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

The Internal Combustion Engine Is Dead? Mazda Has Another Idea With Innovative SkyActiv-X
How The Mercedes-AMG Project One Brings F1 Tech to the Street

Mazda has cracked the code for a compression ignition gasoline engine by using a spark plug to start a small pressure wave which controls the timing of a diesel like compression ignition, with similar efficiency and the ability to do lean burn (about 1/2 stoich fuel to air ratio) that supplies a cooler burn and lower emissions before pollution control. The ability to this lean means that the power band close to peak efficiency is much broader. This should be available in a car around the end of next year, and according to the forbes article toyota is likely the only other car company that will be able to license the tech easily. Others are all working on it though. skyactiv-x is about 44% in the test mules, versus 40% in the prius engine, and 41% in the new camry hybrid engine. Skyactiv-3's goal is 56% thermal efficiency.

Mercedes on the other hand has gotten 50% efficiency out of their hybrid f1 gasoline engine, which doesn't even have to do the compression ignition trick. It uses a 1.6 liter engine in a psd configuration with an electric motor and a great use turbo/supercharing technology. It splits the turbo in half connected with a shaft and a 90 kw motor/generator that can rotate at 100,000 rpm typical for these high power turbos. The motor can act like a supercharger, adding boost to the engine quickly making it more responsive, but the motor also can also use the power of the exhaust and put it back in the battery. Mercedes is trying this in a very expensive 1000 hp production phev. I'll use toyota hsd terms for the motors. In both the f1 car and the production car mg1 is attached through a psd to the engine, but the output shaft goes to a double clutch transmission in the f1, and a traditional 8 speed manumatic in the production car to the rear wheels. The production car will split mg2 into two motors, one operating on the right front wheel, the other on the left front wheel. Because it is such a high power car its got a big battery, a limh that can go 25 km on the nedc cycle, my guess is that is around 12 miles aer on the epa cycle. F1 version gets 50% thermal efficiency, they are talking about 41% in the production phev engine as it needs nvh and low power availability compatible with a street car.

What does this mean to less expensive hybrids and phevs 6 years from now? Well bosche is making a common low price part to add to a traditional turbo charger to mimic the control of mercedes f1 engine. Audi, delphi, and others have been working on making the tech inexpensive.
BorgWarner's e-Booster Could Make Turbocharged Engines 10 Percent More Efficient | WIRED
This won't provide the huge boost of the f1 system, and won't charge the battery much, but mercedes plans to put it in a v6 turbo that will be much more efficient than a v8 normally aspirated engine of similar power. This tech applied to skyactiv x versus the mechanical supercharger might boost it up to over 50% efficiency when other parameters are tweaked as all the power for the supercharger now comes from the engine, while this would mainly come from wasted exhaust energy.

Put this a 2L skyactiv-x with a echarger type system and a 48V bas system, and likely you could build a 250 hp camry that cost about the same as the hybrid, and about the same combined mpg as the hybrid (worse city, better highway and hypermiling). Put a similar 1.2L - 3 cylinder engine in a phev like the prime, with a change in the psd and motors for the higher torque lower rpm range of power of the engine, and acceleration would be greatly improved, while battery depleted gas usage may be 20% lower for around $1500 more for the engine, but battery prices may drop more than that. I expect this technology to be available in a car within the next 7 years from either mazda or toyota.

 

The case of this electric/mechanical super/turbo charging technology they require a lot more power than the 12V advanced lead acid batteries provide in start stop systems, but only need it for short bursts of time. Currently many of these are after market devices that run off super-capacitors and many of these run at 48V. The use of this technology is what will boost the efficiency potential of 48V start stop systems.

From a thermodynamic point of view a miller cycle engine turbocharged engine is much more efficient than an atkinson or otto engine. Unfortunately without adding supercharging there is either turbo lag and/or sizing of the engine and turbo that is not optimal for efficiency. The mechanical supercharger in the first skyactiv-x is inexpensive but it does not power itself from wasted heat, but from engine rotation. Replacing the mechanical supercharger with an electric one plus a small right sized turbo charger would probably allow the engine to be downsized while making it more efficient. Since this would add at least a 48V super-cap, why not add a start-stop bas system and a 48V lithium battery with more energy but enough power to spin the supercharger? This is what mazda has hinted they will do, but they also look like they will use it in phev systems.

Audi and mercedes are attacking it from opposite ends. Audi is doing this with inexpensive electric superchargers and 48V bas systems to add better power delivery to turbo 6 cylinder engines. You won't see much efficiency gain there, what they are going for is a more responsive system with a tiny efficiency gain, but can meet euro and US emissions standards. Mercedes is throwing money at it to really make it more efficient, but this will be very expensive tech, that they hope will fall in price. Their system does need big full hybrid batteries because their supercharger needs lots of power, and given the big battery are starting at a full hybrid high voltage system. The turbo/supercharger is broken to two halves with a shaft and motor in between and no waste gate. Braking it into 2 halves means that exhaust heat does not heat the intake air, but this makes it more expensive to produce. A motor on the turbo shaft is also more expensive than an electric supercharger, because the motor must be built to operate hotter and must spin at higher speeds (10,000 rpm in this case). But by being on the turbo shaft it can use the power of the exhaust gasses to charge the battery, instead of simply dumping it out the waste gate when more energy is comes from exhaust then is needed to compress intake air.

A 48V system with a mechanical turbo + electric supercharger will all things being equal beat a hybrid on the fuel efficiency on the highway. The mazda system also will because of lean burn do well in city cycles but not as well as a full hybrid. The main problem with the mercedes system is cost, and it is a full phev + efficient boost, but my guess is costs will go down.

 

For bob, I think the first real test of whether 48V hybrid/electric supercharger works is the 2019 CLS, which is replacing a thirsty beast of a v8.

From an engineering point of view the change is a very elegant way to build IMA style hybrid without adding much to the cost versus a non-hybrid version of the same car. Mercedes has replaced the flywheel, alternator, and starter, with a 15 KW motor/generator sandwiched between the engine and transmission. The mg can move the car with the engine off for short distances, which I guess means its a full hybrid. Mercedes increased efficiency by using an electronic water pump and air conditioning. They are replaced a 4.7L sequencial turbo V8 making 402 hp/443 lb-ft, with a 3L I6 electric turbocharged car making 367 hp/364 lb-ft hybrid with the electric motor supplying 21hp/181 lb-ft of torque at lower rpms as needed, but the mg on the motor shaft doesn't add to the peak horsepower or torque, while the motor on the supercharger does. Combined this will probably make the car more responsive and smoother than the V8 but will lack as much power at higher rpm.

The competition for the CLS is the bmw 6 series and the audi A7. Both sell mainly with their lower powered 6 cylinder engines. The 2019 audi A7 will have a 48V Bas start stop system, which will give a good comparison of IMA+electric supercharging versus just BAS. Really though, for an efficient car you would use a much smaller 3 or 4 cylinder with same size mg for the IMA. For comparison sake Lexus does not make a competitive model, but the CLS is based off the E series, and Lexus does have a poor selling competitor to that in the GSh.

Compare Side-by-Side



My estimate is the CLS will jump from 21 mpg to 26 mpg with these changes (about 2 mpg better than the lower powered less refined discontinued old V6). The A7 will go from 23 to 25. For perspective the full hybrid GSh gets 31 mpg, but the hybrid system adds 385 lbs and over $10,000 to the price of a GS.

 

I would size those cables and battery to supply bursts of at least 600 Amps. It may need to supply the traction motor, supercharger, and air conditioning compressor, and other loads, but that will be for less than 15 seconds. Pure speculation is that when in electric mode for any longer period of time, it won't use more than 10 KW.

For reference the prius c is 144V and a max motor power from battery of 19 KW, or 132A. For the same power a 48V system will have 9x the power loss in cables as a 144V system, if resistance is the same. This is not too bad. IMHO 4/0 cables are overkill but, they are only 0.80" in diameter and cost less than $7/foot. I'm sure the engineers at mercedes have figured it out. Even if it requires thicker cables, there is a cost savings because more expensive parts are regulated in at 60V.

 

Some older developments. Mazda, subaru, and vw group have been talking about 4 cylinder deactivation. VW has it running in some non-US vehicles. 2018 Mazda CX-5 Adds Cylinder Deactivation - Motor Trend
Mazda is the first to sell it in the US, which means now we get epa results on the cx-5. It improves 1 mpg in city and 1 mpg on highway on the awd version, but highway in the fwd remains the same. Combined stays the same, which is probably due to rounding, but it doesn't raise it by 1 mpg combined for either model.

Software may help and tula technologies software is being used by delphi on a better deactivation model where individual cylinders are deactivated, which gives better NVH and fuel economy.

http://wardsauto.com/engines/delphi-touts-48vdsf-system-inexpensive-diesel-alternative
I really like the idea, and it needs the advanced software, which costs a lot to develop, but once it works can be rolled out to large numbers of vehicles. The article also gives some costs and efficiency gains.

Taken all the way though, camless systems will allow the best variability of intake valving. There is now a car using a camless system that is close to being sold in china. If it works I expect rapid adoption of the technology.
Camless engine evaluation nearly complete in China - SAE International

camless also allows a 4 stroke engine to briefly act as a 2 stroke, which means if it works, it may get widely adopted for motorcycles especially in countries where emissions regulations are not as strict as epa.

Numbers from delphi on 48V and individual rapid cylinder deactivation.

Comparison to toyota hybrid is probably the 2018 camry le hybrid which is $3800 more expensive and provides 62% better fuel economy. In the Lexus GSh though things get much worse, over $10,000 more expensive and 29% mpg improvement. That makes it seem that in higher powered vehicles this may be most cost effective for improving fuel economy. The 2019 GM Silverado will use this cylinder deactivation software to vary a V8 to any combination of 2 to 8 cylinders. We should see how it does soon, but more improvement would be on a 6 cylinder with an electronic turbo charger.

 

I answered on the other thread, but the key is to not make all things equal ;-) I've heard terms like 10x on the tests thrown around, and it really depends on what levels are bad. The particulate regulations are getting tougher in europe and the US, not easier. Most port injected gasoline cars are very low in particulates compared to carburated or diesel vehicles.

The 2 primary technology I was talking about were skyaciv-x and -3, along with mercedes M256 along with the better mercedes tech from their f1 that will make it to a car soon . The skyactiv design should lend itself to cooler burns and lower particulates and NOx before treatment, as they have in cylinder sensors and precise injectors to make a homogeneous mixtures. The mercedes, well sure it will probably have higher NOx than a normally aspirated port injected engine of the same power, but these are trade offs to efficiency. Mercedes is using this instead of a bi-turbo v6 or v8, and with new regulations I'm sure its built to be lower not only in CO2, but these pollutants. The engine not only has an efficient mechanical turbocharger, but also an integrated 15 kw mg that does duty as a ima type hybrid as well as starter and alternator, and a 5 kw electrical supercharger. Say there is 7 psi of boost on the engine and valves close at 7.5 compression. This is equivallent to a 11.1 compression (but cooler because of the intercooler) with 10.5 expansion in a miller cycle. If the first 1.5 of compression is done by the turbo, then this is powered from the exhaust gas ;-), it will be more efficient and more powerful than an atkinson. Its a long stroke engine, meaning there should be plenty of time at lower rpms to mix the gasoline. Boil it to 5000 rpm (prius never goes here) or use high effective compression, and you probably get higher particulates. Mercedes is also working with gasoline particulate filters for higher boost models. The design makes higher rpm less efficient and higher polluting, but there is an eco button, that hopefully locks it out.

The audi SQ7 and lexus/toyota new 3.5 L twin turbo V6 were designed much before election results were known. I expect them to release more unheathy pollution than these engines. I do see the lobbiest from major autos working the politicians, but that is every administration. The switch to boost has been happening in europe for a long time. The diesel cheating scandal is forcing better testing, which means companies need boosted smaller engines to meet fuel economy and ghg goals. IMHO this is progress. 2-mode hybrid and diesels on lighter urban vehicles were backward steps.

 

I'm not sure if we see such different realities, I'm more looking at some of the facts to improve light hybrid and phev fuel economy and pollution. I've added the insight, gen 1 and 3 prius, and recent ima lithium civic hybrid. Mercedes M256 uses a similar IMA as the insight in a much more modern engine. All individual numbers need to be taken with a grain of salt, as they are self reported and self selected.

Compare Side-by-Side

First generation insight clearly beat the gen I prius in city fuel economy and user reported mpg. It took toyota until gen II prius to tie it in epa test for city. The mercedes M256 is a better IMA system than the gen I insight. It has a 15 KW Integrated mg, versus the insight's 10 KW, and 1 kwh lithium ion battery which is about the same size in energy as the original insights, but more of than energy is usable, in a more reliable higher power package. M256 also is beltless with electronic power steering, water pump, and ac, these were innovations that helped gen III prius become more efficient. Honda IMA and hsd both do the partial power thing well, the difference is the electric motor in a hsd can power the car with engine rpm at 0, in IMA it must rotate the engine, even if there is no fuel or pumping losses, but this is minor. On the highway, a more efficient engine and transmission will beat hsd.

I agree with you on the whole car concept. Rolling resistance and aerodynamics play a big role. IMHO. 48 V BAS systems likely won't do all that well in the city, because the automaker hasn't done the rest of the work. The key is that the 48V bas start stop systems will drive down costs to allow some of the better technology to be adopted. If a 48V IMA system with all the improvements are there in the 2018 camry hybrid le, plus better responses, well I think that will be good enough. Regen braking really is about 26 kw max in the prii (correct me if I'm wrong), if a 48 V system can do 15 kw, and you are gentle on the brakes its really not much difference. Say you get a 3% power penalty charging the battery because of the higher current, I don't think that will show much in the real world unless you are jack rabbiting and hitting the brakes just perfect for a prius but not the car you are driving.

There are definite improvements to the M256 IMA concept, but those are mainly that it is way too powerful to compare to a prius or even a camry V6. In its american CLS guise of 367 hp, it should be quite efficient and low polluting, but its in a expensive fairly high rolling resistance car. The AMG 53 version greatly increases boost to the detriment of some efficiency, and it is fitted with a gasoline particulate filter (much cheaper than diesel filters and no SCR needed for NOx). I am thinking more of 3-4 cylinder 1.2L-2L engines of this type. Add in some of those skyactiv 3 improvements spark controlled compression ignition, cylinder deactivation, aggressive cooled egr, and a 1.5L would
beat a camry LE hybrid in terms of responsiveness and efficiency and it would be much easier to hypermile.