicker at lower RPM's and lower power levels.

Here is the problem it appears Honda is having from what information we as a community can gather. It appears they have gone with a "smaller" (i use quotation marks as saying smaller is not an absolute, there are to many variables that go into turbocharger turbine design) turbine. This means that the MGU-H has less work to to do to keep the shaft spinning at lower then 100% throttle loads and below peak power rpm when the turbine is not seeing enough corrected massflow to provide enough torque to the compressor for the required PR. conversely at high mass flows indicative of peak power rpm and full throttle the turbine might be producing enough torque with a low enough back pressure to power the compressor, this will mean there is very little torque on to turbocharger shaft left to turn the MGU-H. Lets say that they then wanted to also generate electricity from the MGU-H. How would this happen you ask? Well the turbine will now have to provide extra torque to the shaft. It will do this by slightly closing the wastegate slightly forcing more exhaust flow through the turbine, because the turbine is now past its peak efficiency range the backpressure will increase rapidly as the load on the shaft from the MGU-H increases. So therefore, while the MGU-H can create electricity it comes at the expense of drastically reduced ICE performance.

Sorry this seemed so jumbled up its just that this isn't really a topic that can be covered in a few posts. There are people who's main design field that they spend most of their life working on is compressors and others, turbines.

[godlameroso]

But I'm talking about torque delivered from the turbine shaft to the mgu-h, and seeing as all 3; compressor/mgu-h/and turbine are all on the same shaft, and the mgu-h can be clutched and geared to the turbine, changing either turbine, or compressor will affect regen torque applied to the mgu-h. Also the spool up characteristics and speed which the compressor and turbine are driven at directly affects the ability of the mgu-h to charge in tandem with the clutching and gearing mechanism. So do you use a high speed low torque turbine compressor combo to create mgu-h regen charge, or a big torquey compressor/turbine? We know that Honda cannot deploy as much as other teams PRECISELY because their mgu-h is lacking, why is my proposition such a leap of logic, when we see how well it works for Mercedes, how well it worked for Ferrari?

Again, I totally think that Honda can make their low torque high speed system work, they'll either have to compromise something, or change the layout to accommodate a more adequately sized compressor/turbine/modified mgu-h.
**Perhaps even going so far as to creating a special shape within the block casting to accommodate the larger compressor/mgu-h/turbine combo. Maybe that's what they already did and why they have such problems with seals and over heating. It's probably quite tricky to create water jackets, and oiling holes needed for such an arrangement.

P.S. I enjoy what you're posting please continue

*I wonder if the trick to being able to minimize back pressure is to figure out a way to design a compressor that can function in both normal and blow down mode, and still generate enough boost and torque to charge the mgu-h.

[gruntguru]

Turbine and compressor selection is actually far more pre-defined than some here are thinking. The only real variablility is the piston engine operating parameters (boost, BP and flow). The tradeoff is crankshaft energy vs MGUH energy. Once the engine developers have established the combination that produces the maximum total energy from the fuel allocation, the compressor and turbine selection are almost automatic - the flow rate and the delta P for each is known. It only remains to select/design a compressor and turbine that have the highest possible efficiency at the chosen operating point. Of course other things will demand some secondary consideration - efficiency at other operating points, rotating inertia, packaging etc etc.

Generally speaking, the smaller the turbomachine the lower the efficiency, due to clearance leakage, skin friction etc. The difference is not going to be that great across the F1 teams however and even the difference in size (tip diameter) itself will not be great. An external view does not give a good indication of size. Folded volutes can deliver a much smaller appearing machine than a radial volute like the Mercedes compressor. OTOH there is probably a slight efficiency penalty in the folded volute design due to extra turns required in the gas path.

[trinidefender]

Turbine and compressor selection is actually far more pre-defined than some here are thinking. The only real variablility is the piston engine operating parameters (boost, BP and flow). The tradeoff is crankshaft energy vs MGUH energy. Once the engine developers have established the combination that produces the maximum total energy from the fuel allocation, the compressor and turbine selection are almost automatic - the flow rate and the delta P for each is known. It only remains to select/design a compressor and turbine that have the highest possible efficiency at the chosen operating point. Of course other things will demand some secondary consideration - efficiency at other operating points, rotating inertia, packaging etc etc.

Generally speaking, the smaller the turbomachine the lower the efficiency, due to clearance leakage, skin friction etc. The difference is not going to be that great across the F1 teams however and even the difference in size (tip diameter) itself will not be great. An external view does not give a good indication of size. Folded volutes can deliver a much smaller appearing machine than a radial volute like the Mercedes compressor. OTOH there is probably a slight efficiency penalty in the folded volute design due to extra turns required in the gas path.

I wasn't be F1 specific, I was just talking about a few of the principles of compressor and turbine design for any situation. I brought in the part about running the MGU-H and compressor off of the turbine for the simple reason that that is what seems to have thrown off the Honda design team the most

[NL_Fer]

I won't be suprised if Honda went for a lower target rpm next year and increase the comoressor, as much as possible while keeping it in the V. The turbine can be increased also, increasng harvesting.

Still, it would be limited, since it's inside the V. Wonder f they could run ERS on a full Monza or Spa lap.

[ME4ME]

I think they will go with a very similair design to Mercedes. Alonso said so much as they knew what was wrong and they would copy their competitor's solution. So probably a big compressor at the front, MGU-H in the middle, tubulair exhaust manifold (which they've already switched to) and increaded cooling and size for the MGU-K and -H.

[Joseki]

I won't be suprised if Honda went for a lower target rpm next year and increase the comoressor, as much as possible while keeping it in the V. The turbine can be increased also, increasng harvesting.

Still, it would be limited, since it's inside the V. Wonder f they could run ERS on a full Monza or Spa lap.

IIRC Fernando said they had just the 33 seconds of boost from the battery, so pretty much just from La Source to Les Combes and a few seconds on the kink back to Bus Stop. Mercedes, Ferrari and Renault also had the energy they harvested from the MGU-H to use the MGU-K longer.

[toraabe]

Exactly. They need to put it in the front of the vee. Then the whole package can be lowered like mercedes layout. I also think that Renault will do the same. They just have to. Another benifit are shorter and lowered exhaust and charging pipes in which is essential in weight distribution.

I think they will go with a very similair design to Mercedes. Alonso said so much as they knew what was wrong and they would copy their competitor's solution. So probably a big compressor at the front, MGU-H in the middle, tubulair exhaust manifold (which they've already switched to) and increaded cooling and size for the MGU-K and -H.

[wuzak]

I won't be suprised if Honda went for a lower target rpm next year and increase the comoressor, as much as possible while keeping it in the V. The turbine can be increased also, increasng harvesting.

Still, it would be limited, since it's inside the V. Wonder f they could run ERS on a full Monza or Spa lap.

IIRC Fernando said they had just the 33 seconds of boost from the battery, so pretty much just from La Source to Les Combes and a few seconds on the kink back to Bus Stop. Mercedes, Ferrari and Renault also had the energy they harvested from the MGU-H to use the MGU-K longer.

33s would mean the full 4MJ, which would suggest that they can't run the MGUH directly to the MGUK.

[NL_Fer]

MGU-K is limited to harvest 2 MJ, that's only 16s. Most tracks have about 45s of full throttle, that is about 5,5 MJ. So the MGU-H needs to harvest 3,5 MJ, that's the target.

Before a qualifing lap, they can charge the ES to 4MJ and drain it during their hot lap, so only 1,5 MJ needs to harvested.

[Tommy Cookers]

to me this seems a good time to ask for clarification ....
does (anyone's) ES actually store 4MJ ?
does anyone actually have even as much as 45 sec per lap of 120 kW mu-k action ?

and does the term harvesting have a clear meaning here ?
(lap net gu-h recovery to store or lap gu-h recovery to realtime (direct) mu-k action ?)
harvesting might imply storage

[R_Redding]

to me this seems a good time to ask for clarification ....
does (anyone's) ES actually store 4MJ ?

I'd expect so .... 4MJ is the maximum energy differential between Min and Max on the batterys charge state.You'd want it fully charged to get the max power allowed from it.

does anyone actually have even as much as 45 sec per lap of 120 kW mu-k action ?

4MJ max from Es to MG-Uk/engine ancilliaries per lap... so they have 33.3 secs of MG-UK at a rate of 120Kw (Watts=Joules/Secs so 4Mj/120kW = 33.3 secs)per lap from ES.
To get your 45secs they would need need an extra 11.7 secs 0f 120kW max MG-UK power from the MG-H (1.4MJ).. Wether they would use the energy streams sequentually or combined is all down to mapping and control methodology.

and does the term harvesting have a clear meaning here ?

Its not a bad description for all the various energy paths available ... to and from MG-*s via ES or not.

Rob

[R_Redding]

... duplicate