First look at Tesla’s new structural battery pack that will power future vehicles

[TheLastStarfighter]

I find it interesting the concern over battery replacement and or massive accidents. How often does this happen? The warranty on the battery is 7 years. Replacing a big battery now is going to run you 15k or so. With this new design it could bump the cost up...20k? 25? Do you really want to spend 15-25k to fix an 8 year old car? A 10 year old car? If you have a massive accident, you're replacing the battery anyway on the existing style, and also the frame work, doors, etc, so nothing's changed there. When you get to that point it's generally best to be looking at a new vehicle anyway, since your 8 year old Tesla is probably only worth 30k.

Rather than think about extreme situations that are likely never to happen and solutions you are likely to never use, think of the positives. You're probably saving 5-10 in manufacturing costs when you purchase. If you're paranoid, put that into saving or investments to cover that .01% chance something catastrophic happens to your car when it's a decade old.

Personally, I'm pumped for the extreme rigidity that's likely going to come from that design. A Cybertruck could feel like a Miata!

Advertisement

 

[rr6013]

The design is intriguing, but I also have to wonder if this is the beginning of the disposable car.

<Snip>

I wonder what this will do to insurance rates.

Ford gets credit for introducing the concept of disposable car in 1970 with the PINTO.

Battery packs are highly charged volatile liabilities after a BEV is incapacitated. As quickly as auctions cropped up to take insurance companies totaled cars, the free market will find a best solution for used battery packs.

Tesla approach to recycling refects back on its impact on the planet, earth resources and lifecycle BEV foot print. Tesla can’t dump its glue and 4680 trash on the planet because it chose a non-reparable method of structural assembly. That would taint its mission and its image.

 

[ScoobyDoo]

Ford gets credit for introducing the concept of disposable car in 1970 with the PINTO.

Battery packs are highly charged volatile liabilities after a BEV is incapacitated. As quickly as auctions cropped up to take insurance companies totaled cars, the free market will find a best solution for used battery packs.

Tesla approach to recycling refects back on its impact on the planet, earth resources and lifecycle BEV foot print. Tesla can’t dump its glue and 4680 trash on the planet because it chose a non-reparable method of structural assembly. That would taint its mission and its image.

We all know SpaceX Starship is really for launching old batteries at the sun. ? ? ?

 

[Diehard]

Tesla can’t dump its glue and 4680 trash on the planet because it chose a non-reparable method of structural assembly. That would taint its mission and its image.

First you make a company to serve a mission. Once your baby grows, you have employees, shareholders and a balance sheet, your company takes life of it’s own and before you know it, your mission is serving your company.

walking is the most green mode of transportation, yet Prius is considered to be green because it pollutes less than Armada. Now that everyone is getting into EV game, all Tesla has to do is to be slightly greener than everyone else and slightly greener than it used to be. Frankly, I think Tesla got off the ground due to it’s SPECS and performance more than the green mission. And will stay in business for the same reason. Green mission helped with the starbucks factor (people feeling good about paying more) but once EV usability and cost can stand on it’s own, no one will have to be green to sell them. When Tesla starts putting the actual purchase price of the car as default on their website, that is when you know we are there.

 

[Crissa]

This looks far more repairable than current battery technology. Sure, you won't be getting your battery pack back in the case of failure: But having the rigid structure may mean remanufacturing is much easier.

Elon did say the cells were going into a rigid structure that needed permission to use. This looks like that.

-Crissa

PS, cars have been disposable for the entire time. It's not something new, and it's not like cars today or yesterday are different in that regard.

 

[Diehard]

Elon did say the cells were going into a rigid structure that needed permission to use.

Can you expand on that? Does that mean owner need Tesla’s permission to open, use, resell, reuse his battery?

 

[Crissa]

Can you expand on that?

The grid has to be made of a material which inhibits fire spread in case of cell temperature runaway. That's part of the purpose of the current foam goo.

The rigid material had not yet been approved for use in vehicles yet.

-Crissa

 

[Luke42]

PS, cars have been disposable for the entire time. It's not something new, and it's not like cars today or yesterday are different in that regard.

I was upset about planned obsolescence about 20 years ago, but then I applied what I was learning in engineering school to the concept.

In a universe which features entropy, no engineered device will last forever. The question is how long it lasts, and what we do about it.

Every fully engineered item has an expected lifetime, whether it's skyscraper or a disposable drinking straw. The question is what that planned lifetime is, and how it's handled. Is it easy to replace?

The old joke about "this is my grandfather's hammer, I've replaced the handle 8 times and the head 3 times" applies. This is a great way to handle this on a hammer (the handle/head interface is simple and standardized), but it's a Hard Problem on complex devices made from multiple interlocking assemblies.

Most modern cars seem to have an expected lifetime of at least 250k-miles, which is around 5x better than it was the year I was born (several decades ago).

So, with that in mind, all of the raging against "planned obsolescence" comes down to the question of whether it makes more sense to replace components, assemblies, or entire vehicles -- because nothing will last forever. When the question of "planned obsolescence" acknowledges these things, the answer changes dramatically.

Also, the answers can be counterintuitive.

Let's take 2004 Toyota Prius as an example. Toyota makes long-lasting cars, which appear to be less about "planned obsolescence" than the competition. However, if you've ever worked on a Toyota, their service procedures strongly favor replacing whole assemblies. For instance, when the spring in the Prius joystick-shifter failed, the Toyota procedure was to replace the entire shifter -- rather than just the spring. This is a wasteful form of planned obsolescence in one sense (most of the joystick-shifter assembly was unnecessarily discarded), but the joystick-shifter lasted longer than a hand-rebuilt version, was easier to install (it had alignment pegs), and kept the entire car on the road with minimal effort. We owned that car for 12 years.

On the other hand, the 2001 Volkswagen Jetta TDI I owned encouraged the owner-mechanic to replace the smallest possible unit, which seems at first glance like a way to avoid "planned obsolescence". However, you had to be a craftsman in order to do anything on that car. Commonly replaced parts (like the oil filter) were difficult to reach, and pretty much everything with a bolt required you to manually align it when you put it back together (while the Toyota had alignment pegs, ensuring that even a trained monkey can do the job right). The maintenance costs killed the Volkswagen for me, and I only owned it for 1 year.

At first glance, it would appear that Volkswagen was designed to avoid "planned obsolescence" more rigorously than Toyota, because you throw away less on each repair.

However, the Toyota was a *much* better car to own than the Volkswagen, because things stayed fixed when you replaced the big-f@cking assemblies with lots of non-broken parts in them. The assembly-replacement strategy Toyota uses is fully engineered, and ensures quality over the planned service-life of the vehicle. The Toyota lasted much longer. So, yeah, total lifecycle engineering is hard, and the answers can be counterintuitive.

Now, let's extend that kind of thinking to the whole car. If you replace the whole car, the problems will stay fixed -- just like when you replace an assembly. It's also easier to ensure that the front-line service technician does the job right, when the job is merely to unload the car and hand the keys to the customer. The million-dollar question then becomes when it makes more sense to repair (rather than replace the entire car), both in economic and environmental terms. Keep in mind that old cars aren't thrown away; they're parted out and recycled by junkyards.

As an engineer, I appreciate that the devil is in the details -- and that this is a hard problem. "Disposable" cars often last longer than cars expected to last "forever", and "disposable" cars can sometimes be more easily maintained throughout their planned service-life. Also, there's no way to get away from a planned service-life, because entropy occurs in our universe.

So, what do you do?

My answer is to just keep increasing the planned service-life, and to have a plan to recycle the components of the car when the end-of-life is reached. Coincidentally, this seems to be the approach Tesla is taking. But it's still "planned obsolescence", even if it's objectively better engineering (all things considered) than what we were doing before.

 

[Riptide]

The design is intriguing, but I also have to wonder if this is the beginning of the disposable car.

With most vehicles today there can be some reconstruction of the frame if something should become damaged in an accident. With the battery pack (arguably the most expensive part) exposed to said damage even a relatively minor accident would total the vehicle. Compound this with the one piece casing of the front and rear sections and it would be even worse for repair costs if you had to replace the entire assembly.

Ill give you an example, side impact in front of the rear wheel at 15 mph, on most vehilce replace the door and straighten the frame paint and repair is complete. On these if the battery pack is tweaked its replace it, unlikely it would be repaired. same with the one piece casting for the rear section.

I wonder what this will do to insurance rates.

We are already at the "disposable" car. Insurance companies are very quick to declare a car a total loss even after a seemingly minor accident, then the cars are parted out for the pieces that can be resold and the rest is recycled. Any accident which causes structural damage to a Tesla battery pack is going to be a total loss, as it's just as protected as the passenger compartment.

By the way, the only cars that have a "frame" are trucks, vans, Jeep Wranglers, G-Wagens, etc. Most cars have a unibody design, and yes it can be "straightened", but if the unibody is seriously damaged the vehicle is probably going to be a total loss, except for the newest, highest-value, vehicles.

My old 1999 Mercedes ML320 had a body-on-frame design - it was classified as a truck. Loved that thing, it could go through 22" of water.

 

[rlhamil]

I was upset about planned obsolescence about 20 years ago, but then I applied what I was learning in engineering school to the concept.

In a universe which features entropy, no engineered device will last forever. The question is how long it lasts, and what we do about it.

Every fully engineered item has an expected lifetime, whether it's skyscraper or a disposable drinking straw. The question is what that planned lifetime is, and how it's handled. Is it easy to replace?

The entire universe may well have a finite lifespan. But for something that would in any event be physical feature-obsolete (well beyond just upgrading the CPU/system board) in probably less than a lifetime, a century isn't unreasonable or unachievable (probably longer for large buildings; doubtless the Empire State Building or Eiffel Tower aren't the most energy efficient or conveniently laid out, but I haven't heard any suggestions to tear either down). The Great Pyramid has done ok in terms of enduring, might have done better if it hadn't been scavenged for the coating.

Sci-fi authors have suggested multiple ways in which more or less physical objects could (given energy and perhaps replenished raw materials stores to compensate for wear) to power their sustenance/maintenance/renewal) exist indefinitely. Some of those may be achievable within a few centuries, perhaps even a few decades (the most plausible - no "tractor beams" needed - being nano-machines either composing or constantly maintaining the larger device, and replicating to renew themselves given energy and suitable materials; or almost equivalently, biological machines). Even those have a sort of lifespan due to deterioration of the information necessary to keep their maintenance uncorrupted; but there are error correcting codes MUCH better and more robust than human genes, so the lifespan could be millennia if not more.

So IMO the better questions are when the break-even on initial vs lifecycle cost arrives, whether the initial cost is prohibitive to most, and whether, depending on the likelihood of sufficiently desirable future improvements exceeding the likely scope of mere maintenance, it's worthwhile to create something with an extreme lifespan. Still, even with a shorter lifespan, it might be reasonable to aim for both near-total recycling and mostly predictable failure modes, that combination offering many of the same benefits as long lifespan. And perhaps for migrating such associated data (media, preferences, activity records, etc) as might be worth preserving.

 

[rr6013]

I was upset about planned obsolescence about 20 years ago, but then I applied what I was learning in engineering school to the concept.

In a universe which features entropy, no engineered device will last forever. The question is how long it lasts, and what we do about it.

Every fully engineered item has an expected lifetime, whether it's skyscraper or a disposable drinking straw. The question is what that planned lifetime is, and how it's handled. Is it easy to replace?

The old joke about "this is my grandfather's hammer, I've replaced the handle 8 times and the head 3 times" applies. This is a great way to handle this on a hammer (the handle/head interface is simple and standardized), but it's a Hard Problem on complex devices made from multiple interlocking assemblies.

Most modern cars seem to have an expected lifetime of at least 250k-miles, which is around 5x better than it was the year I was born (several decades ago).

So, with that in mind, all of the raging against "planned obsolescence" comes down to the question of whether it makes more sense to replace components, assemblies, or entire vehicles -- because nothing will last forever. When the question of "planned obsolescence" acknowledges these things, the answer changes dramatically.

Also, the answers can be counterintuitive.

Let's take 2004 Toyota Prius as an example. Toyota makes long-lasting cars, which appear to be less about "planned obsolescence" than the competition. However, if you've ever worked on a Toyota, their service procedures strongly favor replacing whole assemblies. For instance, when the spring in the Prius joystick-shifter failed, the Toyota procedure was to replace the entire shifter -- rather than just the spring. This is a wasteful form of planned obsolescence in one sense (most of the joystick-shifter assembly was unnecessarily discarded), but the joystick-shifter lasted longer than a hand-rebuilt version, was easier to install (it had alignment pegs), and kept the entire car on the road with minimal effort. We owned that car for 12 years.

On the other hand, the 2001 Volkswagen Jetta TDI I owned encouraged the owner-mechanic to replace the smallest possible unit, which seems at first glance like a way to avoid "planned obsolescence". However, you had to be a craftsman in order to do anything on that car. Commonly replaced parts (like the oil filter) were difficult to reach, and pretty much everything with a bolt required you to manually align it when you put it back together (while the Toyota had alignment pegs, ensuring that even a trained monkey can do the job right). The maintenance costs killed the Volkswagen for me, and I only owned it for 1 year.

At first glance, it would appear that Volkswagen was designed to avoid "planned obsolescence" more rigorously than Toyota, because you throw away less on each repair.

However, the Toyota was a *much* better car to own than the Volkswagen, because things stayed fixed when you replaced the big-f@cking assemblies with lots of non-broken parts in them. The assembly-replacement strategy Toyota uses is fully engineered, and ensures quality over the planned service-life of the vehicle. The Toyota lasted much longer. So, yeah, total lifecycle engineering is hard, and the answers can be counterintuitive.

Now, let's extend that kind of thinking to the whole car. If you replace the whole car, the problems will stay fixed -- just like when you replace an assembly. It's also easier to ensure that the front-line service technician does the job right, when the job is merely to unload the car and hand the keys to the customer. The million-dollar question then becomes when it makes more sense to repair (rather than replace the entire car), both in economic and environmental terms. Keep in mind that old cars aren't thrown away; they're parted out and recycled by junkyards.

As an engineer, I appreciate that the devil is in the details -- and that this is a hard problem. "Disposable" cars often last longer than cars expected to last "forever", and "disposable" cars can sometimes be more easily maintained throughout their planned service-life. Also, there's no way to get away from a planned service-life, because entropy occurs in our universe.

So, what do you do?

My answer is to just keep increasing the planned service-life, and to have a plan to recycle the components of the car when the end-of-life is reached. Coincidentally, this seems to be the approach Tesla is taking. But it's still "planned obsolescence", even if it's objectively better engineering (all things considered) than what we were doing before.

“Junkyard” veteran parts guy. My 1997 Tahoe was GM‘s last year that supported parts. 1998 model year only assemblies are available. Which means in addition to the engineering advantage in assemblies over parts GM reduced its “quantities” on-hand inventory, parts investment and increased revenue per transaction.

My 1997 Toyota 4Runner was infinitely simpler for every reason mentioned in your article. In addition the 4Runner was better offroad capable by design. I ended up appreciating Toyota engineering. But chose the Chevrolet. Bigger is better in the USA. Obsolescence is expensive either way.

After fully restoring both vehicles there wasn’t $1000 diff between them. The Toyota had required fewer parts and less labor. The Chevrolet tons of labor, lots of little parts and multiple iterations over a “problem chain” to resolve. Toyota was one and done and expensive. Toyota was “ like new” thereafter. The Chevrolet was just “ fixed”. It could fail in the same problem domain for a different link in the chain requiring a different part.

Assemblies are better all around for everyone.

 

[ldjessee]

To touch on the battery if a cell o several go bad, it is handled like flash memory, spinning harddrives, and many other components that have small bits that go bad, but you do not want the whole thing to fail... over provisioning.

Harddrives & flash memory are larger than they claim, keeping 'spares' to swap out when a sector or region goes bad. Many companies reserve a section of the battery pack to not be used in normal circumstances.

These can be the over provisioned for when a cell or two fails... assuming there is the ability to switch it so that cell is no longer in use... which I am guessing Tesla has to do now, as now, if this was not the case and a single cell went bad, it would require the battery pack to be replaced (or maybe a module for those that have modules).

Sure, there are other reasons to over provision the battery pack, such as it is never truly charged to 100% nor ever truly discharged. And for some chemistries, this is a huge advantage.

 

[Crissa]

Technically, drives are usually not over-provisioned but lose space when formatted.

-Crissa