Hacker News Reader

Cars? Probably not.

Drones? Military drones or aircraft? (The paper mentions ARL) Quite possibly. What this promises is that anywhere you would have a strut or panel made from carbon fiber as a load-bearing element, you can have it store some energy for you as well.

> Cars? Probably not.

Why not cars?

Because structure makes up a small percent of an electric car mass and batteries make up a big percent of electric car mass.

Let's say that it's 10% for structure and 30% for batteries.

If you decrease energy density of batteries several times now they take up let's say 90% of the car mass and the only thing you saved was 10% :)

10% for structure might be an underestimate. I did a Google and it suggested 25% for suspension and chassis and another 25% for body. Obviously that includes more than what you could make a battery, but I reckon it’s more than 10%.

Also, even if it’s only 10% weight you are saving, that sounds pretty good to me!

When the battery needs replacement, what do you do?

You toss the car in a landfill and stamp out a new one in the factory.

So probably better for drones and things that have a short lifetime anyway.

Drones not likely. Their improved stiffness of 75 gpa is 2x less than carbon fiber.

So you could literally make the frame 2x lighter with carbon fiber and put that 50% of weight in batteries getting you 2x+ more capacity.

Cars more likely as it's roughly on par with aluminum. Still the cost will be way more than plain aluminum or even steel body.

Doesn't the carbon fibre argument work for cars as well? CF is becoming increasingly common and affordable, as are normal batteries.

Some supercars have a carbon frame. But it's extremely expensive. Check maclaren cars for example

BMW i3 has a carbon body, and it's only a little (twice?) more expensive than a normal car. A lot of inexpensive carbon fiber bicycles are available now, while previously you could only get them for big money from big brands. Carbon fiber has gotten cheaper, we're going to see it in a lot of products.

Hopefully it'll be easy to repair. I'd hate to replace entire chassis because of light collision.

If your fender bender short-circuited your battery and ignited it, you might need a wee bit more than just a new chassis.

(This is not a claim that the batteries described in the article are a fire hazard - I don't know)

A mechanic working on a car, with enough juice to fry him, and no way to separate the battery from car?

I guess a full discharge first.

But what about your accident, and the jaws of life? Will they conduct? Will the wrenched apart car have conductive edges?

Interestingly, these are actual concerns that firefighters have regarding both Teslas and self-driving cars. The relevant companies put out explainers and training videos on how to do a rescue in these vehicles.

In the Teslas, the power system is shunted through a single disconnect point... Cutting that line isolates the battery from the rest of the vehicle.

In Waymo cars, firefighters needed to be made aware that high-power electrical conduit and liquid cooling channels go up through the pillars, which are usually empty of energized components or contain only low-power channels (for things like overhead lights).

That's interesting. I never considered this to be a problem but it makes sense. This could be quite challenging if every car manufacturer does something different, but I suspect over time just like with ICE cars they'll iterate towards some local optimum and things start to be very similar.

>I suspect over time just like with ICE cars they'll iterate towards some local optimum and things start to be very similar.

This doesn't happen organically. Manufacturers were happily iterating toward a different optimum before regulation forced their hands.

https://www.nytimes.com/2015/11/27/automobiles/50-years-ago-...

That's not really what I mean though. I'm well aware that without regulations short term incentives and optimization can lead to bad outcomes. What I mean is something along the lines convergent evolution where certain designs keep re-appearing because they just work better than the rest.

That's pretty much the status quo for EVs now; the batteries are hundreds of volts and you can't easily remove it. Draining it isn't an option either, as it damages the battery to be fully discharged. So, you have to work on the car with a live voltage source present.

I'm not sure if most modern EVs have manual disconnect switches on the main power cables where they meet the batteries or if you're just expected to wear rubber gloves and use non-conducting wrenches to unbolt the cables. I'd expect any modern EV will at least have contactors to connect the battery to the motor controller (these are basically electrically actuated switches, like a mechanical relay), which will be disabled when the car is off. In general, though, I'd expect the main power cables to be something that doesn't need maintenance and mechanics can generally steer clear of.

In the article, they say they're using a lithium iron phosphate chemistry. Those tend to be pretty safe. Or at least as safe as you can reasonably expect anything to be that can dump hundreds of amps across a closed circuit.

My thoughts exactly. Load bearing structures tend to be permanent or not frequently changed (on the order of decades). Now someone wants to place a battery in that structure?

I hope they develop a mechanism that allows exchange of the battery and can bear load without the battery with some sort of mechanical movement that can shift the load to a more permanent structure (not that this is necessarily a good idea either but proposing an option). Or, perhaps the battery never needs changed, and by that I mean the life of the product lasts as long as it would with a replaceable battery. They can't just redefine the life of the product as the life of the battery to get around this (unless the life if the product was already defined by the life of the battery for a long time before).

How about the structural problems with composites that occur with use ... Fractures from normal wear and tear (over years) in a composites will lead to catastrophic failure of the part

Repairability is guaranteed to be impossible if you have that level of integration, and I suspect that is a big part of the motivation behind adopting this.

Indeed. And having worked with a wide variety of batteries.... I would hate to see a 'load bearing battery under torsional stress'. Just ow.

Explosion?

If you damage carbon fiber you generally won’t be repairing it. This doesn’t sound like it will be a very big limitation for similar use cases.

carbon fiber repair is definitely a thing. i’ve had a few carbon bike frames break and had them repaired without any issues. there’s a few places that specialize in this

Many ultramodern carbon fiber processes don't use an autoclave so one might be able to do production grade repairs in-place.

I would like a very simple electric car, simple motor(s), simple electronics to repair, no GPS or phone home necessary (I want to download on my time, send me a text/email/whatever about critical downloads). I want something easy and cheap to repair, have the simplicity of electric motors, and run 20 years with very little maintenance. I don't 30 computers running on my computer. I just want to get from A to B and it not cost me 80-100k for a decent electric vehicle. Tesla and others are overengineered for what they do and that complexity adds to failure rate and cost of ownership.

Aren't you describing a Nissan Leaf? I can't really speak to the repairability, because even though I've had it 2 years, I haven't had to do anything but charge it. Oh, once I had to put more air in the tires. Great little car, with about a 70 mile range on my 2017 model.

You're not going to get around the 5-10 computers required for basic 21st century automotive safety. But, the rest of what you are describing is the status-quo for quite a few Chinese 'budget' brands.

I'm betting once EVs start making their way into fleets, you'll find your 300k mile, low maintenance, minimal electronics American/German brand.

One way to get what you're asking for is to do a conversion (or pay someone else to do it, or buy one that someone is selling).

I'm in the middle of converting a Mazda RX-8. Just got the motor/clutch/transmission assembly (mostly) in place last weekend. It's kind of fun, but it's also a lot of work. It seems like a tremendously inefficient way to get the car I want: electric, no phone-home or over-the-air update capability, ability to control how everything works, easy to fix, manual transmission, modern safety features. I thought about buying a Model 3, but decided it was a lot of money for something that wasn't quite what I want. 200+ mile range would have been nice though; it's hard to get that in a conversion without making it unreasonably heavy.

Why do you want an electric car with a transmission? It doesn't have an engine in the first place.

Partly because shifting gears is fun. Partly because you can get more usable power out of the motor with a transmission.

It's a bit of a trade-off. To use Tesla as an example, they take the approach of just putting a massively overpowered motor in each car so that they can overcome the lack of a low gear. That works fine, but the kind of parts you need in that sort of system are expensive. Not just the motor, but the motor controller and the batteries have to handle that much power.

A different approach is to just have a normally-sized motor and an ability to shift gears. It should make for a lighter, more efficient car. (Though you also lose some efficiency in the gearbox.) The conversion I'm doing has a 120 horsepower motor, and it's going into a car that's about 3,000 pounds. I won't win any 0-60 races with a Model S, but it'll probably handle corners better and generally be more fun to drive, not to mention be about 1500 to 2000 pounds lighter.

I'm willing to concede that wanting an EV with a stickshift may be an unpopular opinion, but it's hard to know if there's a market for that kind of thing until someone starts manufacturing it. (And I don't mean the Taycan, I mean something that some ordinary middle-class person might buy.) It seems kind of a shame that we have all these legacy car companies that know how to build transmissions, but none of them seem interested in using that as an advantage they have over Tesla for that part of the market that want an electric car but would rather drive a manual transmission.

I suppose back in the day there were those who missed the buggy whip.

>it'll probably handle corners better and generally be more fun to drive

Sounds like a very educational project.

I wouldn't be so quick to just brush off Tesla engineering as having bad tradeoffs, or make assumptions about the fun of cornering, but I'm sure the project will be an eye opener and maybe will lead you to test drive an M3P sometime.

Definitely it's hard to beat an RX8 with cornering but that's before you load it down with batteries.

Edit: I had some doubts about your weight claims but I wasn't thinking about the fact that you are losing the engine weight, which puts you pretty far ahead! Sounds like a cool project.

I don't think Tesla's engineering tradeoffs are bad, I just think that they result in a car that's different than my notion of an ideal car. It's a preference thing more than anything else.

I did actually consider buying a Model 3, and even put down a thousand dollars to preorder back when it was announced. I test drove a dual-motor version (I would have gotten the single-motor version, but it's what they had). The acceleration is fun. Otherwise it seemed like a solidly built yet kind of generic car. In the end, the main things that kept me from buying it were Tesla's secretive nature when it comes to service information and the prospect of some minor component failing and bricking the car (like that flash chip soldered to a giant motherboard that had issues in early models) after the warranty expires, and the price being a bit more than I cared to spend on a car. Awhile later a friend showed off a Ford Ranger pickup truck he'd converted with an electric motor and I thought: hmm, this isn't actually as expensive or difficult as I thought, maybe I should do a conversion...

The weight breakdown is something like this: the RX-8 is about 3,000 pounds stock. The rotary engine is about 300 pounds with all its accessories attached. The dual-shaft version of the Netgain Hyper9 AC motor I'm using is 130 pounds. The exhaust system is maybe a hundred pounds or so removed. The gas tank is another hundred pounds or so (when full). I'm adding about 450 pounds of lithium iron phosphate batteries, which gets me about 27kwh and maybe 100 miles of range if I'm lucky. (Used Tesla cells have a lot better energy density and are actually reasonably priced, but they wouldn't have fit very well in the places I wanted to put them. They're also a bit more dangerous than LFP cells and require liquid cooling.) I'm also adding in some weight for battery boxes and the motor mount and motor/transmission adapter. In the end I figure the car'll be about 200 pounds heavier or so, and most of that weight is very close to the ground.

Thanks for all that detail! I hope you’re writing this up somewhere sometime where more people will see it :-). Especially after it’s done it would be cool to read about how it is, how it went, the battle scars and or tips and what you would do next time.

I haven't posted anything publicly yet, but it's on my to-do list.

I'm not understanding, perhaps you can enlighten me. The article says "The battery has an energy density of 24 Wh/kg, meaning approximately 20 percent capacity compared to comparable lithium-ion batteries currently available." But the article goes on to say, "since the weight of the vehicles can be greatly reduced." If you need 5x the weight of this kind of battery than if you used Li Ion battery, and a Li Ion battery constitutes 1/3 the mass of a car (at a guess), how is there any savings? Seems like the result will be even heavier than current EVs.

Leads to a lot of potential if the support structure is also the battery. Even as you said, if you have to still have more of it, you don't add weight to store a charge, and if you already need a lot of structure, you're fine.

Having said that, the car chasis is not juuuust about holding up a roof and the humans inside. You also have to handle impact stresses, you don't want it discharging on the poor inhabitants as the car is being hit, etc.

This sounds like a step, out of many, the the direction of lighter cars or aircraft that are electrically powered. It also doesn't have to be the only tech. If we can eliminate say 20% of the battery weight by making the frunk/trunk casings out of this material, some other inner parts of the car, we can use it as part of the solution.

I'm wondering if these batteries will maintain their capacity over time. I think part of the appeal of traditional battery packs is you can replace them as they wear out and hold less charge. If they do degrade, hopefully a chassis/ battery replacement won't be too expensive.

I've seen estimates that put modern battery at the expected age/mileages comparable to ICE expected lifetimes - at 300-500k miles you usually replace the car anyway, especially modern ones with so much electronics inside, those will probably get outdated way faster

> You can't make a car by building the chassis out of smartphone batteries

They're called Structural batteries (or [micro]structural super/ultracapacitors)

"Carmakers want to ditch battery packs, use auto bodies for energy storage" (2020,) https://arstechnica.com/cars/2020/11/carmakers-want-to-ditch...

This is literally what TFA is about

The Ars article I linked has an overview and some history and specific industry applications; whereas OT is about a new approach discovered since the Ars article was written.

You comment is more useful than the article, just so you know.

Note that Tesla Cybertruck is using battery pack as a structural element.

The battery pack yes. The batteries? I would guess not.

It’s a step in that direction others haven’t made yet.

Having the chassis of the car be the battery seems problematic. Once the battery is worn out the whole car must be disposed of.

Even more interesting for aviation industry.

What about bicycles?