Featured SkyActiv-3 and Project One. The future of efficient hybrid engines

Adding to @hill, in 2001 I compared the Insight and Gen-1 Prius looking at the electrical capacities. The Gen-1 Prius had more than twice the electrical capacity of the Insight which at the time I attributed the much higher city 52 MPG over highway 45 MPG (EPA specs at the time.) In contrast, the Insights had lower city 60 MPG versus highway 66 MPG. So my logic was:

• City MPG > Highway MPG - this means the car is handling the lower speed, multiple stop/start EPA test than the Insight. This must be due to better usage of the electrical capabilities.
• Gen-1 total MG1+MG2 power > Insight IMA - at the time I did not realize we would have to wait to 2017 and the one-way clutch to combine MG1+MG2. Later I learned the IMA was providing low-end torque for the small Insight engine. Regardless, I knew that more electrical power translated into better city and highway MPG.

I rented a Gen-1 for two days but could not get the EPA 52 MPG; it did not have cruise control; the dealer wanted $21k; I hated the puke green, and; did not want their GPS map. So we bought a 2001 Echo and after adding all the Prius-like options my wife wanted, $14.2k. I continued to commute in a 1991, manual transmission, Camry.

Turn the clock forward to September 2005 and we lose our Camry in a rain slick accident. So I bought a used, 2003 Prius, black, with cruise control, and no GPS in Fort Worth, 800 miles away. My first leg home, the car only gave me 39 MPG but I was keeping up with Texas highway traffic and I almost thought about driving back. Instead, I drove the next segment at 50 mph and measured 53 MPG. Eventually I plotted this graph:

It turns out there was a knee in the Gen-1 mileage curve, 65-70 mph, and I never looked back.

So this is how I see it:

1. IMA or BAS - these 'fool' the designers who treat them as low-speed, engine torque. This is what fooled Honda and GM which means the motor enhanced engine has to be used for all vehicle power. An engine that burns gas. The voltage doesn't matter as the first Insight used 144V. Heck, I don't care if they use 48V, 12V, or like my first MicroBus, 6V. It is a dead-end with limited efficiency and no answer for inertial losses (i.e., regenerative braking.)
2. Blue Motion - using a single motor that can disconnect from the engine, this system can work with a stepped transmission which is its own problem. The stepped transmission tries to have overlapping speed-vs-efficiency speed ranges which the car has to hunt with pauses between gears.
3. HSD and variants - using MG2 as an independent motor, this eCVT optimizes the engine and motor.
4. HSD with one-way flywheel clutch - finally MG1 and MG2 can operate as a single motor yet retain the engine running at peak efficiency.
5. Range-extender EV - done right, a small, 35-45 hp, high efficiency engine like a SkyActive-3 can provide the energy needed for affordable, long distance, highway travel. The engine only needs to provide up to the top highway speed power level and no more as the battery handles peak loads.

We own both #4, 2017 Prius Prime, and #5, 2014 BMW i3-REx. This isn't a paper exercise but real vehicles having mass and the key-fobs within arm reach.

Bob Wilson

 

Not necessarily - they are claiming 41% engine thermal efficiency, but there's a few other factors to consider.

That 41% is a peak number for both engines. I'd be unsurprised if the Mercedes engine is making its 41% thermal efficiency at a higher power point, which would mean more energy would need to be absorbed into the battery to maintain speed, reducing overall efficiency.

Driveline efficiency is also a consideration - this is a longitudinal drivetrain, which tends to be less efficient. However, it's a geared transmission, which tends to have tradeoffs - sometimes improved driveline efficiency, but less options to keep the engine on its peak.

Sure, it requires bigger MGs, but the more powerful vehicles tend to need more power all the time, so the bigger MGs don't hurt you that much.

And, there's other ways to get things to work in your favor as far as power delivery - the multi-stage hybrid system is one way, as well as 2-mode systems.

 

Insight also had way worse emissions, thanks mostly to "lean burn" which is anything but!

Games. All the auto makers play them. Toyota, Honda, Tesla included.

 

HSD efficiency is based on the ratio of engine RPM to vehicle speed. If MG1 is at 0 RPM, HSD efficiency is maximized. The faster MG1 is spinning, the more power it has to absorb off of the ICE to send to MG2 (or vice versa in some instances where ICE RPM is lowered below the natural point by power being pulled off MG2 and sent to MG1).

 

Oh, meant to reply to this.

So, lean burn simply means that you're burning less fuel than stoichiometric (14.7:1 air to fuel ratio). They were doing that, and that improves efficiency. (Wider throttle opening for the same amount of fuel burned, meaning reduced pumping losses.)

The problem is that air is mostly nitrogen. In high temperatures, nitrogen reacts with oxygen to produce nitrogen oxides.

In most gasoline engines, the first solution is to simply not run lean. This means that the oxygen in the cylinder is consumed by combustion, and isn't available to react with nitrogen (for the most part - this assumes perfect mixture). However, that hurts efficiency - the engine has to work against a closed throttle plate at partial load.

Now, mixture isn't ever perfect, so the three-way catalytic converter is used to clean up what NOx is in the exhaust (as well as other things). Note that it becomes ineffective (or even counter-productive) for NOx reduction if you have excess oxygen, meaning a three-way cat cannot work properly on a lean burn engine.

Exhaust gas recirculation is another way to deal with this. Exhaust gases won't have excess oxygen, and effectively act as an inert gas in the cylinder, giving you some of the benefits of lean burn (reduced pumping losses), without the oxygen. This works excellently in gasoline engines, and diesels use it as well. (I'll get to diesels.) Cooling exhaust gases before feeding them into the intake helps even more, by reducing cylinder temperature, although many modern gasoline engines instead use variable valve timing on the exhaust camshaft to pull gases back in through the exhaust valves. (The Gen 3 and Gen 4 Prii use cooled EGR in the conventional way - pulled from the exhaust, and piped into the intake.)

The intake valves can be controlled in a manner to reduce intake air mass, without increasing pumping losses as a throttle plate would. Various ways to do this exist - variable valve lift systems like BMW's Valvetronic and Toyota's Valvematic can restrict how far the intake valves open (reduced pumping losses compared to a throttle plate), variable valve timing systems like the system used on the Mazda SkyActiv-G engines and Toyota's VVT-iW and VVT-iE systems can be used to delay valve closure during partial load (pushing excess air out), Fiat's MultiAir system can do both valve opening restriction as well as closing the valves early (not letting excess air in in the first place), or you can fix intake valve timing to always close the valves early (as in the Volkswagen Budack cycle engines) or late (as in Toyota and Ford hybrid applications). In all of these configurations, intake air mass is reduced without significantly increasing pumping losses. (Note that in engines that always close the valves early or late, this is done to improve thermal efficiency across the board, at the expense of power and torque, rather than to handle a partial load condition.)

Diesels use two technologies (sometimes one, sometimes both) that are rarely seen on gasoline engines. Lean NOx traps trap the NOx during driving, and when full, the engine switches to a rich mode of operation (which generates high particulate matter (due to the poor mixture quality inherent to diesels, due to a less volatile fuel) and is inefficient, but is only done for a short time) to burn off the NOx. Selective catalytic reduction uses ammonia to reduce NOx. (Both can be combined, and one of the emissions of a LNT is ammonia, so it actually helps the SCR. Also, LNT works best at low load, SCR works best at high load, so they can complement one another.) All of this is incredibly expensive, LNT isn't very effective, and SCR requires that a tank of urea solution be filled every so often, though.

Finally, hybridization itself can be used to help. The electric motor can work against the engine, increasing load and efficiency, and using the excess engine power to charge the battery for later. However, given how undersized Honda's batteries are, they likely didn't want to do this too much. And, the Honda IMA system cannot shut down the engine while the vehicle needs power, so the effectiveness of this approach is greatly reduced as compared to a system like Toyota's.

 

Note that there is another 2-mode system, simplified slightly from the truck one, used in the Cadillac CT6 PHEV, paired with a 2.0T. Engine power is 266 hp, total system power is 449 hp.

Going back to Toyota, for 3.5 liter systems... they've revamped the RWD systems for the LCh and LSh (now using a V6 instead of the V8), and sure, if you use the power, you'll use a lot of fuel, but of course you will, you're using a lot of power. The FWD systems (think HiHy and RXh), however, while the engines were updated, they're still the old 2006 hybrid systems, it's not like the new Camry's system. (That said, the next-gen HiHy will be getting the Camry's 2.5 liter system, based on a press release about that system's production being brought to the US.)

As far as forced induction, why? All that'll do is require boost enrichment at high load, which will utterly kill your efficiency. More displacement is far better than boost, especially with a hybrid system that can avoid low-load scenarios.

Realistically, a 2-speed multi-stage device could be helpful on the FWD powertrains - note that the 4-speed multi-stage device allowed Toyota to get better off-the-line acceleration with a 3.5 V6, compared to the 5.0 V8 in the previous LSh (with 2-speed MG2 reduction, and single-speed PSD output).

 

Regarding "Furure of Hybrid Engines"
Just finished an interesting read regarding the state of traditional auto makers - continuing on with existing hybrid systems & engine efficiency. Even as GM sold 10's of thousand's of PHEV & EV's last year - their industry leadership scratched their heads when 100's of thousands chose to wait for a model 3, instead.

In short - the main thrust of the author is how traditional / successful industries see no real need to consider doing it differently - because what they do, already works.
Remember 1½ decades ago? - how "the Prius killer" was just around the corner? So - what's "just-around-the-corner" now? Like "the Prius killer", the "Tesla killer" will be here .... even as the traditional industry continues to tinker with more & more efficient ICE & hybrid configurations.
"Culture Eats Strategy"
it's phrase we understand, & i'd never heard anyone articulate it.
Tesla Competition: Culture Eats Strategy!
.

 

The original Insight was ULEV certified. Yeah, it emitted more than the SULEV Prius, but it was much better than most other cars available.
It was also one of those few gasoline cars using a NOx trap.

 

Source: http://stuff.jaygroh.com/prius/Prius%20Info/Bonus%20Stuff/The%20Prius%20That%20Shook%20The%20World/The%20Prius%20That%20Shook%20The%20World.pdf

The history of the Prius suggests Toyota had the ability to create and sustain the Prius line when everyone, including Toyota, was printing money making gas powered cars. I'm of the opinion the author is ignorant of how the Prius came into being and resentful that Toyota didn't jump into EVs.

Successful companies do change. Jack Welch proved you can manage a huge, conglomerate company into something much smaller. Instead of working the synergies between divisions, he simply sold off those not meeting his short term criteria. He was no different from any other vulture capitalist.

Bob Wilson

 

I agree the ICE is not dead but rather the whole car has to work. An ICE centric architecture will always struggle against a blended EV and ICE. It is the old Insight/Civic vs Prius contest that dates back 18 years.

Bob Wilson

 

It's a 2 liter turbo, though, and what's being questioned is the power handling capability of the system, which is why I quoted engine horsepower.

Citation on it being not as refined?

That's... actually really terrible (and it's 62, not 64, for 2018). My guess is that a lot of the problem is related to it being rear-wheel drive, which creates a double-whammy of inefficiency. First, being a longitudinal powertrain, that means that the power has to go through a power-sapping bevel gear before making it through the rear wheels. Second, RWD vehicles can't do as much regenerative braking, hurting efficiency more.

I feel like this misunderstands why the MG2 power rating is so high?

Essentially, total system power is (roughly) engine power plus battery power. To handle all situations without having to reduce engine output, MG2 power should be MG1 power plus battery power - that way, the maximum power that MG1 can pull off of the engine can be sent to MG2, as well as the maximum power that the battery can supply. (That said, if you don't mind being MG2-limited, having a more powerful battery as in these PHEVs absolutely helps efficiency.)

My best guess is that they simply weren't ready to update to the parallel axis system on that platform, note that the HiHy/RXh are still on the old K platform (that dates back to the 2001 Highlander). The next-generation Highlander is clearly moving to TNGA, I'd guess the same would be true of the RXh.

More power, yes. More efficiency at higher power levels, eh, not really? Again, boost enrichment (I'll get to your comments about the Mazda engines though), and note that the boosted engines get worse mileage in the real world.

Lower weight and more low-end torque, yes.

Lower total friction and total pumping losses, yes... but if you're also getting more power, the hybrid system can save that power for later, and then shut down the engine, making that a non-issue, as long as the hybrid system is sufficiently flexible.

And, better thermal efficiency in most operating conditions? Show me a road-legal turbocharged gasoline engine that's 40% efficient at peak or better (again note that the hybrid system can be designed to run the engine at a minimum power output of wherever the peak is, unlike non-hybrids that are forced to spend time below peak), that isn't an absurdly detuned Formula One engine. Meanwhile, I can show you four naturally aspirated ones - the ESTEC 2ZR-FXE as used in the Gen 4 Prius, the Hyundai Kappa III 1.6 Atkinson as used in the Ioniq and Niro, the A25A-FKS as used in the Camry, and the A25A-FXS as used in the Camry Hybrid.

It's a bit more complicated than just cooled EGR, I'd argue, although that absolutely helps - they're not intentionally sucking hot gases back into the cylinder with retarded exhaust cam timing. (Cost is really the reason why manufacturers don't do that, of course.) However, the really clever bit with Mazda's setup is arguably the exhaust manifold setup, although they've done similar things for naturally aspirated engines.

Three problems with their setup, though.

First, it's a bear to package (this is the 4-2-1 setup off of a 2014 Mazda3):


This caused problems for the 2012-2013 Mazda3 with the 2.0 SkyActiv-G (which was a pre-SkyActiv platform), as well as the CX-3 with the same engine (a smaller version of the SkyActiv platform) - there wasn't room for that manifold, so they had to reduce compression and detune the engine. Basically, the engine bay has to be sized with more room between the rear of the engine and the firewall.

Second, while they've figured it out, there can be problems with catalytic converter warmup in such a design - the exhaust has to flow through a lot more piping to get to the cat.

Finally, cost. That thing ain't cheap at all.

Constant speed is less of a question, it's acceleration. And, in the city, acceleration happens a lot more, and people do a lot more city driving than they think they do. This is especially when manufacturers use undersized turbo engines, which is especially rewarded in the European market (where you see things like 1.0Ts in midsize sedans, and the things are always in boost and always in boost enrichment, and people wonder why they guzzle fuel).

I'm not incredibly opposed to... design a reasonable amount of base power, then turbocharge for a high performance variant, because then in normal driving you're not really using the boost and getting enrichment. But it does increase complexity and cost for a "normal" variant, and in a hybrid, you won't see as many of the benefits (because a sufficiently good hybrid system can prevent low-load low-efficiency operation, but it can't hide high-load low-efficiency operation as well).

 

The fact that Toyota was getting government help in Japan gets overlooked in histories of the Prius. There was a MITI LEV program looking to reduce car emissions that dates back to the '70s that was a partnership between the Japanese government and industry. It looked at BEVs back when NiMH was bleeding edge technology and most EV cars used lead acid packs. The group think at Toyota that the public only wanted short range city BEVs and hydrogen FCEVs are better stems from that.

The final stages of that program, or outgrowth of it, was subsidies for electric and hybrid car sales. From the moment it was released, Toyota was getting a direct subsidy equal to half the price difference between it and the Echo for each Prius sold. In a moment of serendipity, the $7500 plug in tax credit is half the difference between the gen1 Volt and equivalently equipped Cruze.

The tl;dr is that Toyota has always been a fast follower company. They'll drink from the government trough with everybody else, and will work to make that better benefit them. They got worried when excluded from the US PNGV program, but likely would not have had the public resources available for the Prius if the Japanese government and people didn't want to clean up the air in their cities.

IMA failed mostly because of Honda's battery management, and the resulting fallout from how they handled wronged customers. Price/cost was also a factor, but some of that was simply because it was Honda and their dealers.

BAS failed because of price, and I won't be surprised if old GM over promised with it. The gen2 system did better, and part of it was because they didn't call it a hybrid. Still not a big seller, but then neither was the Malibu it was apart of, its sales did help prop up hybrid statistics. The gen3 is now an option on their pick up truck, and I think it will be standard on the base engine of the new LaCrosse.

The Hyundai hybrid system is essentially BAS with a more powerful IMA. The Ioniq gets all the press here, but the Sonata hybrid has quietly gone into its second generation without major issues and disappointments. We will likely see more full power parallel hybrids coming to market as they might be lower cost than the equivalent power-split. They have required finer control of the subsystems than was possible upon the Prius's introduction for smooth integration of the ICE and motor, which could have been an issue for IMA.

That integration of operation and design is important. The high improvement claims of 48 volt hybrids assumes the car and engine is designed with them in mind, and not just simply bolted on. The roadmap for these systems will be BAS at first, but move to an improved redo of IMA.

 

IMHO it is a "Blue Motion" system with an independent shaft into the transmission. I first saw the engineering mockup in the January 2009, Detroit Auto Show:




It is the transmission connection of the motor/generator that makes this a distinct difference from the BAS and IMA. It means the car can move with the engine off and not spinning. But this is not a 48V system.

Bob Wilson

 

They are all a parallel hybrids with the traction motor between the engine and transmission. In addition to the BlueMotion, Hundai's system is close to Honda's newer double clutch system that is in the Jazz hybrid. Not a full hybrid, and lacking clutches to isolate the motor and engine, IMA was the engineering amphibian to these later systems. We all kinda looked like tadpoles at one point.

In addition to potential cost advantages, current car manufacturers may favor parallel hybrids because they are easier to work into their current design process. While mechanically simpler, power-split hybrids required a whole new transmission to be designed for them.