I've always seen rotary engines as one of the best examples of a particular approach to design that solves something very elegantly but relies too much on one critical component where all the pressure of the design is applied. In the case of a rotary engine, it's the apex seals.
They're a great warning for designers/architects/engineers to not get too enamored with the elegance of a system if parts of it are not yet completely solved. It's so easy when designing to try to shove aside some complex problem and say you'll solve it later, or play some shell game where every time you hit some hard to solve problem you wind up shuffling it around to someplace else [1], but that kind of instinct ultimately leads to unworkable things in practice.
[1]: 'we'll solve the seals problem later.. maybe materials has an answer' or 'just add oil in the mix to protect the seals there, we'll solve emissions later'
I would just add that sometimes ideas exceed the technology of their time. So revisiting a design that had deficiencies (weak points, high production costs, bad emissions, etc.) with new tools, materials, etc. can lead to breakthroughs. Not that that's what is happening here, just why some ideas that previously didn't work seem to circle back around.
I feel like software is the poster child for this. My hunch is that a lot of techniques have been dropped for performance penalties that may be on the order of 20-30% (totally guessing here) when Moore’s Law often has covered that gap in mere months.
I read a comment on HN a while back that you can look like a genius in software by going back about 10 years and finding something forgotten. The whole web development shift from server-side to client-side and now drifting back to server-side as if it’s something new seem to validate this. Though in this case it just seems like the extension of the decades long back and forth of going from mainframe with super minimal client, to PC only, to networked, to client/server, to server-side web, etc, etc.
No, rotary engines are terrible. For one, the equivalent of variable valve timing doesn't exist/cannot exist for rotary engines, unless Mazda creates a miracle. VVT and VVL lets you do some insane things in terms of efficiency and good engine behavior, so the rotary falls behind a lot.
I'm not saying rotaries are great, but two stroke piston engines can have variable port timing (typically called a "power valve"). I don't see why the same method couldn't be applied to a rotary.
I think it only turns to charge the battery, so probably it turns at it's most efficient speed at all times. Does variable valve timing help in that case?
That said, the article doesn't describe any benefit to using the rotary engine here either.
Mazda implemented "auxiliary ports" in the fb to Fc transition ~85, the 5th and 6th ports were vacuum actuated rotating sleeves that increased intake volume and duration; this was available before VTEC and Variocam.
Mazda has a super interesting way of doing this.
When paired with an electric motor/generator, they can speed up/slow down specific parts of the combustion cycle to essentially get vvt.
I think they can only really do this as a range extender due to the fact it makes power delivery a little bit weird.
In hub motors might be another automotive example of this. They were used by Ferdinand Porsche and invented in 1896. Ever since engineers keep trying them out as a simple elegant solution. Eventually they will break through I think, but there’s the same allure in the simplicity of it all.
Does it produce more low end torque verses a traditional gasoline combustion engine? If not then why is it better suited for electrical generation? Is it more efficient with less load?
The article mostly makes it sound like Mazda just loves the wankle and wants to find any possibly way to bring it back - even though it has “high” emissions… so coupling it with a hybrid electric motor makes it happen..
Wankels have tremendous power-to-weight and power-to-size ratios. Their main problem is reliability. The generally accepted solution to improve rotary engine reliability (oil injection) results in poor emissions. The wide, flat-ish combustion chamber doesn't help the emissions problem, either.
The Wankel is at its most efficient and its most reliable when operating at a constant RPM. Conveniently, the EV generator application demands a pretty flat RPM band. As a result, the engine doesn't need to lean as hard into those emissions-increasing compromises.
Thus, EVs allow the Wankel's benefits over a reciprocating-piston engine to be reaped without the same costs as before. In theory, at least. It remains to be seen if the benefits will outweigh the drawbacks. I'm glad they're at least going to give it a try.
Technically, the Wankel rotary engine design allows much higher power output than an I-4 or even a V-6 engine design due to no strains on the rods, and much less strains on the piston(s) as those literally blow up all the time in I4 engine designs in similar power output ranges.
I've seen many very reliable 3 or 4 rotary wankel swaps in RX7 FD3S that output far beyond anything imaginable any V8/V12 could produce. 2000hp is not a joke here, when being run on pure ethanol as fuel input. Apart from fuel injectors, clutch and gearbox, these engines run very stable and reliable.
There was an RX-8 Blue model being sold in Japan which was burning hydrogen directly, effectively producing water as output, which, in the prototype was being converted back to hydrogen via a fuel cell. And this was in 2004.
I wish there were more Wankel engines being used as "pocket generators", because they can reliably run on synthesized alcohol and hydrogen and be a potential generator replacement for all that Diesel based crap that's being used in rural areas.
Imagine a solar roof on your house that produces hydrogen with some fuel cells (which also produce heat for your home). This could be the optimum cycle for use in a decentralized home, as chemical energy storage has no loss compared to li-ion batteries that have a limited lifetime. The multiple use of hydrogen (e.g. a stove just needs to burn the gas) also makes it very low tech, and possibly much more reliable than a circuit based system where transformers might fail over time.
But of course, can't sell decentralized approaches via gas stations, so it will never take off...
Reliability is something easily discounted because the data to characterize it is much more difficult to capture than performance data. In most applications you can work around this with redundancies and diverse technologies, but no one makes a fault tolerant powertrain due to cost.
I don't think there's a good reason to keep pushing down deadend reliability paths. We should be responding to our hard earned decades of learning and be pressing advantages. Not every novel and viable solution ends up being an enhancement.
What happened to the inside out rotary that was supposed to be extremely compact and would enable cheaper PHEVs? IIRC the patents were about 5 years ago.
> The Wankel is at its most efficient and its most reliable when operating at a constant RPM. Conveniently, the EV generator application demands a pretty flat RPM band.
If I understand the article correctly, the "series" hybrid configuration means precisely that it cannot operate at the ideal rpm when charging the battery, because it is always driving the wheels directly as well.
> Does it produce more low end torque verses a traditional gasoline combustion engine? If not then why is it better suited for electrical generation? Is it more efficient with less load?
You've got it backwards.
The Prius's Atkinson engine makes low-end torque *worse*, and then relies upon the EV Motor to drive the car at low speeds (0mph to 10mph) before the ICE kicks back in.
If ICE is operating, its at higher RPMs where the generator can still be useful (low RPMs like 500 are too low for the Atkinson engine to be effective in any way, the computer instead increases the RPM to maybe 2000, and uses all the power to drive a generator instead)
------------
So you see, the name of the game is efficiency at all costs, with EV-motors assisting whatever compromise you built into the motor. In the case of Toyota, its absolutely undrivable crap for low-end torque ICE, but a powerful enough 60hp to 100hp electric-motor that can handle the low-speeds and stop-and-go traffic, smoothing out any problems.
------
IE: The engineers build a highly compromised ICE engine (the Atkinson engine) that has a far narrower band of usable RPMs than a normal vehicle. Then they smooth out those problems with electric motors.
It sounds like Mazda is doing the same trick here with their Rotary engine, but the Rotary engine doesn't have the crazy-good efficiency curves that the Toyota Atkinson engine has. Efficiency isn't the "only" name of the game however, but Mazda now needs to find out a good way to market this engine / highlight its strengths.
Isn't it a Miller cycle engine, not Atkinson? Miller cycle exploits variable valve timing to make the compression stroke effectively shorter than the expansion stroke (but reduces the amount of air being compressed, reducing power); Atkinson has some funky extra joints in the rod or crankshaft, I think.
Wankels can be made extremely compact so that might have something to do with it, i.e it has both very high power to weight and power to volume specs. I honestly don't know if that is the reason though, perhaps someone more knowledgeable of the specifics of range extenders might chime in but I imagine that is an important factor.
They have a pretty narrow power band (produce efficient max power at a small range of RPMs) which always somewhat limited their use for normal ICE cars, but is a pretty workable constraint for an electrical generator.
So you have compact size, good power to weight ratio, and power limitations that don't really matter for range extender purposes. Lots of potential.
I can't find the article right now, but I'd swear I remember discussion here a handful of months ago about a startup marketing a similar EV range extender engine design.
Also for those who don't know, it's pronounced /ˈvaŋkl̩/ or /ˈvaŋkəl̩/ in the International Phonetic Alphabet (IPA). In German, "W" is pronounced as "V" in English, and the "e" in the syllable "kel" is reduced to a schwa ([ə]) sound, common in many German pronunciations of unstressed vowels.
In a hybrid vehicle, the gas engine can be run mostly as a generator, which makes it possible to further optimize it for a very specific load. It's possible that such an updated+tuned Wankel could be a great fit for certain applications where space and weight are at a premium.
They can simulate lots of it but to get real answers, they have to build the engine and see how it holds up.
Also Mazda is a small-ish manufacturer at the Japanese scale. Since Wankels are part of their identity they could decide to build a car with it even though the downsides wouldn't make sense for a "rational" brand like Toyota. It can give them that creative freedom that help make desirable cars and keep Mazda relevant.
Low end torque is not that relevant to electrical generation, which typically involves the motor constantly running at a constant rpm. So, since low end torque is a weakness of the Wankel engine, that actually makes it more suitable for electrical generation than for driving directly.
First, like most of the Japanese manufacturers, Mazda bet against electric vehicles. They focused R&D on improving engine efficiency and getting their engines to run on hydrogen. If Mazda wants to make electric vehicles now, they have to play catch-up, or license key technologies from other manufacturers.
Second, batteries are heavy. For sedans and mid-size crossovers, this isn't much of a problem. EVs of that class are about the same weight as combustion vehicles. But for a lightweight sports car with decent range, batteries would be a big chunk of the total weight. Tesla's 85kWh battery weighs around 1,200lbs. If your desired weight is 2,500lbs, that only leaves 1,300lbs for the actual car. Yes you can save some weight by making the battery part of the structure, and you don't need an exhaust system, engine block, alternator, intake, etc, but it's still a tough set of constraints to work within.
Why do customers want sports cars to be light? Well all else equal, a lighter vehicle will have better performance. But even when all else isn't equal, vehicle weight can drastically affect driving enjoyment. I have a 4,048lb Model 3 Performance and a 2,182lb Mazda Miata. In terms of specs, the Model 3 is better in every way. It can accelerate, brake, and turn better than the Miata. It even has more range than the Miata. But the Model 3 feels like it's using brute force to beat inertia into submission. (Don't get me wrong, that can be fun.) The Miata is the opposite. Its light weight means that there's very little inertia to overcome, and something about that is extremely satisfying. It's almost like having a street legal go-kart. Until battery technology improves, an electric version just won't have the same appeal.
Engineering organizations fall in love with superficial attributes of solutions that worked especially well for them in the past. When RIM/BlackBerry realized the iPhone was a serious threat, they built a touchscreen phone where the entire display produces a physical clicking effect because they were so convinced that what people really want from a smartphone is the click of a keyboard.
Mazda is BlackBerry, and the rotary engine is their clicky keyboard.
Since the achilles heal of the rotary is wearing seals has anyone seriously invested in testing coating/material tech for this? It may not make sense for mass production but spending ~10k to make a bulletproof rx7 would be an incredibly good investment.
Wankels have only one advantage - they are about two times more powerful on same volume (you could consider them as very clean two stroke engines).
Unfortunately, Wankels have extremely huge mechanical problems - complex geometry (classic ICU are very close to just cylinders), need of better materials, depend on much better oil.
And also big problem is production scale, as I talked with people, they considered Wankels as toy, you will just utilize when it run out guarantee term.
For example, for standard ICU, considered big repair, sleeves, so they will continue working, sure, less heavy duty than new.
You do want efficient horsepower to drive a generator. More low end torque means the HP comes at lower RPM, which should mean less fuel consumption.
Sibling says the achilles heel of the Wankel rotary is low end torque, but you don't take the direct output from the engine, it goes through gear reduction for final drive output. The real achilles heel is the awful emissions. It's more or less a 2 cycle engine from that POV.
When driven at variable speeds it's hard to wrangle.
The reason it's well suited for electrical generation is its mechanical simplicity, compactness, low weight, low NVH, and not least important, "rotary" brand value. I suppose that when run at constant RPM and constant or smoothly changing load, the emissions is easier to deal with.
Reliability for rotaries hasn't been a concern for a long time. Modern apex seals work well and last a reasonably long time. There is a need to stop parroting facts from the 1980s.
One interesting detail is that Mazda never designed a wankel after the 90s. They have claimed since then that computer design and simulation has allowed for dramatic performance gains.
Low end torque tends to be the rotary's Achilles heel. I think the claim being made is that efficiency is better at high, steady RPMs, but tbh, I've always found that claim a bit dubious. If you love the rotary engine, this does have some nice perks as the electric motor basically fixes the rotary engine's main weakness.
Having said that, I'd have been much more excited about this 10-15 years ago.
It really isn't. Go and drive a 13BT car. They make almost all their torque by 2500rpm and the curve is flat. They're more torquey than any 4 cylinder I've driven.
I used to own the first gen electric i3 with its tiny range. I didn’t wish to have the hybrid ICE version to drive further, but I did wish for more electric range and more fast chargers along the road.
Nowadays, I have a cheaper car with a lot more range, almost 4 times more in real life conditions, and plenty of fast chargers everywhere. I don’t see why I would bother with an ICE. It makes no sense for me.
It’s because I live around Oslo in Norway, a place where it’s the age of electric cars.
I think ICE for cars has a very limited future in the age of electric cars. I see it reserved for specific applications where the energy density is a must, and some car enthusiasts activities.
Hybrids are some kind of temporary solutions for places where the EV infrastructure aren’t good enough yet. Once the infrastructure is good enough, some people will still buy ICE for a little while as they are unsure, but most switch to full electric eventually. At least that what happened around Oslo and happens now in the country side of Norway.
I am a big proponent of EVs, but I personally think hybrids will have a really long tail, especially as the technology improves: we could feasibly end up with a situation where much city driving in a hybrid is zero-emission EV-mode.
Two reasons:
First, in relation to your point
> Once the infrastructure is good enough
I live in the UK and I think this is gonna take a long, long time here. Not only will we need to build an enormous amount of fast chargers, but there will need to a significantly greater number of them than petrol stations, to offset the fact that even the fastest chargers take 5-10x as long as filling up with petrol (2-3 mins vs 20-30).
Of course, the ideal scenario with EVs is that most charging is done at home, with fast chargers used only on long journeys. Problem is, by some estimates 2/3 of UK households do not have off-street parking. We would need to roll out en-masse solutions for on-street charging and, to my knowledge, we have not even began to think about this outside limited trials.
Second is cost. Almost all cars have got crazy expensive over the last few years (I’m unsure if the rise of the PCP is a cause or effect of this), but over here full EVs are still not affordable for a huge number of people—myself included.
I really wanted to go electric, and was looking at the MG4; widely considered to be the best value in EVs in the UK right now. But for the model with a range that would suit us, and the cost of installing a charger, you’re looking at close to £30,000. I just don’t have that sort of money, and a finance deal would be half my mortgage again. And for context, I make nearly double the median UK salary.
If you think about cost, long term it’s cheaper to skip the ICE and have a slightly better electric power train with a bigger battery capacity. That’s what Carlos Tavares said to the French government many years ago. Hybrids have two systems and that’s expensive and complex.
Installing electric power plugs everywhere isn’t that challenging. I’m sure UK can manage it if the local government decides to do it seriously.
By the way, 35k€ for a brand new car that is cheap to run isn’t a bad deal. New cars are expensive. Maybe you can wait a decade or two before switching with a cheap used EV that has enough range.
> even the fastest chargers take 5-10x as long as filling up with petrol (2-3 mins vs 20-30)
I just drove to the airport in a bit of a rush.
Drove straight up to pump, tapped credit card, authorized for $100 and opened cap at same time. Stuffed the pump in and held it at max flow. 45l of gas later (small car), cap back on, pump away and moving again.
Is this intended to be an anecdote about life in Oslo, or are you suggesting that you expect the world outside of Scandinavia will somehow become more like Scandinavia over time?
Yes, with the right political decisions most countries could switch to electric vehicles. EVs also are nicer to drive and cheaper over the vehicle lifetime. It’s just a matter of time in my opinion.
I have never understood why the Volt Series Hybrid idea never took off. It is more efficient to turn gasoline into electricity and then drive the car with that than to directly connect the engine to the wheels. Is it perhaps that the cost involved is just too much more than a plugin hybrid to make the small extra fuel savings worth it?
This was how the BMW i3 worked. It was a rather novel design that included an optional small electric scooter motor in the rear that had a 2.5 gallon gas tank. When the battery was low, it would be charged by running the small generator.
This was clearly a wonderful idea but it was hamstrung by a silly California rule requiring the gas range to be less than the electric range to qualify for rebates. With a 6 gallon tank, the car would have been able to do ~300 miles instead of 170 and would have been parked in everyone’s driveway.
An added benefit was that the car could use existing gas station infrastructure when you needed to travel long distances.
> it was hamstrung by a silly California rule requiring the gas range to be less than the electric range to qualify for rebates.
The problem is that BMW wanted to get the same amount of credits as a pure-BEV. California anticipated that a car with 300 miles of gas range might spend much of it's life being driven on gas if there was no penalty for doing so... and in fact that is exactly what happened with many European plug-in hybrids sold as company cars.
California never gave BMW the credits, but BMW decided to keep the dinky gas tank anyway, so that's on them.
Some friends of mine in the Seattle area had one and absolutely loved it. It was a very practical car for commuting within its range. The extender was barely audible when running.
I think the goofy styling was probably as much of an issue as the tank size.
> It is more efficient to turn gasoline into electricity and then drive the car with that than to directly connect the engine to the wheels
I thought so too, but my research suggested that the efficiency is pretty much the same if not worse and power delivery is worse. Do you have some links? I‘d like to be wrong on this one.
Surely the "generator -> charger -> battery -> inverter -> motor" chain is less efficient than a driveshaft. Perhaps the only benefit is the engine can run at an optimal speed, but an appropriate gear ratio should handle that.
Supposedly has rooms for improvements but there is actually that system and real world data: Nissan NOTE e-POWER with LEAF motor and no mechanical engine->wheels connection has actual user reported mileage of 19.72 km/L(46.4 mpg) on regular fuel[0]. This is worse than reported 21.25 km/L(50 mpg)[1] for Prius[2].
> but an appropriate gear ratio should handle that.
See Prius's "Powersplit Device".
I'd describe the Powersplit device to be a combination of generator/alternator, starter/electric motor, reversed-differential (2x power inputs -> 1x driveshaft), and effective gear-ratio. All in one planetary gearset.
EV motor2 determines the speed of the car.
The ICE motor can spin at any speed that the computer determines to be useful. If EV Motor2 is 0-rpm, then the ICE is 100% in generator mode (2000rpm but the car isn't moving: all the energy goes to charging the battery). If the EV motor is at 10mph but ICE is off (0-rpm), then its 100% electric drive mode. And any combination in-between is possible.
EV Motor1 (a smaller, weaker motor) controls the 3rd set of gears (I think the planet gears?? I forget), which determines how ICE relates to EV (changes the effective gear-ratio)
So yeah, the PSD allows the ICE to always function at the appropriate speed (which is either 0rpm or ~2500rpm for efficiency). While the combination of EV-motor1 (changes effective gear ratio of ICE) and EV-motor2 (hard-wired to the final speed) handle the different speeds the user wants in practice.
I do think that the Prius (and Prius Prime) have surpassed the Volt's design, and the proof is in the pudding. Prius Prime has 52mpg, a figure far more efficient than the Volt ever had.
Prius Prime also has 220 horses today for a 0-to-60 time of 6.6 seconds. So today's Prius Prime is a lot faster than the Volt too.
Volt was good when it came out, but technology has gotten better since then. Toyota has seemingly perfected this "power split device", and its beginning to lead into exceptional acceleration and good driving feel (as opposed to being 100% economy focused like before). Volt had better feel than the 2010-era Prius, but 2024-Prius is a totally different car.
I think all Volt fans are in "but what if GM didn't kill the Volt and kept investing in the technology?". And... yeah... that's a fun what-if. But... GM killed the Volt. It sucks, because it seemed like great tech. Apparently GM has kept the drivetrain technology ("Ultium") and has continued to provide R&D, but Toyota's recent advancements are jawdroppingly good.
The Volt worked best as an EV with an ICE range extender. As an EV, the 40-52 mile range was sufficient for 90+% of daily driving. Adding more range would have little true benefit. It was a series hybrid, but it was mediocre as a hybrid due to the added battery weight.
Nissan has the E-power hybrid that is the pure series hybrid that you describe. AFAIK it is not as efficient as a regular, parallel hybrid. The advantage is in cost as running the gas engine as a generator uses fewer components than running it in a parallel hybrid system.
I got my honda partially because it has the closest thing to nissan's e-power (in many ways the same, but with a couple twists).
At low speeds (< 45 mph) it's either off the battery the whole time, or the motor runs and directs power to the traction motor or the battery.
One wrinkle is that the motor assembly has a clutch that can engage the engine direct to the wheels, but like in city driving, the engine isn't going to be connected to the wheels. But I def would say that at least Honda's hybrid doesn't feel as simple as e-power.
Now this probably isn't really accurate, but architecturally I think I like it because the only thing keeping it from being a BEV is the battery size (only a couple kWh); the electric motor is strong enough to be usable on its own.
I think Toyota has increased the power on the normal prius, and definitely on the prime, but it used to be as part of the power split system, the electric drive motor wasn't sized to be enough on its own.
That seems unlikely. A mechanical coupling should be close to near perfect efficiency where as using the engine to drive electric motors requires several conversion steps.
The trick is having a large buffer, so that typically the engine can operate at the point with the best specific fuel consumption or shut down entirely. Getting energy from the engine to the wheels is less efficient than a mechanical transmission, but the increased average efficiency of the engine can more than offset that.
Aside from your suggestion violating the laws of physics, that's actually not how the Volt worked.
The original concept for the volt was that the engine would only generated electricity, but in production models, the engine was connected to the drivetrain.
Almost - In production models the engine does act like a series hybrid generator under most circumstances. At high speeds, the engine was directly connected to the drive train as it is more efficient to do that. the transaxle of the Voltec system is a marvel of engineering. It supports a dynamic switching between series and parallel hybrid mode as well as using the two separate electric motors for either power delivery or can switch one to regen and can rapidly switch between those modes, too.
It wasn’t my suggestion, but what I read way back in the day, and it may not be accurate but it does not violate the laws of physics. It’s entirely possible that (just as one hypothetical example) being able to keep the engine at optimal RPMs at all times in a series hybrid creates more efficiency even after the extra conversion losses.
Exactly. The Volt uses the same planetary gear set style transmission that the Prius does.
Everyone bemoaning the death of the Volt can now just buy a Prius Prime (the PHEV variant). It's the same thing just newer/better. It even looks sporty-ish now.
That remains true for more recent "series hybrids" as well, such as the Honda mentioned in the article. The efficiency gain from engaging the ICE when cruising on the highway is just too good to pass up.
“It is more efficient to turn gas into electricity”
I don’t think that’s correct. It may be true for highly variable/low loads where the pumping losses in the pistons dominates. However the majority of fuel consumption in a car happens at traveling speed (highway miles). That is the area that needs to be optimized for.
My 2015 Honda Accord Hybrid takes this approach. At below-freeway speeds the gas motor runs in series to drive an electric motor. At highway speeds, it engages a clutch and directly connects the engine to a low-loss 1-speed transmission.
> However the majority of fuel consumption in a car happens at traveling speed (highway miles).
Unless of course you drive primarily in stop-and-go traffic, e.g. delivery drivers, taxis, commuter cars, etc. Quite often you won't exceed 50 kph. For whatever reason, I've never seen (or have and forgotten) a car marketed towards this market—probably for exactly the reason that plug-in hybrids perform better in this scenario.
> > “It is more efficient to turn gas into electricity”
> I don’t think that’s correct.
You're right, it's not.
In fact, GM wrote an SAE paper about their "2-mode hybrid" transmission (which was used in their 2008-2013 light-duty trucks and SUVs, and then in later modified form in the 2nd gen Volt), where this is plainly explained.
In the paper they describe exactly the tradeoffs made to optimize fuel efficiency in an eCVT... it turns out that you want to set up the planetary gears to minimize the energy transmitted from input to output via electricity and maximize the amount transmitted mechanically because that is most efficient. You especially want to avoid a round trip through the battery in most cases (except when that allows installing a smaller, more efficient ICE).
That has implications for the CVT's mechanical ratio: GM's "2 mode" which has what basically amounts to an auxiliary overdrive integrated in the eCVT so that it can use smaller motor-generators over a wider range of speeds. Smaller motor-generators means more energy is transmitted mechanically, which means higher efficiency.
This is also basically the same reason the 1st gen Volt gets significantly worse highway MPG in range-extender mode than the (contemporary) Prius Hybrid (~35MPG vs ~50MPG): The 1st gen Volt eCVT was envisioned as an EV with a range extender (where energy usually comes from a battery), while the Prius's eCVT was optimized for driving primarily on the ICE, with the battery only supplementing acceleration.
Stellantis (Dodge) has an upcoming version of the RAM 1500 that does this. It's an electric truck with a 145 mile range on electric battery; once that is expended the gas engine kicks on to run a 130 kilowatt generator.
It's worth mentioning that serial hybrids, like the BMW i3 rx, might not be able to drive on the freeway on the output of their generator. That's why it's a "range extender" - at some point you have to pull over and charge.
EDIT: this is a big secret that none of the marketing materials (want to) make clear.
I'm uncertain the output of the wankel, but maybe with its power-to-weight it might get closer to being able to drive on gasoline in a self-sufficient way.
The problem with the Wankel is and has always been apex seals. You need to rebuild it at least once during a vehicle lifetime. I would not say it is safe from this fate, even with modern seals and a hybrid application.
I thought it was the horrible compression and oil burning. Apex seals seem to last if you don’t do silly things to them but you can’t solve inherent issues with the cycle design.
For comparison, rotaries get about 100psi of compression, a modern gas car can more than double that. More compression, more efficient burn.
I've read that oil burning (which is directly tied to the apex seals) has been fixed in modern engines. Compression is a somewhat configurable matter, and can be high in a low-torque generator (battery hybrid) application.
Readit News
They're a great warning for designers/architects/engineers to not get too enamored with the elegance of a system if parts of it are not yet completely solved. It's so easy when designing to try to shove aside some complex problem and say you'll solve it later, or play some shell game where every time you hit some hard to solve problem you wind up shuffling it around to someplace else [1], but that kind of instinct ultimately leads to unworkable things in practice.
[1]: 'we'll solve the seals problem later.. maybe materials has an answer' or 'just add oil in the mix to protect the seals there, we'll solve emissions later'
I read a comment on HN a while back that you can look like a genius in software by going back about 10 years and finding something forgotten. The whole web development shift from server-side to client-side and now drifting back to server-side as if it’s something new seem to validate this. Though in this case it just seems like the extension of the decades long back and forth of going from mainframe with super minimal client, to PC only, to networked, to client/server, to server-side web, etc, etc.
Kelly Johnson of Skunkworks had a keen mind for differentiating between them.
The rotor is pretty much unable to be cooled too.
That said, the article doesn't describe any benefit to using the rotary engine here either.
I wonder if we're missing something?
https://carbuzz.com/news/mazdas-new-rotary-engine-variable-v...
Which, now that I think of it again, is a bit silly because most EV mass is battery, not motor.
The article mostly makes it sound like Mazda just loves the wankle and wants to find any possibly way to bring it back - even though it has “high” emissions… so coupling it with a hybrid electric motor makes it happen..
That can’t be the whole story?
Wankels have tremendous power-to-weight and power-to-size ratios. Their main problem is reliability. The generally accepted solution to improve rotary engine reliability (oil injection) results in poor emissions. The wide, flat-ish combustion chamber doesn't help the emissions problem, either.
The Wankel is at its most efficient and its most reliable when operating at a constant RPM. Conveniently, the EV generator application demands a pretty flat RPM band. As a result, the engine doesn't need to lean as hard into those emissions-increasing compromises.
Thus, EVs allow the Wankel's benefits over a reciprocating-piston engine to be reaped without the same costs as before. In theory, at least. It remains to be seen if the benefits will outweigh the drawbacks. I'm glad they're at least going to give it a try.
I've seen many very reliable 3 or 4 rotary wankel swaps in RX7 FD3S that output far beyond anything imaginable any V8/V12 could produce. 2000hp is not a joke here, when being run on pure ethanol as fuel input. Apart from fuel injectors, clutch and gearbox, these engines run very stable and reliable.
There was an RX-8 Blue model being sold in Japan which was burning hydrogen directly, effectively producing water as output, which, in the prototype was being converted back to hydrogen via a fuel cell. And this was in 2004.
I wish there were more Wankel engines being used as "pocket generators", because they can reliably run on synthesized alcohol and hydrogen and be a potential generator replacement for all that Diesel based crap that's being used in rural areas.
Imagine a solar roof on your house that produces hydrogen with some fuel cells (which also produce heat for your home). This could be the optimum cycle for use in a decentralized home, as chemical energy storage has no loss compared to li-ion batteries that have a limited lifetime. The multiple use of hydrogen (e.g. a stove just needs to burn the gas) also makes it very low tech, and possibly much more reliable than a circuit based system where transformers might fail over time.
But of course, can't sell decentralized approaches via gas stations, so it will never take off...
I don't think there's a good reason to keep pushing down deadend reliability paths. We should be responding to our hard earned decades of learning and be pressing advantages. Not every novel and viable solution ends up being an enhancement.
If I understand the article correctly, the "series" hybrid configuration means precisely that it cannot operate at the ideal rpm when charging the battery, because it is always driving the wheels directly as well.
You've got it backwards.
The Prius's Atkinson engine makes low-end torque *worse*, and then relies upon the EV Motor to drive the car at low speeds (0mph to 10mph) before the ICE kicks back in.
If ICE is operating, its at higher RPMs where the generator can still be useful (low RPMs like 500 are too low for the Atkinson engine to be effective in any way, the computer instead increases the RPM to maybe 2000, and uses all the power to drive a generator instead)
------------
So you see, the name of the game is efficiency at all costs, with EV-motors assisting whatever compromise you built into the motor. In the case of Toyota, its absolutely undrivable crap for low-end torque ICE, but a powerful enough 60hp to 100hp electric-motor that can handle the low-speeds and stop-and-go traffic, smoothing out any problems.
------
IE: The engineers build a highly compromised ICE engine (the Atkinson engine) that has a far narrower band of usable RPMs than a normal vehicle. Then they smooth out those problems with electric motors.
It sounds like Mazda is doing the same trick here with their Rotary engine, but the Rotary engine doesn't have the crazy-good efficiency curves that the Toyota Atkinson engine has. Efficiency isn't the "only" name of the game however, but Mazda now needs to find out a good way to market this engine / highlight its strengths.
Yes: https://en.wikipedia.org/wiki/Atkinson_cycle
However, these modern Miller cycle engines are being called Atkinson or Atkinson-Miller cycle for some reason.
So you have compact size, good power to weight ratio, and power limitations that don't really matter for range extender purposes. Lots of potential.
I can't find the article right now, but I'd swear I remember discussion here a handful of months ago about a startup marketing a similar EV range extender engine design.
Also for those who don't know, it's pronounced /ˈvaŋkl̩/ or /ˈvaŋkəl̩/ in the International Phonetic Alphabet (IPA). In German, "W" is pronounced as "V" in English, and the "e" in the syllable "kel" is reduced to a schwa ([ə]) sound, common in many German pronunciations of unstressed vowels.
They can simulate lots of it but to get real answers, they have to build the engine and see how it holds up.
Also Mazda is a small-ish manufacturer at the Japanese scale. Since Wankels are part of their identity they could decide to build a car with it even though the downsides wouldn't make sense for a "rational" brand like Toyota. It can give them that creative freedom that help make desirable cars and keep Mazda relevant.
A true electric successor to the RX-7 would capture so much attention.
First, like most of the Japanese manufacturers, Mazda bet against electric vehicles. They focused R&D on improving engine efficiency and getting their engines to run on hydrogen. If Mazda wants to make electric vehicles now, they have to play catch-up, or license key technologies from other manufacturers.
Second, batteries are heavy. For sedans and mid-size crossovers, this isn't much of a problem. EVs of that class are about the same weight as combustion vehicles. But for a lightweight sports car with decent range, batteries would be a big chunk of the total weight. Tesla's 85kWh battery weighs around 1,200lbs. If your desired weight is 2,500lbs, that only leaves 1,300lbs for the actual car. Yes you can save some weight by making the battery part of the structure, and you don't need an exhaust system, engine block, alternator, intake, etc, but it's still a tough set of constraints to work within.
Why do customers want sports cars to be light? Well all else equal, a lighter vehicle will have better performance. But even when all else isn't equal, vehicle weight can drastically affect driving enjoyment. I have a 4,048lb Model 3 Performance and a 2,182lb Mazda Miata. In terms of specs, the Model 3 is better in every way. It can accelerate, brake, and turn better than the Miata. It even has more range than the Miata. But the Model 3 feels like it's using brute force to beat inertia into submission. (Don't get me wrong, that can be fun.) The Miata is the opposite. Its light weight means that there's very little inertia to overcome, and something about that is extremely satisfying. It's almost like having a street legal go-kart. Until battery technology improves, an electric version just won't have the same appeal.
Not entirely sure that would be the case even with a Red Sun label on it.
Engineering organizations fall in love with superficial attributes of solutions that worked especially well for them in the past. When RIM/BlackBerry realized the iPhone was a serious threat, they built a touchscreen phone where the entire display produces a physical clicking effect because they were so convinced that what people really want from a smartphone is the click of a keyboard.
Mazda is BlackBerry, and the rotary engine is their clicky keyboard.
No, they're not. Go test drive a Miata. They make the most fun cars to come out of Japan. And also have one of the better design languages.
Unfortunately, Wankels have extremely huge mechanical problems - complex geometry (classic ICU are very close to just cylinders), need of better materials, depend on much better oil.
And also big problem is production scale, as I talked with people, they considered Wankels as toy, you will just utilize when it run out guarantee term.
For example, for standard ICU, considered big repair, sleeves, so they will continue working, sure, less heavy duty than new.
What does low end torque have to do with electrical generation?
You do want efficient horsepower to drive a generator. More low end torque means the HP comes at lower RPM, which should mean less fuel consumption.
Sibling says the achilles heel of the Wankel rotary is low end torque, but you don't take the direct output from the engine, it goes through gear reduction for final drive output. The real achilles heel is the awful emissions. It's more or less a 2 cycle engine from that POV.
When driven at variable speeds it's hard to wrangle.
The reason it's well suited for electrical generation is its mechanical simplicity, compactness, low weight, low NVH, and not least important, "rotary" brand value. I suppose that when run at constant RPM and constant or smoothly changing load, the emissions is easier to deal with.
There are better designs than Wankel for "series hybrid" application.
https://newatlas.com/automotive/innengine-one-stroke-engine-...
One (potential?) advantage not mentioned in the article is lighter weight relative to a similar traditional engine.
It does prefer a narrow RPM band, which is fine.
Reliability is the biggest concern TBH, but maybe that's not a huge bummer if its more of a backup/assistant engine.
Having said that, I'd have been much more excited about this 10-15 years ago.
As others have pointed out, the article doesn’t do a great job of explaining how the rotary helps.
Nowadays, I have a cheaper car with a lot more range, almost 4 times more in real life conditions, and plenty of fast chargers everywhere. I don’t see why I would bother with an ICE. It makes no sense for me.
It’s because I live around Oslo in Norway, a place where it’s the age of electric cars.
I think ICE for cars has a very limited future in the age of electric cars. I see it reserved for specific applications where the energy density is a must, and some car enthusiasts activities.
Hybrids are some kind of temporary solutions for places where the EV infrastructure aren’t good enough yet. Once the infrastructure is good enough, some people will still buy ICE for a little while as they are unsure, but most switch to full electric eventually. At least that what happened around Oslo and happens now in the country side of Norway.
Two reasons:
First, in relation to your point
> Once the infrastructure is good enough
I live in the UK and I think this is gonna take a long, long time here. Not only will we need to build an enormous amount of fast chargers, but there will need to a significantly greater number of them than petrol stations, to offset the fact that even the fastest chargers take 5-10x as long as filling up with petrol (2-3 mins vs 20-30).
Of course, the ideal scenario with EVs is that most charging is done at home, with fast chargers used only on long journeys. Problem is, by some estimates 2/3 of UK households do not have off-street parking. We would need to roll out en-masse solutions for on-street charging and, to my knowledge, we have not even began to think about this outside limited trials.
Second is cost. Almost all cars have got crazy expensive over the last few years (I’m unsure if the rise of the PCP is a cause or effect of this), but over here full EVs are still not affordable for a huge number of people—myself included.
I really wanted to go electric, and was looking at the MG4; widely considered to be the best value in EVs in the UK right now. But for the model with a range that would suit us, and the cost of installing a charger, you’re looking at close to £30,000. I just don’t have that sort of money, and a finance deal would be half my mortgage again. And for context, I make nearly double the median UK salary.
Installing electric power plugs everywhere isn’t that challenging. I’m sure UK can manage it if the local government decides to do it seriously.
By the way, 35k€ for a brand new car that is cheap to run isn’t a bad deal. New cars are expensive. Maybe you can wait a decade or two before switching with a cheap used EV that has enough range.
I just drove to the airport in a bit of a rush. Drove straight up to pump, tapped credit card, authorized for $100 and opened cap at same time. Stuffed the pump in and held it at max flow. 45l of gas later (small car), cap back on, pump away and moving again.
I don’t think it would be possible to be faster.
Total time stopped: 7 minutes.
This was clearly a wonderful idea but it was hamstrung by a silly California rule requiring the gas range to be less than the electric range to qualify for rebates. With a 6 gallon tank, the car would have been able to do ~300 miles instead of 170 and would have been parked in everyone’s driveway.
An added benefit was that the car could use existing gas station infrastructure when you needed to travel long distances.
The problem is that BMW wanted to get the same amount of credits as a pure-BEV. California anticipated that a car with 300 miles of gas range might spend much of it's life being driven on gas if there was no penalty for doing so... and in fact that is exactly what happened with many European plug-in hybrids sold as company cars.
California never gave BMW the credits, but BMW decided to keep the dinky gas tank anyway, so that's on them.
Thankfully it’s trivial to change.
I think the goofy styling was probably as much of an issue as the tank size.
I thought so too, but my research suggested that the efficiency is pretty much the same if not worse and power delivery is worse. Do you have some links? I‘d like to be wrong on this one.
0: https://minkara.carview.co.jp/car/nissan/note_e_power/nenpi/
1: https://minkara.carview.co.jp/car/toyota/prius/nenpi/
2: The latest one has a worse fuel economy figure than the 1st gen, by the way
See Prius's "Powersplit Device".
I'd describe the Powersplit device to be a combination of generator/alternator, starter/electric motor, reversed-differential (2x power inputs -> 1x driveshaft), and effective gear-ratio. All in one planetary gearset.
EV motor2 determines the speed of the car.
The ICE motor can spin at any speed that the computer determines to be useful. If EV Motor2 is 0-rpm, then the ICE is 100% in generator mode (2000rpm but the car isn't moving: all the energy goes to charging the battery). If the EV motor is at 10mph but ICE is off (0-rpm), then its 100% electric drive mode. And any combination in-between is possible.
EV Motor1 (a smaller, weaker motor) controls the 3rd set of gears (I think the planet gears?? I forget), which determines how ICE relates to EV (changes the effective gear-ratio)
https://eahart.com/prius/psd/
---------
So yeah, the PSD allows the ICE to always function at the appropriate speed (which is either 0rpm or ~2500rpm for efficiency). While the combination of EV-motor1 (changes effective gear ratio of ICE) and EV-motor2 (hard-wired to the final speed) handle the different speeds the user wants in practice.
All in like, 15 gears or so.
https://www.youtube.com/watch?v=jofycaXByTc
---------
I do think that the Prius (and Prius Prime) have surpassed the Volt's design, and the proof is in the pudding. Prius Prime has 52mpg, a figure far more efficient than the Volt ever had.
Prius Prime also has 220 horses today for a 0-to-60 time of 6.6 seconds. So today's Prius Prime is a lot faster than the Volt too.
Volt was good when it came out, but technology has gotten better since then. Toyota has seemingly perfected this "power split device", and its beginning to lead into exceptional acceleration and good driving feel (as opposed to being 100% economy focused like before). Volt had better feel than the 2010-era Prius, but 2024-Prius is a totally different car.
I think all Volt fans are in "but what if GM didn't kill the Volt and kept investing in the technology?". And... yeah... that's a fun what-if. But... GM killed the Volt. It sucks, because it seemed like great tech. Apparently GM has kept the drivetrain technology ("Ultium") and has continued to provide R&D, but Toyota's recent advancements are jawdroppingly good.
Here are the differences explained in more detail:
http://roperld.com/science/ChevyVolt.htm
Nissan has the E-power hybrid that is the pure series hybrid that you describe. AFAIK it is not as efficient as a regular, parallel hybrid. The advantage is in cost as running the gas engine as a generator uses fewer components than running it in a parallel hybrid system.
https://www.nissan-global.com/EN/INNOVATION/TECHNOLOGY/ARCHI....
At low speeds (< 45 mph) it's either off the battery the whole time, or the motor runs and directs power to the traction motor or the battery. One wrinkle is that the motor assembly has a clutch that can engage the engine direct to the wheels, but like in city driving, the engine isn't going to be connected to the wheels. But I def would say that at least Honda's hybrid doesn't feel as simple as e-power.
Now this probably isn't really accurate, but architecturally I think I like it because the only thing keeping it from being a BEV is the battery size (only a couple kWh); the electric motor is strong enough to be usable on its own.
I think Toyota has increased the power on the normal prius, and definitely on the prime, but it used to be as part of the power split system, the electric drive motor wasn't sized to be enough on its own.
The original concept for the volt was that the engine would only generated electricity, but in production models, the engine was connected to the drivetrain.
Deleted Comment
Everyone bemoaning the death of the Volt can now just buy a Prius Prime (the PHEV variant). It's the same thing just newer/better. It even looks sporty-ish now.
I don’t think that’s correct. It may be true for highly variable/low loads where the pumping losses in the pistons dominates. However the majority of fuel consumption in a car happens at traveling speed (highway miles). That is the area that needs to be optimized for.
My 2015 Honda Accord Hybrid takes this approach. At below-freeway speeds the gas motor runs in series to drive an electric motor. At highway speeds, it engages a clutch and directly connects the engine to a low-loss 1-speed transmission.
Unless of course you drive primarily in stop-and-go traffic, e.g. delivery drivers, taxis, commuter cars, etc. Quite often you won't exceed 50 kph. For whatever reason, I've never seen (or have and forgotten) a car marketed towards this market—probably for exactly the reason that plug-in hybrids perform better in this scenario.
> I don’t think that’s correct.
You're right, it's not.
In fact, GM wrote an SAE paper about their "2-mode hybrid" transmission (which was used in their 2008-2013 light-duty trucks and SUVs, and then in later modified form in the 2nd gen Volt), where this is plainly explained.
In the paper they describe exactly the tradeoffs made to optimize fuel efficiency in an eCVT... it turns out that you want to set up the planetary gears to minimize the energy transmitted from input to output via electricity and maximize the amount transmitted mechanically because that is most efficient. You especially want to avoid a round trip through the battery in most cases (except when that allows installing a smaller, more efficient ICE).
That has implications for the CVT's mechanical ratio: GM's "2 mode" which has what basically amounts to an auxiliary overdrive integrated in the eCVT so that it can use smaller motor-generators over a wider range of speeds. Smaller motor-generators means more energy is transmitted mechanically, which means higher efficiency.
This is also basically the same reason the 1st gen Volt gets significantly worse highway MPG in range-extender mode than the (contemporary) Prius Hybrid (~35MPG vs ~50MPG): The 1st gen Volt eCVT was envisioned as an EV with a range extender (where energy usually comes from a battery), while the Prius's eCVT was optimized for driving primarily on the ICE, with the battery only supplementing acceleration.
https://www.cnbc.com/2023/11/07/new-ram-pickup-ev-has-gas-po...
EDIT: this is a big secret that none of the marketing materials (want to) make clear.
I'm uncertain the output of the wankel, but maybe with its power-to-weight it might get closer to being able to drive on gasoline in a self-sufficient way.
For too many freeways, for too much of the day, I can ride a bicycle at speeds greater than the surrounding traffic.
And in those conditions hybrids, and electric cars are perfect.
For comparison, rotaries get about 100psi of compression, a modern gas car can more than double that. More compression, more efficient burn.