Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[T.S. Zarathustra]

Having less wasted space will make the cells more structural, especially as they're compressed.  Right now they use big I-beams and the cells will instead be packed tightly with cooling rails atop or below them.  All that means they can mount them in something that itself is rigid instead of gooey or soft like the material they use now.

Basically, no ribbons, plus rigid mounting medium, means no I-beams, which means even more space saved.

And Tesla is looking at sheathing them in even less material!  Denser means more structural.  The cylinders already expanded less than pouches!

-Crissa

How are they cells compressed? So the cells will be "packed tightly with cooling rails atop or below". Mounted in "something that itself is rigid". That sounds like regular standard batterypacks, i.e. loose cells in structural container. The cooling rails will simply take up space atop or below the cells, instead of between them, unless the cooling rails will be structural part of the cell container. I guess we'll see what the Tesla geniuses have designed when it actually reaches the market.
Cylinders expand less than pouches because the expansion is inside the cylinders. The cylinders have to be slightly bigger, little bit of wasted space, to allow for the expansion of the materials inside.
I still call phooey on the claims of 5-6 times improvement, simply because they're not comparing same size cells. I'd say the cell changes are at most 10% improvement (probably closer to 5%) for same volume. And they could have continued using the 18650 or 21700 cells with the cooling rails moved "atop or below" the cells "packed tightly" and get similar improvements. And they could have made the no ribbon improvement to the 18650 or 21700 cells. I think Alans suggestion of them simply using fewer cells to drive down costs is the main reason for the new cells. Battery technology these days is big investment for small gains.

[Alan Stewart]

And they could have made the no ribbon improvement to the 18650 or 21700 cells.

I think the removal of the cooling ribbon is made possible by the tabless design. In addition to reducing resistance it conducts heat from the interior of the cell to the end caps, so cooling via top and bottom plates becomes effective. As I said earlier it seems like the tabless design could be applied to other (cylindrical) cell formats such as 18650 and 2170 to give the same benefits as it does to the 4680 format.

[Crissa]

How are they cells compressed?

In between the cooling rails?

So the cells will be "packed tightly with cooling rails atop or below". Mounted in "something that itself is rigid". That sounds like regular standard batterypacks, i.e. loose cells in structural container.

Yes and no.  Tesla is trying to get approval for a safety medium which is also rigid.  So right now the box itself is the structure, but they want something which is laminated together where the cylinders are part of the structure instead of contained in it.  I don't know if that means they're countervailing forces or just glued in place; but their anti-fire resin itself is supposed to be rigid structure.  Which is wholly unlike how Zero or Energica cells are potted.

I still call phooey on the claims of 5-6 times improvement, simply because they're not comparing same size cells.

Per cell, which is how fast they come out of the factory, their numbers hold.  But yes, that's nearly all from just 'being bigger'.  But no one could make bigger cylindrical cells and have them perform as well until now.

Bigger cells means they can make more battery capacity in the same per-unit time.  Bigger cells means they can try their 'structural medium'.  Dry cells means smaller factory, so they can fit more production lines in the same place.  Bigger, in their case, is better.  And the tabless allows that to happen.

-Crissa

[T.S. Zarathustra]

My English comprehension seems to be failing me. 
But to the part I understood.

But no one could make bigger cylindrical cells and have them perform as well until now.

Bigger cells have been made, maybe not cylindrical, but there are huge Lithium based cells available. Once they get up to this size the fact that they're cylindrical starts to hold them back. IMHO, they'd make bigger steps by switching to pouch cells. Don't forget that the original 18650 were only used because they were readily available for "low prices" in high quantities.
There are many performance parameters, which one you think is most important is up to you.
Energy density, Operating temperature, Safety, Durability, Charging time, Output power, Elimination of cobalt. These cells

[Richard230]

The cells used in my 2008 Electric Motorsport GPR-S were made by Hi Power (not exactly the right name for that company  ) and were about the size and shape of a home Webster Dictionary. Their claim to fame were that they couldn't handle power draws of over about 50 amps very well and when used to power an electric motorcycle sometime tended to expand like a balloon and vent gas at inopportune moments.    I suspect they were designed more for storing electricity to power decorative lights on Asian buses than they were for EV motivation. 

[Crissa]

There's also how fast the cells can be manufactured, how cheap it is to manufacture, how much resources it takes, etc.

That's why Tesla went with cylindrical cells:  They are cheap and fast to manufacture.  But manufacturing machines can only run so fast, so at that point, to speed up you need more factory or bigger cells to have more capacity.

Bigger cells were not reliable in cylinders because of the tabs.  Hense the new design.

-Crissa

[T.S. Zarathustra]

Congratulations on 2000 posts Crissa.

Here is a nice youtube video at the EEV blog about the 4680 cell. A guy who actually knows a little bit about electronics and is mildly impressed.

Edit. Real explanations instead of marketing lies speech. If I'd found this video in the beginning I would have linked to it. It would have saved me few posts. 

[Richard230]

That certainly was an interesting video, Zarathustra. Thanks for posting it.  My feeling has always been the the real advantage for Tesla by using the cylindrical cells is the ease of quickly manufacturing millions of them using robotic machines, compared with manufacturing pouch cells. Any other advantages are just gravy.    When you are planning to sell millions of electric vehicles a year, money becomes a big deal, especially when government subsidies are shrinking and in their case, disappearing. 

[Alan Stewart]

Here is a nice youtube video at the EEV blog about the 4680 cell. A guy who actually knows a little bit about electronics and is mildly impressed.

If the ideas behind the 4680 cell format are so obvious why has it taken a number of decades for anyone to implement them? Many innovations are obvious after the fact, physics is just physics. Brilliance is in seeing how to take best advantage of physics, and in the case of commercial products, costs also. The largest barrier to EV adoption is cost.

This guy is not considering the financial and manufacturing aspects of the 4680, which is the primary reason for the design. 5x energy implies 1/5 as many cells are required to make a battery pack of a given capacity, a great reduction in manufacturing and quality control operations and parts counts and thus, costs. There are also possible benefits in reliability. 6x power is significant because a battery pack of a given capacity will be capable of 20% more power, which also has implications for potential charge rates and charging times. The 16% range increase is a result of weight savings for a battery pack of a given capacity, so smaller battery pack can be used to achieve the same range goal, again reducing vehicle cost.

The primary commercial and practical significance of a battery cell design is the battery pack design it enables.

[Alan Stewart]

The Tesla Battery Day presentation is focused is on how to scale up battery production and decrease battery costs, not on the 4680 cell design or EV range.

https://tesla-share.thron.com/content/?id=96ea71cf-8fda-4648-a62c-753af436c3b6&pkey=S1dbei4

[Crissa]

If the ideas behind the 4680 cell format are so obvious why has it taken a number of decades for anyone to implement them?

Well...

The spending to many the continuous or tabless cell just didn't make economic sense to small cells, which is where the vast majority of cylindrical cells are used.  It wasn't just the tabless idea, you also need more precise cell manufacture.  Usually saving money in manufacture is about adapting to a less precise but cheaper method.

For larger cells, it made sense to just make larger cells instead of slowly ramping up.  Remember, pretty much only Tesla decided to use cylindrical cells as a good atomic unit for larger batteries.  This is because just building a large battery was sort-of cheaper than making many small ones.  But small ones can be made fast and cheap... But have you seen how complex Tesla battery packs were?

Obvious doesn't mean 'there is money to do it' because often the inobvious part is 'how much will this save me vs the investment to make it so?'

-Crissa

[T.S. Zarathustra]

This guy is not considering the financial and manufacturing aspects of the 4680, which is the primary reason for the design. 5x energy implies 1/5 as many cells are required to make a battery pack of a given capacity, a great reduction in manufacturing and quality control operations and parts counts and thus, costs. There are also possible benefits in reliability. 6x power is significant because a battery pack of a given capacity will be capable of 20% more power, which also has implications for potential charge rates and charging times. The 16% range increase is a result of weight savings for a battery pack of a given capacity, so smaller battery pack can be used to achieve the same range goal, again reducing vehicle cost.

Please don't fall into the trap of thinking that manufacturing costs will be the same for both cells. The 4680 cell will requires 5-6 times the raw materials of 21700 cell. Mining costs for materials that much higher for each 4680 cell. There will be some savings in machines required to make the cells but will that be a game changer? I still don't see what would be that cost reduction game changer.

[Alan Stewart]

Please don't fall into the trap of thinking that manufacturing costs will be the same for both cells. The 4680 cell will requires 5-6 times the raw materials of 21700 cell. Mining costs for materials that much higher for each 4680 cell. There will be some savings in machines required to make the cells but will that be a game changer? I still don't see what would be that cost reduction game changer.

The cost savings is in decrease in the reduction in the number of cells to make a battery pack of a given capacity. One-fifth as many parts and manufacturing operations saves money. Tesla moved to 2170 for the Model 3/Y to save money, per Elon Musk. It’s battery pack cost that matters in the end, not cell cost.

[Crissa]

Most of the per-cell cost is manufacturing it.  More cells, more cost.  That's why GM went with large prismatic cells, to lower the per-cell cost.

-Crissa

[Alan Stewart]

The new visualization from Tesla shows how they’ll be packing the cells closer together and eliminating the cooling ribbon of current battery packs. Probably the cell bottoms will be fully glued to the cooling plate and the interstices between cells also filled with glue to make a very solid honeycomb-like structural component. Sandy Munro thinks they’ll use manifolds welded to the positive and negative contacts at the top of the cells.

https://insideevs.com/news/456644/tesla-shows-4860-cells-and-pack/