Mazda’s rotary engine in the age of the electric car | Hacker News Reader

Mazda’s rotary engine in the age of the electric car

nippon.com

240 pointsgascoigne291 comments

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.

>20-30%

Your total guess is orders of magnitude off.

Most software today is in the ballpark of 10,000x slower than it should be.

That's not what they're saying. They're talking about things that couldn't be done due to performance issues but are now possible thanks to improved hardware.

Software developers rarely question the hardware they are developing on/for. Unless you were developing in the pentium and earlier era, it has almost definitely not been true for you that “the hardware is not capable of running this idea” (ignore certain obvious high performance sectors).

Dunno, there are some pretty large exceptions to that.

The AI boom is almost entirely a thing due to advancements in hardware. You simply couldn't build ChatGPT with 2014 hardware.

10 years ago we designed a system to approximate a chunk of maths we did not have time to do every second, too slow.

If you started from scratch you'd never even consider doing anything other than the original obvious "correct" maths. Approximation be damned.

This factoid about web dev misses the fact that the JS developers today are often the same people who were writing CGI, Perl and PHP, some of us also ASP and later ASP.NET, or some Java. We always knew there are good parts about the old ways and merging the two was always the goal - it just takes some time to get the client side right and then to get the merging right.

This especially applies to React core devs. Remember XHP?

It’s hard to tell if you can improve technology until it’s good enough (jet engines) or if it’s a futile dead end.

Kelly Johnson of Skunkworks had a keen mind for differentiating between them.

It is indeed and sometimes it can go either way depending on the time point. A good example is what eventually became the PW GTF, the PW8000, discussed here:

https://cornponepapers.blogspot.com/2006/04/short-life-and-u...

Just a little bit to early but when they came back to the idea they ended up with a technology that they use on all their new (big) commercial engines.

For anyone that doesn't know the name Kelly Johnson I recommend "Kelly: More Than My Share of It All". A rare person who combined technical genius with an ability to get large scale things done.

Adding to the list of deficiencies, Apex seals!!

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.

The rotor is pretty much unable to be cooled too.

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.

The same reason we don’t use two strokes anymore? They’re environmental disasters and wildly inefficient.

The comment I replied to said a rotary couldn't have variable timing and so I was simply refuting that claim.

I'm not trying to make a case for either engine here.. there's clear reasons why four cycle piston engines won out for the most part (other than a few powersport and power equipment niches).

I completely agree, there's no reason any engine can't have VVT. I believe that famous northern european supercar guy, I think "Keurig" or something like this is is name - he took an old beater car 4 cylinder engine and superpowered it with some pneumatic / AI controlled valves - I think it's called "FreeValve" or something like this.

As for two strokes being "less efficient", efficiency is in the eye of the beholder. We see on the recent key bridge collapse lore that indeed these mammoth cargo ships run two stroke engines, in this case, two stroke turbodiesels which formerly were the mainstay engine of american trucking. The humble weed eater or chainsaw also tends to use two-stroke - why? because they're far more efficient, almost twice as efficient, in power-per-pound of engine i.e. in applications where lightweight is needed. An added benefit is mechanical simplicity. The double-power notion follows from the name, these engines give power every two strokes versus every four strokes. Two strokes are also really interesting in that exhaust manifold geometry interacts with the engine stroke with pressure waves, look at any two-stroke dirtbike and some will have these goofy-appearing exhaust manifolds for this reason. Granted the major challenge of two-strokes is dirty/pollution and being oil-burning by their nature.

The two stroke diesel isn't really comparable to typical two stroke gas engine ala. chainsaws and dirtbikes.

Because they use forced induction, they don't use the crankcase for scavenging and thus have a "normal" four stroke style oil sump.

Since the total loss lubrication system isn't used, they aren't actually any dirtier on principle than a four stroke.

The big ship engines also turn very slow so a lot of the physics behind it is different, and different fuels are used compared to road going diesel (different grades of bunker fuel).

You are correct that GM used to make a two stroke diesel for over the highway and medium-sized equipment use but they were phased out because they weren't competitive on fuel consumption and their crude superchargers were incredibly loud. Growing up my Dad owned a compactor with one and I'd have to wear ear plugs under my ear muffs to not be deafed by "the screaming jimmy" (Detroit 471, IIRC).

That hasn't been true for a long time. Two strokes barely produce more power per CC, and the emissions are not compatible with any modern standards. They're used in yard tools becuase they're CHEAP.

Power per CC is the wrong way of looking at it. A 250cc displacement two stroke is still much lighter and more compact than a 250cc four stroke engine so you want to focus on power per weight.

The cost argument probably is the main driver for yard tools (Honda makes a great 4 stroke string trimmer for instance) but Endurocross bikes, trials bikes, and mountain snowmobiles especially still use two strokes for the power to weight advantage. The BRP snow machines even have direct injection which really helps emissions, though obviously they still have total loss lubrication so they're dirtier than a four stroke.

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.

I wonder if we're missing something?

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.

https://carbuzz.com/news/mazdas-new-rotary-engine-variable-v...

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.

I admire the idea of hubmotors as well, but once mentioned them to a car engineer, who basically said "think about unsprung weight".

Which, now that I think of it again, is a bit silly because most EV mass is battery, not motor.

Shaving unapeung weight down, even by a small fraction of the total vehicle weight, will dramatically affect handling and perceived ride quality. Even though the battery makes an EV heavier than an equivalent ICE car, there is still the need to reduce unsprung weight. Thus, it would be better to have the motor inboard and run a cv axle out to the wheel than have the entire mass of the motor out on the wheel.

Unsprung weight does not have a linear impact.

Increasing the unsprung weight by 10% causes a 100% deterioration in handling and ride quality.

You're suggesting increasing the unsprung weight by about 400%.

> once mentioned them to a car engineer .... now that I think of it again, is a bit silly

No offense, but this right here is peak HN.

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..

That can’t be the whole story?

It has nothing to do with sentimentality.

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...

While I also share in your love of Wankel engines I think you are grossly skipping over their unreliability (mostly around the seals).

There's an extremely good reason "rural" applications are reliant on diesels. It's not uncommon for a diesel engine to hit 300k+ miles and still operate reliably. And not to mention simply. They are relatively easier to maintain than a standard gasoline engine.

Diesels also greatly benefit from being on a fixed power band. It's why they work so dang well on large ships.

Where as rotary engines(rx-7's at least) need a rebuild every 80-100k miles. So I think without really knowing what you're doing mechanically or having access to a mechanic that can repair those kinds of engines you aren't going to be super reliable with a rotary.

Green hydrogen approaches via rotary engines does seem quite interesting. I think if the reliability of the rotary could be improved that could have some serious merit. Or at least making small enough generators that you can easily ship them out to be replaced/repaired could have serious merit.

My guess is this: Mazda believes its battery-only range is enough for most driving situations. If they're right, the rotary engine's more limited lifespan would be a moot point since it will still last the vehicle's entire lifetime. One can only hope.

The maintenance thing is concerning, though. Nobody knows how to work on these things. As an RX-7 or RX-8 owner, you're either doing most of your engine work by yourself, or you're driving very long distances to go to a specialty shop. We can pray that it's reliable enough for this to not be a problem, but if these engines start failing, it's going to be a huge mess.

Side note - the geometry of a Wankel rotary engine makes a diesel version impossible - you can't get enough compression. Diesel rotaries do exist though. An American company called Liquid Piston is making them for the Army. The intent is to use them in diesel generators, since the resulting generator is much smaller and lighter than more traditional diesel generators.

"The multiple use of hydrogen (e.g. a stove just needs to burn the gas)"

Oh, no. You don't want to be cooking over a flame capasble of melting platinum. You will utterly destroy your pots and pans temper. I do jewelry work, hydrogen gas is one of the hardest gases to properly work with. I'd rather let a newbie play with oxy-acetylene than a hydrogen torch.

Could you elaborate on why this is? Naively, the temperature of a simple hydrogen flame is not too much hotter than a methane flame. And in torch form, it seems like MAPP is several hundred degrees (C) hotter? And oxy-acetylene a few hundred more on top of that. Any torch seems to have way more than enough heat to melt platinum (1700C)?

A hydrogen-oxygen flame (this is what happens naturally since you're burning hydrogen in open atmosphere) burns closer to 2800C, and also has the fun property of causing embrittlement (which is especially worsened at higher temperatures where metal expands and becomes more porous.) Your steel pots and pans are essentially toast under a hydrogen flame.

People think "Fire is fire" and no, no it is not.

When I was doing a bit of research, I was coming up with 2050C or so for hydrogen open flame (in air; with pure oxygen it was 2600C), and about 1950C for methane in air. Natural gas is mostly methane and has a pretty similar flame temperature, but it does not melt our pots and pans. Why would hydrogen? It is hotter, sure, but not dramatically so.

Hydrogen really likes to strip out all the carbon in your steel, leaving you with pure iron (given enough time). It's a well known phenomenon called hydrogen embrittlement.

Where are you seeing all these magically rotaries? Yeah the guys in NZ and maybe some Americans(Rob Dahm) have some high horsepower rotaries but let's not pretend it's the norm and that they are more reliable than an equivalent piston engine.

Built Nissan VR38DETT V6 produces 2000hp. A full billet block will produce 3000hp+. Nissan GT-R guys(T1, ETS,AMS) Let's not even get into the big block V8's because these things will run all day at that hp and be a whole lot more reliable than anything mentioned here.

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.

Except in Mazdas case the data was there to characterize it. They were notorious for failing very early and suffered bad seals around the rotary shaft. This has been well known for a long time. Many Rx owners knew ride it for 100k miles then sell it before it is too late.

That rate isn’t a deal killer for a plug in hybrid. The engine might see ~1/4th the ware per mile spending the vast majority of its time off.

Meanwhile hybrids really want a small and lightweight engine because of all the extra equipment.

Right but then why not use a more efficient and reliable engine? The motors will last hundreds of thousands of miles. The power circuitry should have a similar run if kept cool. Why have an ICE that's more likely to fail with lower thermal efficiency?

The engine used in Toyota's most popular cars runs in Atkinson most of the time with a 40% thermal efficiency.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7425626/

Because the engine isn’t the only thing being optimized. Suppose your options are 50 miles of effective plug in range and a 40% efficient engine or a 35% efficient engine and a 100 miles of plug in range. The second option might use a lot less gas.

It’s not just the cost of batteries that’s a concern resulting in short plug in hybrid ranges. Weight is a real limiting factor when you also need a gas engine and large fuel tank.

That would be a worthy trade, but that is vastly overstating the weight of aluminum block engines compared to Li+ and not at all realistic.

Atkinson engines have poor power to weight ratio at maximum efficiency, if you’re going to maintain that 40% then the engine needs to be oversized and under utilized. This is why many hybrids have seemingly much more power than they need. Thus the 194HP Prius vs 75 HP on a CX-90 plug in SUV.

Also it’s not just that you can have extra Li+ it’s also pushing around less dead weight when the engine is off. So you can get fairly close to a 50 mile range difference depending on exact setup.

Alternatively you can have a smaller Atkinson engine and a larger reserve on the battery, but that also costs range while still being heavier.

It could make more sense to use a conventional engine, but get extra expansion (the kind an Atkinson/Miller cycle engine gets from the longer expansion stroke) by turbo-compounding. That is, exhaust gases (still at greater than ambient pressure) go through a turbine that is coupled to the output shaft, or to another generator, rather than being used to drive a compressor as in a conventional turbo. Or, attach a motor/generator to a conventional turbo and be able to have it operate in both modes and use the battery to help overcome turbo lag.

Possibly, but that’s a lot of separate systems.

It might make sense to look into a turbine engine on a hybrid. Though that’s probably been investigated I doubt anyone has invested the kind of time and money needed to make a real shot of it.

You will never be able to match the weight of a Wankel with a regular reciprocating piston engine, all other parameters being equal.

> The engine uses port and side direct fuel injection systems (PFI/GDI, referred to by Toyota as D-4S); a cooled, external exhaust gas recirculation system (cEGR); and a wide range of authority variable valve timing with electric phasing on the intake camshaft and hydraulic phasing on the exhaust camshaft. Atkinson Cycle is implemented using late intake valve closing (LIVC). Effective compression ratio is varied by varying intake camshaft phasing.

Clever indeed.

> The motors will last hundreds of thousands of miles.

They might, but with the current average build quality of the car industry, the rest of the car will die well before that..

I had an RX-8 for years. It was made in 2007. Someone defaulted on the loan or lost it in a divorce and drove it into a lake, marking it salvage.

I bought it for 25k and was the owner from 2009 to 2019, from 60k to over 150k mi. The maintenance was a bit pricey, due to the fuel injection problems and the electronics would fail in extreme heat of sustained 110+ outside temp (southern california heat) or cold (seattle mornings). I eventually sold it for 9k to a collector. Not a bad deal at all.

$25k USD? That does sound really high for a salvage RX-8 in 2009. My 2005 RX-8 was about $12k CAD in 2011 and was not salvage status. Though I guess yours was just a couple years old, mine was older, but still.

In regards to reliability, not long into owning it I had to get the engine rebuilt due to low compression (luckily JUST within the extended warranty period which Mazda offered due to the very issue everyone in here is mentioning, loss of compression due to failing seals).

Wait you paid 25k for a used/salvage RX-8 in 2009? That sounds crazy. I think I bought one new for around 26k in 2008.

A new RX8 Gran Touring in 2009 was ~38k out the door, iirc.

I completed the purchase of one red new gran touring from Mazda, for delivery. The inventory was already taken, so they tried to swap a yellow and I cancelled. The next shipment was in another 8 months. Obviously, the internet RX8 community was active and I knew what was a reasonable pricing, at the time. I looked locally and I had a salvage checked out at Mazda. It was clean. As I mentioned, I sold it to a collector for 9k ~10 years later, as it was still in near-mint condition, regardless of the title status.

That does sound crazy. The sticker price for the brand new Subaru BRZ limited manual transmission I got in 2012 was $28k.

> I don't think there's a good reason to keep pushing down deadend reliability paths. [..] Not every novel and viable solution ends up being an enhancement.

That's a strange take.. It took decades to get reciprocating engines to be reliable, even deployed at scale. It also took decades to get jet engines reliable enough to comply with ETOPS requirements, to the point that it was long considered a pipe dream.

If it hadn't been for continuous investments over long periods of time and incremental adoptions of improved technologies, we wouldn't be enjoying any of their benefits today.

>In most applications you can work around this with redundancies and diverse technologies

Are you proposing that Mazda should put two engines in their cars? Or am I misreading this?

>hard earned decades of learning and be pressing advantages. Not every novel and viable solution ends up being an enhancement.

That's precisely what Mazda is doing. They poured decades of R&D into the rotary engine, and are leveraging that research to their advantage. Their engine's diminutive size means they can provide more interior space in a compact vehicle, which is one of the key benefits of an EV that you can't get with a conventional hybrid. There are innate material advantages to the rotary in this application, it's not just hand-waving.

There's nothing overly novel about series hybrids, either. Chevy proved they can deliver EV driving dynamics and efficiency with ICE convenience, even at a larger scale than Mazda appears to be targeting.

This entire comment perplexes me. I can't tell if you like what Mazda's doing, or if you dislike it.

Formula 1 cars are also series hybrids, for about a decade now. Electric motors drive the wheels at all times, and power comes from a turbocharged V6.

This is incorrect.

The F1 MGU-K is geared to the crankshaft via the timing gears. The wheels are absolutely NOT electrically driven, it’s working as a torque fill for the conventional powertrain.

These three paragraphs would have been a far more enlightening article than the posted article.

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.

If you're referring to Liquid Piston's "X-engine," as they call it, it's doing well. It landed a substantial military contract and will likely be used to make super compact, lightweight diesel generators.

> 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.

Exactly the opposite. In a series hybrid, there is no mechanical connection between the engine and the wheels - only an electrical one, with the battery still providing considerable power in most cases, even when it's depleted.

No hybrid ever runs its battery all the way to zero, and series hybrids are no exception. The Chevy Volt (the other notable series hybrid in Western markets) keeps its engine at a near-constant RPM by using the battery to supply torque, even when the battery is "dead." I would expect the MX-30 to do the same.

IME in several thousand miles driving/riding in a Volt, it hums away within a very narrow RPM band unless you hoon it on a dead battery in cold weather. Even then, it's not nearly as wide of a rev band as a conventional hybrid.

The part that's got me confused is the illustration near the end, with the caption: "In the MX-30 e-Skyactiv R-EV’s electric propulsion unit, the already compact rotary engine shares a common axle with the generator and motor."

I don't get how the engine can be decoupled from the wheels if it's mounted to the same axle as the motor?

The Volt is a series/parallel hybrid. It has situations where the motor is directly connected to the wheels. The BMW i3 does has a series only set up, but that’s probably because it also has a BEV only version.

I think you have series and parallel hybrids backwards. Series hybrids have no mechanical connection between engine and wheels. It's just an EV with a generator wired in.

> 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.

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.

It's not like the ICE doesn't do anything at low speeds, HSD uses the two motor-generators like a gearbox

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.

Maybe you’re thinking of the Edison kits? They seem to be putting a genset into the truck bed to do EV conversions: https://news.ycombinator.com/item?id=38893567

Then you may as well use a gas turbine, which is pretty much better in every way that matters for generating purposes.

Apparently they really don't scale well. I found this reddit post explaining it better:

> Gas turbines scale extremely poorly. They rely on small clearance between the rotating blades and the housings for efficiency. The smaller the turbine, the greater the relative clearance and the more energy is lost. Gas turbines, at least with established technology, make very little sense below 300ish HP. As a real life comparison: A Robinson R44 piston helicopter and an R66 turbine Helicopter have almost identical design, dimensions, and weights. Power is around 250 / 300hp. The former burns between 50-60L of gas per hour at cruise, the latter around 90-110L of Jet fuel.

1 - https://old.reddit.com/r/cars/comments/s8vkv8/are_wankel_eng...

Thank you, I've wondered for years why we didn't have tiny gas turbine engines...

They did come about, in the late 60s. Williams worked on these to varying degrees of success. One of their designs became quite famous. The F107 powers the BGM-109 Tomahawk and AGM-86 ALCM.

https://www.smithsonianmag.com/air-space-magazine/the-little...

There are companies that have tried, like Capstone (formerly Capstone Turbine, then Capstone Green Energy). Capstone declared chapter 11 bankruptcy last year (since emerged and continuing under new leadership.)

I still believe it's possible - for example by having all the blades enclosed in a ring you can avoid gas escaping 'round' the blade.

And with metal 3d printers rapidly falling in price, I think it will get within the scope of university students to prototype soonish.

Fun fact, before the oil crisis the French TGV was originally a gas turbine train.

They definitely exist for small RC aircraft, but they're very expensive.

And chew through fuel at truly extraordinary rates.

This one, for example, weighs 450 grams, but can consume its own mass in fuel every three minutes!

https://www.kingtechturbinesaustralia.com.au/product-page/k4...

300 HP = ~220 kW

I believe a Tesla can charge its battery pack at approximately this rate.

What's your point?

There's no reason to have a generator that charges the pack in a hurry. It really only needs to cover the maximum sustained average draw - driving up an extended grade at high speed. That's a lot less than 300 hp - it's probably not more than 80 hp or so at most.

I remember back in the 90's in auto shop we calculated that you'd only need 15-25hp continuous to essentially power all of a car's needs if you could smooth demand for power from the peaks over the length of the trip. It stuck with me as a surprisingly small number, but it mathed out, even including heating and AC. Cars are both larger and more aerodynamic nowadays; I wonder if the amount would still be the same?

A Model 3 or Model Y's "rated efficiency" of 250-300 Wh/mi corresponds to 16-20 kW assuming a 65 mph speed. That's well within that range you quote.

Will car owners be allowed to run their (25 hp) range extenders unattended? It could be useful to allow it to top up battery while parked, and the car could detect its in an enclosed garage.

For many 'standard' driving patterns (relatively 'flat' urban commutes with approx balanced medium length up and down grades) there are lightweight optimal solutions for EV's that can minimise both battery pack size (and weight) and the need to draw on a small rotary engine for recharge.

The UK's drone engines come in light and small with models that range from 5 BHP to 120 BHP with 40 BHP being suitable for broad swathe of "typical" driving.

https://www.aieuk.com/wankel-rotary-uav-engines/

Except extremely complex and usually hard to repair and inefficiently sized for a consumer car.

Actually far simpler than a piston engine; the fuel consumption and high production cost was what made them become extinct.

https://en.wikipedia.org/wiki/Chrysler_Turbine_Car

Of course Jay Leno has one; not sure if he’s featured it on his youtube channel (I don’t recall seeing it). Awesome

Here's his video on his 1960's one, which is worth watching and some neat animation footage from the period talking about the design: https://www.youtube.com/watch?v=b2A5ijU3Ivs

You can tell Jay genuinely loves cars and the history of the auto world. He's indulging his hobby interest in a way that will preserve these vehicles for future generations to see and learn about. As far as ways rich people can spend their money that's a pretty cool one in my book.

Yep I've seen it on his YouTube. The "EcoJet" https://www.youtube.com/watch?v=10q9_pB6unU

He said that they had to put really big brakes on it. I guess for some reason you can't just have a clutch. Personally if I was going for eco I'd want something more clever than "bigger brakes" but maybe building something like a Hybrid Synergy Drive is harder than building a turbine.

> He said that they had to put really big brakes on it.

There's no compression braking (engine braking) with a turbine engine.

Also: While Leno seems to be very interested in restoration, he also is clearly very interested in staying alive. He's got a ton of old cars that have been modified to have very good, modern disc brakes.

The EcoJet is just one of many examples of his cars that have been improved in this way.

Gas turbine cars have been built, but they tend to set anyone on fire if they walk behind the car.

https://bandimere.com/previewing-the-brakes-plus-jet-car-nat...

You're probably being silly, but for anyone else, gas turbines as used for generators don't have jet exhausts on them, i.e. https://www.youtube.com/watch?v=yR5oVn-Fvvg

So what do they do with the exhaust?

The reason jet engines have that sort of exhaust is because the primary purpose of the gas turbine there is to dump all the excess energy into the exhaust to make it go fast (so that the plane can be pushed ahead by the reaction force). They produce very little power, just enough to power the Auxiliary Power Unit (that manages the plane electronics, air-conditioning etc.).

If you want to use a gas turbine for producing power, you will set it up such that most of the energy goes into the work generated, rather than the exhaust, so it would be a cooler, slower exhaust, similar to an IC-engine.

Other people have already mentioned the distinction between turbojets (what you’re describing) and turbofan engines, but I think there’s another inaccuracy:

> They produce very little power, just enough to power the Auxiliary Power Unit (that manages the plane electronics, air-conditioning etc.).

The APU is a completely separate gas turbine that doesn’t rely on the main engines. As a consequence, the APU on an airplane will also have its own exhaust.

> If you want to use a gas turbine for producing power, you will set it up such that most of the energy goes into the work generated, rather than the exhaust, so it would be a cooler, slower exhaust, similar to an IC-engine.

Yes, obviously. At the same time, I was under the impression that turboprops and turboshafts on airplanes and helicopters still produced enough jet exhaust to represent a safety hazard, and in those applications you would also expect that most of the energy would go into the work generated rather than the exhaust. So is it just that this residential generator is even more efficient than the main engine of a heavy-lift helicopter? Is it because it’s a less powerful machine in the first place? I could continue to speculate about this but I don’t actually know.

It's because to extract all the energy from the exhaust, you need huuuuuge turbine blades to extract the last little bit of energy from very slow moving air.

So instead you extract most of the energy, and then use the last little bit as extra thrust in a plane/helicopter.

Depending on the plane and engine. Big jetliners have high-bypass turbines where they do intentionally produce a lot of torque, to spin a large compressor fan, but most of the fan air does not go through the combustion step, it's just used to react off of.

It's ironic how propeller airplanes evolved into jets for for thrust and back to propellers.

(A ducted fan is still a propeller.)

We may go to unducted fans soon. The latest iteration of this concept has two sets of fans, one set of which doesn't rotate.

Some models do, some others do not: https://www.youtube.com/watch?v=b2A5ijU3Ivs, https://www.youtube.com/watch?v=CdCudNSti_Q

Oh! The Batmobile!

Gas turbines are economically effective when larger than about 300hp. This border with time slow lowered, but I don't know, when it will appear somewhere about 50Kw of most popular automobiles (Toyota Corolla, WV Golf, Ford Focus).

When scaling down, gas turbines become much more expensive than ICE. For example on small planes market, exist many dual-powered models (in range 200-300hp), and with gas turbine it typically priced twice of ICU-powered with very similar parameters. Range of 1000hp+ on planes are near totally gas turbines.

Cosworth appears to have developed a microturbine for this purpose. And it (maybe?) has found a home in Ariel's latest insane car: https://www.carscoops.com/2022/09/ariel-hipercar-is-an-1180-...

AFAIK I think there is some difficulty in getting turbines to scale down efficiently with a need of a recouperator to maintain a higher core temperature for better efficiency. The recouperator is additionally expensive on top of the already expensive turbine cost.

Having a rotary with a turbo should be able to work better at a lower scale for a pretty cheap production cost.

Want to comment explicitly, though I upvoted a similar comment ... a gas turbine in a consumer car will be a maintainability nightmare (where do you find the technicians that can do anything with it?), at least in the current automotive ecosystem.

Chrysler tried it in the 50's and 60's

https://en.m.wikipedia.org/wiki/Chrysler_Turbine_Car

The ideas and innovations back then seemed amazing

Except for consumption, waste heat generated, size.

If all you need is power generation, use a nuclear reactor. Time to bring back the Ford Nucleon ! [1]

Comes with daily free health checkup (in form of an X-Ray).

[1] https://en.wikipedia.org/wiki/Ford_Nucleon

Aaahktually, the radiation given off the reactor would be too “hard” and useless in soft-tissue imaging.

Maybe neutrino imaging?

> I'd swear I remember discussion here a handful of months ago about a startup marketing a similar EV range extender engine design.

Possibly https://liquidpiston.com

Prior discussions: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

They have low compression though, which really hurts the thermal efficiency.

nit: it's spelt Wankel not Wankle.

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.

Sadly enough I did know that but wrote it wrong anyway! Fixed, thanks. :)

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.

I bet wankle development is also excellent for internal education of their engineers.

Mazda Rotary 'Launch':

https://www.youtube.com/watch?v=Hdjj52FUsEo

>These are some of the earliest ads I wrote and directed. Many of these ads cost our client just a couple of hundred dollars. When you look at this early attempt ant animation, it's not surprising that it didn't cost much more as a decent lunch these day!

>It's hard to believe now how absolutely 'revolutionary' the rotary engine appeared to be in those days. Many of us expected that they would eventually take over from reciprocating engines completely.

>This ad was produced just before the moon landing.

>The Agency was Hayes Advertising, Sydney. My producer (and dear friend) was Max Cleary and the Account Director was Vic Violet.

Mazda RX-3 Commercial:

https://www.youtube.com/watch?v=oHzeGEHWMjo

>Piston engine goes boing, boing, boing, boing.

>Mazda engine goes Hmmmmmmmmmmmmmmmmmmmmmmmmmm.

Felix Wankel:

https://en.wikipedia.org/wiki/Felix_Wankel

>Wankel and the Nazi Party

>During the early 1920s Wankel was a member of various radical anti-Semitic organizations. In 1921 he joined the Heidelberg branch of the Deutschvölkischer Schutz- und Trutzbund and in 1922 he became a member of the NSDAP, the National Socialist German Workers Party (or "Nazi Party"), which was banned soon afterwards. Wankel founded and led youth groups associated with a cover-up organization of the NSDAP. With them he conducted paramilitary training, scouting games and night walks.[3] When his high esteem for technical innovations was not widely shared among the German Youth Movement, he was offered instead the opportunity to talk about the issue of technology and education to Adolf Hitler and other leading National Socialists in 1928.[4]

>In the meantime Wankel's mother, Gerty had helped founding the local chapter of the NSDAP in his hometown of Lahr. Here Wankel not only rejoined the party in 1926, but also met the local Gauleiter, i.e. regional head of the NSDAP party, Robert Heinrich Wagner. In 1931 Wagner entrusted Wankel with the leadership of the Hitler Youth in Baden. But they soon fell out with each other, because Wankel tried to put a stronger emphasis on military training, whereas Wagner wished for the Hitler Youth to be a primarily political organization. In a particularly bitter and ugly controversy Wankel publicly accused Wagner of corruption. Wagner retaliated by stripping Wankel of his office by early 1932 and managed to have him expelled from the party in October 1932.

>Wankel, who sympathized with the social-revolutionary wing of the NSDAP with Gregor Strasser, then founded his own National Socialist splinter group in Lahr and continued his attacks on Wagner. Since the Nazis' seizure of power on 30 January 1933 had strengthened his position, Wagner had Wankel arrested and imprisoned in the Lahr jail in March 1933. Only by intervention of Hitler's economic adviser Wilhelm Keppler and Hitler himself, was Wankel set free in September 1933.[5] A fellow native of Baden and member of Reichstag from 1933 to 1945, Keppler had been a friend of Wankel and an ardent supporter of his technological endeavors since 1927. He now helped Wankel to get state contracts and his own Wankels Versuchs Werkstätten experimental workshop in Lindau.

>Wankel tried to rejoin the NSDAP in 1937, but was turned down.[6] With the help of Keppler, however, he was admitted to the SS in 1940 in the rank of Obersturmbannführer.[7] Two years later his membership was revoked for unknown reasons.[6]

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.

I don’t understand why Mazda doesn’t just make a drift-tuned electric car. You could do amazing stuff with software focused on that driving style.

A true electric successor to the RX-7 would capture so much attention.

Several reasons.

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.

“Adding power makes you faster on the straights. Subtracting weight makes you faster everywhere.” - Colin Chapman (Lotus)

I agree with the adage, but brute force seems to win in this specific case. Even though it's the lightest model made, my Miata only gets 0.82g on the skidpad.[1] The Model 3 Performance gets 0.96g thanks to its wide tires, which are needed to transfer all its power to the asphalt.[2] This difference isn't just due to the 1990 Miata's older suspension and tire technology. Even the latest Miatas only get 0.90g of lateral acceleration.[3]

1. https://www.caranddriver.com/reviews/a15141519/1990-mazda-mx...

2. https://www.caranddriver.com/reviews/a36329678/2019-tesla-mo...

3. https://www.caranddriver.com/reviews/a22678665/2019-mazda-mx...

> I agree with the adage, but brute force seems to win in this specific case. Even though it's the lightest model made, my Miata only gets 0.82g on the skidpad

The current ND2 (2019+) Miata regularly pulls ~0.95 stock in magazine tests, almost identical to your model 3 Performance number.

> https://www.motortrend.com/reviews/2019-mazda-mx-5-miata-clu...

You can exceed 0.95 and get ~1 on an ND Miata with slightly wider than OEM tires (still on stock rims) and a little more negative camber, which is widely done to the car by the enthusiast community. Similarly, you can get more out of your very own NA (1990) Miata with simple tire/alignment changes, even more with cheap new sway bars or springs etc.

> https://help.flyinmiata.com/align-your-suspension-chakras-By...

The ND2 is a ~1070kg car.

To use a more fitting Lotus example here, the ~900kg Lotus Elise (50% of the weight of the Model 3 Performance) pulls 1g when tested by Car and Driver:

> https://www.caranddriver.com/reviews/a15146116/2007-lotus-el...

Yes and you can modify the Model 3 to increase cornering ability, but like in the case of the Miata it means increased tire wear, worse comfort, and worse mileage. Not to mention money.

My point was simply that even if you know vehicle A is twice as heavy as vehicle B, you don’t know for sure which one is faster in the turns.

> increased tire wear, worse comfort, and worse mileage… Not to mention money

It literally means none of this to change a miata as I described - we are talking a single degree of camber here not a race car. An alignment is normal maintenance - no change in price - and tires stay the same price if you go up a single size, so if you do this when getting new tires anyway it costs essentially nothing.

It’s a minor camber change (done at a standard alignment as normal, no extra special bits - just ask tech nicely). The tires will last just as long for your driving style and gas mileage unaffected. Comfort unchanged - no spring, damper or tire pressure changes.

The point this all makes is simple factors beyond weight have a huge bearing on constant lateral load car will sustain, to the point it’s almost pointless to compare weight and max corner load. You will never see car enthusiasts comparing weights of their cars and arguing in favour of more weight, almost ever. This entire comparison is pretty odd. No one who knows what they are talking about is going to question the classic Colin Chapman quote because physics didn’t change since his death - the concept of same car but lighter was faster in the 60s and 70s, and is still faster round a circuit today. It’s why race cars set faster lap times as the fuel tank depletes, which proves the point beyond doubt.

If you haven’t had a good alignment done to your NA recently get it done and don’t be scared of small adjustments, they won’t ruin anything - it does quite the opposite! - and the numbers that work great for all miatas are insanely well documented online. Steering feel will thank you for it. It’s the first thing I will have done to any generation of miata - they all benefit a lot, and usually arrive from factory not very accurately setup at all - you will see this when you have first alignment done and brand new car has initial numbers all over the place.

I’ve owned and maintained multiple examples of all four generations of the car over the last 20 years - a precision alignment with a touch more camber/toe is one of the easiest, best and cheapest (100-150 dollars typically in major US city) things you can do to the car - the miata is all about that steering feel which is easily corrupted.

Going with this theme, the idea of a battery car with longer range is appealing to me. However a smaller battery but quick charging would mostly remove the need.

I’m not sure I want to drive around with a capacitor in the boot, but a huge battery isn’t ideal either.

That's the whole thing right - most people don't actually care about having 600 miles of range, they care about being able to "refuel" quickly. My Mercedes AMG would only do like 200 miles on a tank of fuel and I don't ever recall having any kind of range anxiety with it, because you could gain all of it back within like 5 minutes and keep going.

It would be theoretically possible to have a small battery ("just a 60 mile/100km range, or even smaller) combined with a generator, but I don't know if markets would appreciate that.

This company Toyota started a pretty popular line of hybrid gas-electric cars, maybe 20 years ago, called Priuses. I think they sell pretty well. I see a lot of them running as taxis. The new ones can plug in and drive a few miles on the highway on pure electric before starting the gas engine.

Priuses do use the ICE to drive the wheels though, no?

Yes, they are parallel hybrids.

Its called a series hybrid, there have been a small number of plug-in hybrids that used that design, they weren't successful in the US and are no longer in the US market. But that may not be anything particular about the technology; I wouldn't generalize from such a small set.

The McMurtry Spéirling is claimed to be under 1000kg. A lightweight electric car may not be an oxymoron after all. Just some concessions to make.

Our VW e-Up is just below 1200kg and has 150 miles range from a 36kWh battery, fits two of us, baby seat, and Costco shopping. You can absolutely have a lightweight electric car, just be realistic about what you're getting.

The McMurtry Spéirling is not street legal and it costs $1 million.

The Bugatti Veyron launched in 2005 with 1K horsepower and cost > $1M. As of last year, you could get a Dodge Challenger with 1K horsepower for <$100K. Those prices are unadjusted for inflation, so the price difference is even greater than 10x.

While it is not a guarantee, the innovations in today's supercars do tend to become much more common with time.

I think you're drawing a lot of conclusions from that correlation.

what innovations did the Veyron introduce that the Challenger used?

And what a car it is!

It seems Mazda is going ahead with an electric Miata, they just don't know how electric it will be:

https://www.motortrend.com/news/2026-mazda-mx-5-miata-electr...

I think it’s likely to be a mild hybrid.

Sports cars, in general, are much loved but seldom bought.

You’d think its a different time now that the kids who grew up lusting over these cars now have money for one, enough money to create a new car market where even a pickup truck can be almost six figures optioned out

Yeah but by the time you're able to afford one you will be married with kids and therefore prioritize practicality over drifting.

I'm literally waiting for it, have been for quite a while. Small cars have numerous benefits over just being sporty. An electric, or even a hybrid 86/BRZ or miata would be great, but can't be compared to the mini or fiat, and while tesla might be fast, it's huge. Even with a price increase these could be more affordable than a lot of sports cars. The 86/BRZ has been a huge seller too.

The EV market is so frustrating right now. Everything seems to more or less be an SUV.

I wish someone would deliver something small, light, aerodynamic, stripped down and without features of marginal utility. Sportiness sort of comes for free.

The only important features for an EV are (excluding safety issues) are change speed and range. I might add a heat pump for the cabin (and battery in cold climes). Skip the screen and just let me use my phone and give me physical controls.

If a car marker thought seriously for a moment and resisted the full techno wank that is inflicted upon us at the moment, they'd make a lot of money.

It's a car, it's not that complicated, get back to basics.

I'm not really that interested in cars. We drive a Honda Fit (Jazz for the rest of the world). I was really excited 8+ years ago when Honda said they might do an EV version of the Fit/Jazz - just perfect for my wife and I, combining the great utility of the Fit/Jazz with our preferred power source, all in a reasonable sized and reasonably slick package (incredibly internal visibility also).

Not only has this not happened, Honda have even stopped selling the ICE version of the Fit in the USA. The closest thing to this concept - the Nissan Leaf - has also been discontinued in the N. American marketplace.

Truly pathetic.

The MG Cyberster is coming out pretty soon, also the electric Cayman, though it's a completely different price bracket.

>A true electric successor to the RX-7 would capture so much attention.

Not entirely sure that would be the case even with a Red Sun label on it.

As someone who owns a white turbo fc, sans stickers, I would be interested.

The unique sound is a key selling point for rotary enthusiasts. Kind of raspy, almost like a 2-stroke.

Kind of raspy, almost like a 2-stroke.

That's not what the TV commercials from the 70s told me:

"Piston engine goes 'boing, boing, boing, boing'"

"But the Mazda goes 'hmmmmmmm'"

(Oh, of course there's a YouTube video: https://www.youtube.com/watch?v=oHzeGEHWMjo)

It can be the whole story.

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.

> Mazda is BlackBerry

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.

As much as I love Mazda, if the only thing they'll have for them is fun, they are dead as a mass manufacturer.

Every car/SUV they make is highly rated. Reliable, well priced, good looks, premium interiors and, yes, fun.

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.

> Does it produce more low end torque verses a traditional gasoline combustion engine? If not then why is it better suited for electrical generation?

What does low end torque have to do with electrical generation?

It's not a direct connection.

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.

The side-port exhaust used in the rx-8 generation RE substantially improved the emissions situation, but the fuel efficiency is still trash by modern ICE standards.

It's been awhile since I gave a damn about wankels (or ICEs in general), but ISTR there being a relatively low limit to achievable static compression ratio due to the fundamental geometry of the swept volume. Modern ICE engines are largely exploiting the combination of direct injection and the high compression ratios it enables to improve their thermal efficiency. Between the relatively low compression ratio and sub-optimal combustion chamber shape and the fact that it migrates around the housing with the power stroke, the wankel is pretty much doomed in a world that cares about efficiency.

disclaimer: I've worked hard to discard all my gearhead knowledge, but went fairly deep down the rx-7 rabbithole in my 20s-30s. Take the above with a big grain of "I may be senile and overconfident in stale once-deep knowledge" salt.

to the best of my knowledge, you are correct on all counts. the new rotary used as a generator does also have direct injection, but i don't think they gain much compression ratio from that, for the reasons you stated. it's more for efficiency. without reading up on it, i believe the combustion chamber mixture and flame front propagation is not great in a rotary (hence RX8 has 2 plugs per chamber, leading and trailing) and DI should improve on it.

Once you discover a hammer, everything looks like a nail.

There are better designs than Wankel for "series hybrid" application.

https://newatlas.com/automotive/innengine-one-stroke-engine-...

Pretty much.

One (potential?) advantage not mentioned in the article is lighter weight relative to a similar traditional engine.

Would be good as a range extender motor. I think smaller battery cars with range extenders should be a better option than what we have now.

Oh my God, if Tesla sold a range extender gas generator that sat in the frunk, and could add an extra 150 miles.. take my money.

Power/Weight is extremely high. A tiny wankel will do the job, and weight is everything on cars.

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.

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 fixes the terrible efficiency by having an entirely different power source?

As others have pointed out, the article doesn’t do a great job of explaining how the rotary helps.

A YTer said it allows the rotary engine to operate in its best circumstances. It’s essentially a range extender while battery tech improves.

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've owned an RX-7 and an RX-8. They definitely didn't have flat torque curves. Even the twin sequential setup on the FC was far from flat.

But even if it was, that was the most powerful stock configuration made and it topped out at ~184 lb/ft. :)

Torque was never the strength of this motor.

I had a turbo FC and beg to differ. A dyno graph of a stock car also begs to differ.

Dyno graphs aren't consistent with that:

https://www.google.com/imgres?q=1993%20stock%20rx-7%20dyno&i...

They are fun engines, but low end torque just isn't their thing.

sorry, why would low-end torque be important?

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.

I don’t think it would be possible to be faster.

Total time stopped: 7 minutes.

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.

> and in fact that is exactly what happened with many European plug-in hybrids sold as company cars.

The counter force to that is that lease car drivers have to meet certain fuel efficiency goals, so, adopt an efficient driving style and plug in the plug-in hybrid instead of ragging it around; I'm not sure what the consequences were though, probably more taxes. Either way, enough of a push for my colleagues with plug-in / hybrids to do the thing.

I'm curious which country it is that has these fuel efficiency goals.

That and the fact that in the US the car is tuned to only turn on the range extender when the battery is nearly dead and it can’t always keep up.

Thankfully it’s trivial to change.

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.

I thought the idea is that the battery acts as a buffer and the engine is always either off or at constant rpms in its ideal power band.

You also have to keep in mind that the average consumer is…quite stupid, or more charitably quite sensitive. There are lots of cars on the market today with CVTs (continuously variable transmissions) that enable the motor to always be driven in the most optimal RPM. However in software, the engines are intentionally driven at less efficient rpm’s so that the driver can feel steps and lurches as the car accelerates which is somehow more “correct” to the average car buyer.

I just bought a car with a CVT and now I’m upset.

No need for that. While it's possible to point out particular "performance" car CVTs that do indeed compromise efficiency for aesthetics, that's not done in most cases, at least not to the degree that efficiency is significantly compromised. The actual reason for not doing as suggested and placing the entire burden of acceleration on the CVT (thus keeping the engine at a constant RPM) is that this would be a foolish design.

Between the precision electronic fuel delivery and ignition, variable valve timing, variable plenums and other features of your ICE, all of which are invariably present today when an ICE is the prime mover, the ICE power curve has wide RPM bands in which its efficiency is high. By taking advantage of these wide bands of efficiency, the CVT can have less size, mass, mechanical loss, cost, etc., than would be the case if the CVT were required to precisely enforce a constant engine RPM.

Minimizing the burden on the CVT is crucial because CVTs are a mechanical compromise: they are inherently not as strong as a conventional transmission and generally have higher mechanical losses.

The manufacturers responsible for these designs aren't actually fools, slavishly beholden to some old fashioned transmission aesthetic. They're responding to a large number of pressures and doing the math. The math says that constraining the ICE to an efficient, yet non-zero, RPM band is better than trying to manufacture miraculously agile CVTs.

Priuses do a power split where the engine drives the wheels and the motor-generators act sort of like an electrical CVT. You would have to be out of the power band a lot to make a series hybrid more efficient than a parallel hybrid. The reason locomotives are series hybrids is just that mechanically switching that amount of power isn't practical. To my surprise, even off-highway mining trucks mostly use mechanical transmissions.

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].

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

> 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)

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.

> 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.

This is partly because the Prius is optimized as a hybrid-first design, while the Volt is optimized as a BEV-first design. The difference is in the gear ratios of the planetary CVTs: Prius is optimized to minimize the amount of energy transmitted electrically, which maximizes overall efficiency.

However, that design is sub-optimal for a car that you want to act like an EV. As your link shows, the Prius must have the ICE spinning at speeds greater than 42 mph. Not a problem for a hybrid, but that doesn't work for a "range-extender EV" like the Volt. Additionally, (before the current model year) Toyota's plug-in hybrids can't provide full acceleration in EV mode--they always kick in the ICE when you floor it. That's again a consequence of how the transmission is optimized for hybrid operation.

By making compromises to the gasoline efficiency, GM was able to create a car which acts like a real EV most of the time: A 1st gen Volt will go at least 80 mph in EV mode, and won't turn on the ICE (come hell or high water) until the battery is below 5%. And, it will do that for an honest 35 miles of freeway driving.

The second generation Volt uses GM's "2 mode" hybrid transmission, which closes some of the MPG gap by adding one fixed ratio (100% mechanical transmission) as well as high and low speed eCVT ratios (to minimize electrical power in those two speed ranges).

https://www.gm-volt.com/threads/gen-2-volt-transmission-oper...

> However, that design is sub-optimal for a car that you want to act like an EV. As your link shows, the Prius must have the ICE spinning at speeds greater than 42 mph. Not a problem for a hybrid, but that doesn't work for a "range-extender EV" like the Volt. Additionally, (before the current model year) Toyota's plug-in hybrids can't provide full acceleration in EV mode--they always kick in the ICE when you floor it. That's again a consequence of how the transmission is optimized for hybrid operation.

True in 2007.

But 17 years later, the Prius Prime 2024 has 100% EV mode even at highway speeds. Toyota has improved the design since that webpage was made in 2007.

Despite this change, the Prius still achieves 57 mpg, even better than ever before.

Prius Prime is a proper EV-only mode, albeit a touch underpowered but its EV mode now covers all possible driving conditions.

------

I believe the modern Prius 2024 has 17,000 RPM limits now, as well as tweaked gear ratios and far larger EV and ICE engines. Despite the larger 220hp aggregate engine, the Prius remains absurdly efficient. Both in EV mode and in ICE mode.

This is why I was saying that the Volt has wasted it's opportunity. GM was ahead in many respects 10 years ago, but Toyota has caught up. The GM advantage has been squandered.

> By making compromises to the gasoline efficiency, GM was able to create a car which acts like a real EV most of the time: A 1st gen Volt will go at least 80 mph in EV mode, and won't turn on the ICE (come hell or high water) until the battery is below 5%. And, it will do that for an honest 35 miles of freeway driving.

Prius Prime added a button to enter EV-only mode in this 5th generation design starting in 2023.

It took Toyota too long to add this feature, but now that it's here the 5th generation Prius is a far better choice.

Especially because Prius still has it's trademarked 50+ MPG and 130+MPGe ratings. Top of the line efficiency.

It's somewhat frustrating to see Toyota catch up when GM was so far ahead for so long. Also because GM killed the Volt.

> All in one planetary gearset.

In the first version. In later versions they added a second separate gear set. Then even later they merged the two gear sets back into one but with two separate planetary arrangements.

I'm not a particular fan of EVs in general, but I do very much like the Prius and the careful engineering that went into this drive train. This video[0] does a great job of explaining the overall system in the context of an operating vehicle.

[0]: https://www.youtube.com/watch?v=PIYNAroYEk0

The Volt was a great idea, but I couldn't make it work for me. I need a four door car which regularly holds four people, and the Volt (especially the newer one, which was when I was shopping) has a low roofline. Only time I've hit my head harder on a car roof was when I tried to sit in a new Supra. So I ended up finding something where the physics of getting in and out were more agreeable. Still a sedan, just with fewer headaches.

I wish they had offered that same powertrain design in something other than a Cruze.

Both the Chevy Volt and Toyota's hybrids rely on a planetary gear set to blend power between the gasoline engine and electric motors.

Here are the differences explained in more detail:

http://roperld.com/science/ChevyVolt.htm

At least one version could run as a series hybrid or as a torque combining hybrid, depending on which clutches were engaged.

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.

https://www.nissan-global.com/EN/INNOVATION/TECHNOLOGY/ARCHI....

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.

Yes, that was one of the limitations of the Honda Clarity. The EV motor wasn't big enough and it was all too eager to switch to the gas engine. Looked like it was a comfortable car, though.

GM setup the Volt to alway have full power in EV mode and the vehicle would never automatically switch to gas unless the battery was discharged.

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.

Still seems like a traditional hybrid would be more efficient. I'm not sure what's gained by removing the mechanical coupling altogether (other than cost/reliability).

Honda has an interesting solution in their latest gen hybrids, where the electric motors power the wheels at lower speeds, the combustion engine runs at its optimum RPMs as a generator while the electric motors are working, and at higher speeds a clutch activates that changes the coupling so that the combustion engine directly drives the wheels.

Whatever the Prius HSD does is great. It'll do 50 MPG city and highway. No clutch at all.

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.

https://www.cnbc.com/2023/11/07/new-ram-pickup-ev-has-gas-po...

have a look at the new Mitsubishi outlander EV, it kinda works like the volt

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.

Weeelll, "freeway driving" depends on the freeway and time of day.

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.

I don't know where you get this idea, my Chevy Volt is perfectly capable of reaching 70mph+ while holding the battery level steady (in the Hold mode)

The volt is not a serial hybrid. The motor is physically connected to the wheels (in an interesting way btw), so it can drive on gasoline alone.

The Mitsubishi Outlander plug-in hybrid can do up to 60mph on a flat road in serial hybrid mode.

Does that matter in practice? What's the effective range for highway driving in these cars?

I recall the i3 rx only had about 2 gallons of gas.

I was always impressed by this engine in my childhood, but it never really caught on. Here is a variation I've seen on the internet, i'm not sure how practible it would be. https://arstechnica.com/cars/2023/11/this-inside-out-design-...

https://www.youtube.com/watch?v=X5UOqZn43gY It has longevity-issues due to an internal roller bearing on the main shaft.

Depending on how it's being used. Rotary engines are better in some ways when being run hard (ie as a track car) compared to more traditional layouts.

wow, thanks. I didn't expect to watch the whole thing, but I did.

Mazda and Toyota formed an alliance to share technologies to fill gaps they each had. I'm guessing this is fruit of that partnership.

Where do read Toyota into this announcement?

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.

The fortnine video about this type of engine is quite interesting:https://youtu.be/-3HBAvkc4a0?feature=shared Mazda might be one company with the expertise to make it work. The range extender, not directly driving wheels, can work in its optimum temperature, rpm and on a pretty constant load. The ability to work with multiple types of fuel adds to its versatility as a range extender. Since Mazda has alliances with other Japanese car makers, if it is indeed a good proposition, I reckon it would appear on other Japanese cars too, and maybe in BMW as well.

Almost feel bad for GM, the article did not mention the Volt when talking about series hybrids. But I guess it is a Japan-focused news source.

The funning thing is, people I know who have a Volt love their cars. Haven't tried one myself. My family has a couple of paid of Toyota gas vehicles that are 8+ years old and will probably live longer than we will.

I drive a 2017 Volt gen 2 and love it. I drive about 90+% of miles in EV mode and mostly use the gas engine for road trips. It is fun to drive and (other than a couple of common, and easily fixed issues) has been reliable. I do plan to replace it in the next year because it sits very low and has a very low roofline. My spouse is uncomfortable getting into it so we almost never drive it when together.

I don't understand why GM never put the Voltec drivetrain into a compact SUV. It would have sold much better than an odd looking hatchback with little interior space.

Word is that PHEVs are expensive to build but customers didn't seem to be flocking to them. Traditional buyers didn't see the need to save money on fuel and didn't believe that climate change "stuff". More adventurous buyers were flocking to the new full EV coolness.

I think GM decided that the future was BEV that they would be able to come out strong with EV version of their products and sell them at reasonable prices. With the component shortages and inflation, their price targets became untenable. Then they ran into assembly problems with their battery packs so they couldn't even deliver at any price.

Now they have backtracked and said that they will bring out hybrids (maybe PHEV?). It's not clear which platform they'll use for that. I think that they have an Equinox-sized hybrid in China that they might bring over.

The flailing around EVs vs hybrids and the on again off again story on the Bolt don't make it look like they know what they are doing, sadly.

BTW the Bolt actually has a surprising about of room in it, but even so, its on the small side. Equinox would fit more customers.

My friend has a few volts from 2013-2018 (that he loves) and I usually borrow one for trips to see family and friends 12-16 hours away rather than drive my truck. I think they're great vehicles. ~40mpg at 80-90mph on the highway is really good though it does take premium for that 40+ mpg and it has a smallish tank so you have to stop a little more often. Once you reach your destination all driving is done on electric and charges overnight on a regular outlet/level 1. Their maintenance is surprisingly low for what seems like a complex system.

Crazy to me that Chevy has abandoned their PHEV platform considering others, like Toyota, are in such high demand. Speaking of which I spent a week with a Rav4 Prime and found that equally as impressive range-wise as a volt.

Vehicles with range extenders are the exception, even if they seem a good option, they didn’t gain market share.

The point is the wankle engine. No sense in mentioning the Volt.

They did mention a couple other series hybrids, however, neither of which has a Wankel.

As mentioned elsewhere in this thread, the Volt is not actually a series hybrid despite being marketed as such. There is a planetary gear set (ala. Toyota hybrids) connecting the engine to the wheels at most speeds.

I like the LiquidPiston engine technology as an inverse Wankel engine. A lot more efficient.

A good article.

It barely addresses the elephant in the room however: The Toyota Prius and its Atkinson engine.

The Atkinson engine is 99% the same as a regular engine (aka: Otto cycle), except the timing is different. Instead of closing the valves when the piston+cylinder is full of gasoline+air mix... the Atkinson engine waits a bit and "leaks" some gasoline+air back out before closing the intake valve, effectively burning only 70% of the fuel, but getting maybe 85%+ of the power of a regular Otto-cycle engine. Its a simple and cheap tweak to a traditional engine that grossly improves efficiency (but at huge costs to low-end torque).

Basically, any regular ol' carmaker who is mass producing Otto cycle / regular ICE engines can easily tune their piston timing to be an Atkinson engine instead. I believe Toyota even has computerized controls today that switches between efficient Atkinson (lower-power but higher efficiency) modes and powerful Otto cycle (higher power but lower efficiency) modes, though this control isn't really used too often in practice.

---------

This article makes a good point that Mazda has a culture of this... rotary engine. It does compare it (somewhat unfairly) to the Atkinson engine though (inside the Prius and RAV4), I don't think anyone expects any traditional engine (Otto or Rotary) to keep up with Atkinson Engine efficiency.

Its a good try however. But it does raise the question of what benefits can this rotary engine give over other engine types.

I am mildly annoyed that Toyota calls late intake valve closing (LIVC) an "Atkinson-cycle" Atkinson made 3 engines none of which are anything like a 4-stroke with LIVC. The closest is perhaps his two-stroke engine, which kept the exhaust valve closed during the power stroke and open for part of the compression stroke.

The Miller-cycle is much closer to what the Prius uses than anything Atkinson made; it was a traditional 4-stroke with LIVC, but required a super-charger to generate power at low RPMs.

The Prius has a CVT and electric motor, overcoming any issues at low RPMs. Also now that VVT is more common, many engines run with LIVC under certain conditions.

And, of course, Mazda had been selling LIVC engines for years before the Prius appeared. Like all of Mazda's other weird engines they were expensive, bad, and unreliable.

I was gonna say, pretty sure the millenia? Had some sort of Miller cycle

Good memory; Wikipedia agrees with you: "[The Miller cycle] was adapted by Mazda for their KJ-ZEM V6, used in the Millenia sedan, and in their Eunos 800 sedan (Australia) luxury cars[1]

1: https://en.wikipedia.org/wiki/Miller_cycle

If the engines are direct injected rather than port injected, they can just delay fuel injection until the intake valve is closed, and no fuel should be pushed back into the intake manifold.

I have read that modern Toyota engines use both port and direct injection, but am not sure what the Prius does.

You want both to keep the intake valves clean, yes?

What was the purpose of using the rotary engine in the first place, for ICE cars? From the comments it seems it's maybe lighter for a given power output. Is that it, other than just the novelty of it?

It's about double the power output for a given size / displacement. It's very smooth and high-revving so that makes for a really fun driving experience in a light-weight sports car.

Lighter and mechanically simpler were the selling points. Poor fuel efficiency and trouble with the seals were the downsides.

My uncle had an RX-7 back in the 80s when I was a kid. I remember when it was idling I could see the exhaust puffing, in pulses. It only had a single combustion chamber after all.

I've had three RX-7s. A first generation and two second generations. I drove the first until the wheels literally fell off (I was a young, stupid kid that ignored the crunching sounds of the rear wheel bearings falling apart. The rear wheel(s) seized while I was driving down a bridge causing the axel to snap in half. That was an interesting "drive" down the rest of the bridge). The second was a base model second gen that I gave up when I suddenly had two cars because of getting married, then divorced. The third was my favorite. A late model second gen fully loaded. That one eventually caught on fire due to negligence at a tire place. I loved that car.

The third generation never got cheap enough for me to consider one, but oh, I wanted one badly. The RX-8 never really caught me. Plus they had some early issues. That Iconic concept definitely has my attention, though.

>It only had a single combustion chamber after all

Just to be clear: At had two combustion chambers. Both the 12A and 13B engines had two rotors (1.2 and 1.3L "displacement" respectively), each one has a combustion chamber. I had an '84 GSL-SE (13B), I don't remember it particularly "puffing", not saying it didn't just don't remember it.

The engine in this article they say is an "8C", which I assume means 0.8L displacement.

It's mechanically simpler until the apex seals blow out. Then you are fucked

rx-7 had two combustion chambers.

Power density is indeed the main advantage. It's also much easier to balance; hunks of metal changing direction many times per second (6000 RPM is 100Hz) is somewhat mechanically exciting.

> NSU and Mazda are the only manufacturers ever to have mass-produced rotary engine vehicles.

Not quite true. Suzuki also made and sold a rotary engined vehicle--specifically the RE5 motorcycle.

https://en.wikipedia.org/wiki/Suzuki_RE5

One of the cool things about this engine is that it can run on hydrogen with minimal modifications. Not sure if the article mentions this, I don't have time to read it now. But that could make it more interesting when renewable hydrogen becomes common (right now it's not)

Why would you burn hydrogen in a car? It's more efficient to burn it in a power plant and run the car on electricity.

I'm sure there's some niche uses where it makes sense. Fast refueling for example because batteries are so slow. And much heavier than a Wankel engine plus hydrogen tank.

Like in a public transport bus that needs to be on the go all day. In fact I'm pretty sure that's what the hydrogen buses in my city do, they're way too noisy to be fuel cell electric.

Besides, if you can run on electric you can bypass the whole hydrogen step altogether.

Have you looked at hydrogen prices? when the first hydrogen cars came out it was about $17/kg, and I think last time I looked it was around $33. I think that's a LOT more expensive than gasoline per mile.

Meh, market forces can go either way. By the time green hydrogen actually becomes available it might have changed again.

The problem is more finding a less lossy way to create it out of water with electricity.

Also I don't advocate hydrogen as a 'solution'. To me the solution is less cars and nmuch better public transport. I could see it work for buses though.

I think I heard this pitched by Neil Diamond 15 years ago. He had an f body or similar old big body car from the early 80 retrofitted with a rotary engine charging the battery… I’m so glad to hear this finally making production.

I have so many bad memories from this engine. I finally brought them to court to win. But all this to say Mazda has been out of the game so long living on dreams with this engine.

Mine went 110k miles without ever blowing up and then I decided to rebuild it.

It's got as much longevity as other cars from the era if you pay attention to oil/omp and (over)boost (if you have a turbo). Everyone puts in at least a downpipe and then neglects all of the other mods (wastegate mod, better fuel bump, bigger injectors, wideband o2 sensor and new fuel map, better intercooler, etc) that you'll be forced to do along the way or suffer from boost creep and an eventual grenaded engine.

It's all of the other things under the hood that I've had to replace from all that heat though. Three alternators. A wiring harness that caught fire...

The mid-90s RX-7 was sexy as hell.

But the RX-8 just wasn't.

A cool engine still needs to be wrapped in a car that people want to drive.

The MX-30 doesn't quite fit the bill... just seems so mundane to look at.

Can you buy it as an EV and then drop in the generator later? It makes a big difference for registration in some places.

Instead of a Wankel, why not use a small custom turbogenerator, like the APU in aircraft?

They're efficient at higher power levels but really not at small-power-generator level. An 8kw jet powered generator I saw recently used 0.5kw-hr per kg, which is 5x more than an equivalent one based on a traditional diesel engine. Their main advantage is high power for low mass of generator, but consume massive amount of fuels.

Seems unreasonable for a powertrain, but pretty reasonable for a series hybrid generator.

Not really because the power rating at which they start to make sense is in the multiple hundreds of kW. Cars don't need that much average continuous power while driving, on the order of 15-30kW depending on speed and grade. So a "more efficient than piston" turbine engine would be way too big and powerful and would cycle about once per hour. Overall excess mass and bad reliability due to frequent cycling. It's better to have a smaller engine that's sized for slightly over 15-30kW continuous duty. Realistically this will be a piston engine, or a fuel cell. I would love to use a turbine engine for my own application but the math doesn't work, unfortunately.

> Despite all this, though, the MX-30 R-EV still falls somewhat short, with a catalog fuel efficiency of 15.4 kilometers per liter (as measured in WLTC mode). With no similar vehicles on the market, like-for-like comparison is not possible, although the Toyota RAV4 plug-in hybrid gets 22.2 km/l, while the Prius plug-in hybrid gets 26.0 km/l. The MX-30 R-EV is clearly inferior in terms of its fuel economy.

That's because the engine serves an entirely different purpose. The Prius plug-in hybrid is a 13.6kWh battery mated to a regular engine. The MX-30 R-EV is a 17.8 kWh battery mated to a tiny ultralight emergency range extender.

As to similar designed cars, um, hello, the BMW I3 Hybrid? Exactly the same design.

The failing of the MX-30 R-EV is that its battery size is pathetically small for what amounts to an EV with a range extender. It is an embarrassment.

I wonder how many people drive far enough to regularly need the range extender before they have a chance to plug it in. The fuel economy of this thing is not great if you drive it far enough that it runs out of petrol and you need to fill it up again - but what if you rarely do that? What if you usually drive less than about 50 miles before the next opportunity to recharge it?

Is the rotary better or worst with pollution ?

The long combustion chamber, according to an ex-petrolium engineer co-worker of mine, is inherently less efficient because the charge along the walls of the combustion chamber doesn't burn as well.

A Wankel with a turbo at redline sounds you just violently shook a big metal can full of hornets.

Reply on news.ycombinator.com