actually I suspect calendaring would be EASIER on a slightly thicker electrode, and my reasoning is that it would be less prone to damage from slightly uneven grain size by increasing roller distance. Where as a 5micron particle might damage a roller or film lamination with a 5micron lamination thickness and you avg large particle is 5micron, if you increase the coating to 7micron, you clearance allowance saves the rollers. In particular with the dry coating process I bet it's like many adhesion problems, to thin is harder to do than slightly thicker.
Another wonderful video, thank you for that Jordan. I think another thing to be thankful for is not becoming a hype machine like some channels or industry watchers. We know in the Tesla community, you will get some who pump any scrap of info as game changing and something we will get very soon. It is so nice to have a source of content that is not jumping on the hype train, but trying to look at things through logical glasses instead bright rose colored glasses. Battery tech is a huge part of why the current growth in the EV market has been happening. There would be no Tesla without the battery tech moving forward. Thankfully, Tesla has done much beyond batteries giving them launch pad to become the large scale builder we see today. I agree with the idea of keeping projects on the low side, because it is better to come in low and then get better results. As you noted, you would rather have come in low than to have pumped people up with idea of 20% gains in 6 months than say 10% is the high side and it will likely take a couple of years if all goes well. That seems to be the right tone to me and I appreciate you willing to do that. I am sure you would rack up a bunch more likes if you had just said Tesla battery tech is going to blow away the world this year. Note sure if you will cover this or not, but I know Tesla uses virtual testing more and more for many aspects of the company, is there good virtual testing programs out there that would really help speed the process along. We hear about it for factories and recently the Joe Justice video of him talking about how AI is helping Tesla to design better vehicles and processes thanks to the combo of massive data collection by Tesla and the ever improving AI software. It might be interesting to hear about the software tools out there for helping and who has the best and some of the evidence of the impact of using it. Just a thought. Keep up the great work Jordan.
I'm really glad to hear it! lol, yes, my thumbnails and titles do limit my reach, but I also feel like I have the best subscribers. Lots of smart people commenting and supporting the channel. Yes, I have the unboxed process on my mind 😀
Asymmetric lamination is definitely going to be a thing of the future not because of managing cell swelling, but because different active materials have different energy densities. This is usually addressed through an n:p ratio. When making a cell in a lab, you need to know what the relative capacities are for both the cathode and anode to pair them close to 1:1. If you don't do this then the cell will fail quickly. When you do pair them effectively, I can guarantee that they will be different thicknesses, which is okay. Thicker electrodes are challenging because current densities shoot through the roof when charging at normal rates. Current density dominates the rate at which materials can accept ions. When this happens, the capacity decreases substantially, but can be remedied by charging slowly and I mean extremely slow. You also run into to diffusion issues because you've increased the tortuous paths for the ions to travel and find a binding site. There are research areas looking into this such as porosity engineering and electrolyte optimization.
Asymmetric lamination may work out but given that adding Silicon is already used in 2170 and battery industry, seems like adding Silicon would be more straightforward way to improve 4680. Of course as Tesla is finding out, everything with batteries is harder than it looks for large scale production.
@tomodakisama as you said - more levers to pull equals more complexity and more problems to resolve. That means more engineering required and money/staff/time limit what you can do
Thanks Jordan! I assume that a prismatic cell can be formed with thicker coatings, due to not needing the electrodes to be wound into a spiral. (In simplistic terms, parts with thick paint don't like to be bent.) But this limitation may be in different directions for different materials. For example, graphite may be weak to compression, due to its porous nature, so the inner layer may be limited in thickness. On the other hand, metals and oxides may be stronger against compression, and can be made thicker on the inner layer. That may allow Tesla to increase coatings on the mechanically unconstrained side, to increase the active material, as you stated.
The limit of the coating thickness is the liquid slurry separating. With that said, the points you make are relevant. There are many factors to consider with each form factor and cell design.
Depends on the improvement. Going with a high nickel cathode is easier, but going with a high silicon anode is tough. It's the type of thing where you don't know how difficult it is until you start trying to solve the problem, so there's no way to know as an outsider
Thank you Jordan. As first answer I was thinking of different diameters. But temperature makes more sense as this can be varied more easy to adjust thicknesses. When I get it right the different thicknesses would create a kind of battery with 2 different C-rates. Most extrem in the same way when there is a battery and a capacitor in parallel. But as we have the same voltage it's not possible to move energy from the 'big slower' buffer to the smaller faster and more flexible buffer. But I am not a specialist.
@@thelimitingfactor "The voltages will be continuously variable during charge and discharge" I don't think this is relevant. Not voltage change over time, but voltage would be equal across battery pack and capacitor pack.
Asymmetric coating is a common feature in cylindrical cells, with certain 18650 cells exhibiting a 1-2% thicker coating on the electrode side facing outward. This asymmetry, after rolling, enhances overall cell uniformity.
@@thelimitingfactor 1-2% is the optimal value, it ensures a more or less constant anode to cathode ratio in the 18650. Think of the jellyroll as if it consisted of increasingly larger anode and cathode rings. There must be less material per surface area inside the ring than outside for the ratio to be constant. However, in 18650 it does not matter much, and in 4680 it is completely negligible due to the larger diameter. By the way, 4680 are already made with asymmetric coverage. Roughly 1:1.3, so it's not a rumor. Proof in supplementary material of "Lithium-Ion Cells in Automotive Application: Tesla 4680 Cylindrical Cell Teardown and Characterization" Example of asymmetrical coating in 18650 at Figure 3.C of "Aging of Tesla's 18650 Lithium-Ion Cells: Correlating Solid-Electrolyte-Interphase Evolution with Fading in Capacity and Power".
At 2:32 of the video you show a clip from Sky News. I think this is taken from our facility at UKBIC in the United Kingdom as I recognised the sleeving on the cylindrical can!
Great vid and to see where things are going. So far the biggest disappointment with 4680's in the cars has been DC charging performance but good to see some of these improvements can help address charging performance.
Would love to see you do a vid about the charging performance of the Porsche Taycan. Curious to see how you think they accomplished this and if it will lead to accelerated degradation of the pack? And if we can expect to see any of the car's approach to charging (especially aggressive charge curve) with 4680s?
Porsche likely just adjusted things like material thicknesses and improved the cooling system. That is, I don't think there's genuine breakthrough there, just engineering tade-offs (as you pointed out). Without a teardown, there's no way to know what's going on there.
@@GridConnections Porsche owners will tolerate a battery pack replacement at 100k, as maintenance. Just tell them it's part of the "ownership experience" and they will line up
Hey ! Did you hear about the Panasonic deal with Novonix from a few days ago ? It was about time, but it finally seems they are on the way to be a succes 😀
Thanks for the video. I'm missing something. The volume of a cell, and hence the amount of active material, is fixed. Seems to me that It doesn't matter if the electrodes are thick or thin, there's only so much active material that you can put in a cell and that the energy density is fixed (unless, or course, you change the composition). I'm assuming the density of the coating is the same for symmetrical and asymmetrical coatings. What am I missing?
When you increase the thickness of the coating, although the total amount of material remains fixed, the proportion of active material increases. This is just the same as if you're looking at one set of layers. If the active material layers are twice as thick as the metal foil layers...you get twice as much active material in the cell as metal foil. If you have active material layers that are 4 times as thick as metal foil, you get 4 times as much active material as metal foil. NOTE! That doesn't mean it scales linearly. There's decreasing returns... 1/4 vs 1/2 is a bigger difference than 8/10 vs 9/10
With asymmetric lamination it seems like the simpler way to equalize the compression density of the thick and thin sides would be to laminate one side at a time using a two stage application approach
@@thelimitingfactor It’s unlikely that both sides need the same exact amount of compression to have proper adhesion and desired density, if that is the case the side that requires or tolerates higher compression could be stage one then stage two can be done a a slightly lower pressure
I notice that you assume that the modification would be to thicken one side. Might they also be reducing the thickness to one side to allow faster charging?
At practical level going thinner isn't a limiting factor. They already do thin out electodes today to improve rate. The real benefits and issues all appear with increased thickness, and balancing power and energy density.
Tesla is not filling their battery boxes fully now. I think this would only be used for cost savings in production (less battery in a pack). So unless the battery lasts longer I see no advantage to a purchaser at this point. It's excellent that battery tech has a long way to go in efficiency. Good video. Thank you for sticking with what's known.
I wonder if it wouldn’t be simpler to have a few banks of batteries with thinner electrodes that the BMS controls differently than trying to make asymmetric layers
I was thinking this. Whatever benefits you can get from asymmetric coating need to be compared to what you could get by building your pack with two or more varieties of cell. We could instead make "thick" cells and "thin" cells and mix them in the same pack. While I could be missing something, this seems to give the same electrical characteristics as the asymmetric cells. The obvious benefit is that you no longer need to worry about the asymmetric manufacturing challenges. There are some other more speculative benefits: this gives you many more parameters to play with to optimize your battery pack for whatever. In the asymmetric cells, the only parameters you have to play with is two thicknesses. With mixed cells, you can vary the ratio of cell types, you can have more than two cell types, and if it proves useful, you can segregate the cell types so that the battery management system can deal with them differently. My knowledge is insufficient to say that mixed cells always win over asymmetric cells, but absolutely the comparison needs to be made. The increased complexity of inventory control for mixed cells needs to be balanced against the increased complexity of manufacturing and lower flexibility of asymmetric cells.
Yes of course I enjoy the video.., as a suggestion to look at reasons to invest in battery companies that have a high potential, as Tesla and possibly Talga do when the Intrest Rate interference +/-, with Investment, gets done with in the next cycle.
14:35) A-symmetry to optimize spinning foil into the can. Centrifugally, would a thicker outer layer conform to better use of available physical space? (Of course, I ignore Time, And Relative Distance in Space).
Well done! I think the lack of research is likely due to the fact that the average battery scientist is focused on material science and has less capacity to think about the physical engineering of the battery.
I came across information about Taiwans Prologium creating a giga factory for solid state batteries. I would love to see a video about this topic. If their claims are true they would have the best batteries I have seen in my research.
In best part is no part style, you get each battery to have a high charge region and a high-density region. No need for super-capacitors, electrical management system, etc. Each battery plays its part
Nice. Your video describes very well the various nuances of asymetrical electrode designs and what Tesla is trying to accomplish. I wish this could be practically implemented as it would then allow Tesla's R&D team to focus on the other big bottleneck of fast charging: improving electrolyte ionic pathways. In conventional designs, ions can only move in a linear way. This causes problem when using thick electrode designs with fast charging as it means only the cathode particles in direct contact with the electrolyte get ideal discharge/charge performance. Obviously, better electrolytes, thinner electrodes and higher temperatures can be used to increase the practical charging limit, but the base problem still remains. One way that has been done in research and has likely been implemented by CATL in their bleeding edge LFP packs is using optimized parallel electrolyte electrode pathways; this allows for proper electrolyte wetting deeper in the electrode and greatly heightens the fast charging performance ceiling throughout the 10-90% SOC range without requiring active management. I hope Tesla and their R&D teams are already at work on this so that charging finally stops being the bottleneck for EVs, so that the focus only becomes cost and energy density.
@@thelimitingfactor @thelimitingfactor Yes. Hopefully the links don't get blocked this time (they did get blocked it seems). This is the main paper I'm referencing. It's rather complex, but illustrates my talking points in much greater detail. PMID: 36072550 The second paper I paraphrased: DOI 10.1149/2.1171805jes This one is something that I found back in 2022, but never thought that it would amount to anything serious that could be implemented in mass production: Electrolyte Membranes with Biomimetic Lithium-Ion Channels (nanolett.0c01910). With how much I comment at this point, I should just become a Patreon member lmao.
It seems that solid batteries can be made stable by introducing a thicker jelly (...). Is this the case, many of us reading/watching 'that' paper/study and Joe's comments, or at least someone tested all the variations including this one and Joe picked this up by analogy? If I remember, the problem was the lithium to swollen.
@@thelimitingfactor True, my only reference, which anyone could have watched, is this one, ruclips.net/video/XiisqnpGy_E/видео.html .I associated the mismatch of both sides and the gel to complement the thinner asymmetric side. after all, controlling the swollen lithium (don't know how is used) and other factors seem interesting in both cases, the Uni study and any experiment with the 4680. Do not worry, I'm just an old man curious about all this and thanks for your attention.
As someone who turns off your very good videos about half way through because they go over my head. I understood the second half of your video better then the first half. I did not understand the display with material on both sides of the anode and catholic (copper/aluminum). I dont recall seeing anything like this before. Or is my brain going through a brainfart?
Good question! A lot of people assume it's coated just on one side. However, that's just to simplify it. Typically, cathodes and anodes are coated on both sides. So, it's not new, but usually the visual representation is simplified
I am absolutely certain at this point they're going to be no less than three simultaneous high impact breakthroughs in the next year. Tesla is going to have to create a low mileage version of each model. Maybe one that can accept a later add-on of batteries
I am curious why some sort of electrostatic dispersion techniques aren't used to put the anode/cathode materials on the foils. Using rollers seems clumsy.
Makes sense. Electrostatic dispersion would apply powder in lower densities, precisely and uniformly and then rolled. I guess using laser printer tech is not good enough.@@thelimitingfactor
I don't agree at all with the general thesis. This is the same kind of argument that people use when they suggest that mixing supercapacitors with batteries (that have thick electrolytes) could increase performance and range. No one has done that because it is not cost effective or weight/volume effective. There is an optimal thickness for the electrolyte and I think that any power gained by the thinner part would be negated by the fact that only half of the cell has this advantage and likewise the thicker for range part. However, what if this variation is just the slight difference one gets because one layer is on the inside of the foil and the other is the outside? When wound into the jelly roll, equal thickness when laminated when flat will become thicker on inside and thinner on outside. So maybe just enough to compensate?
But with asymmetric laminations. Isnt that difference going thru the jelly roll. But the 4680 cell is going across the width of the cell. NOT thru the entire length of the jellyroll like most other cells.? Like using the butterfly cap and Not the tabs. The current is going vertical Not lateral. This is a non issue going tabless and changing direction across the jellyroll..
Wouldn’t it be easier to do a double roll of thick and thin? This would have thick/thin battery’s but they should balance through the current collectors.
Making even thick and thin layers (easy to make) then wrap thick-thick-thin-thin it makes almost the same battery. Every fourth layer is a thick against thin making it slightly less gain, so maybe 5% only.
@@thelimitingfactor The way I see it, having 3 huge machines churning out 3 simple products that require no R&D would be cheaper than having 1 huge machine churning out 1 complex product that requires additional R&D. Where is the fault in my reasoning? does the self balancing behavior seen at the cell level not translate directly to the pack level? I assume I must have a faulty premise, seeing as this is not the industry standard by now. Model Y has 830 4680 cells, would there not be some configuration of High Capacity Slow Charging / Low Capacity Fast Charging cells that'd be beneficial to gaining faster charging and higher capacity?
The more technical the video is, the more paid supporters I get - who are the backbone off the channel. That is, I'm not making tech porn for casual viewers.
I'm not an expert in this whatsoever, but this seems really similar to that one company that made a 750mile battery for Model S. If I remember correctly, they mixed 2 different batteries, one that charges really fast but has low density and other that charges slow and has a really good density. This seems really similar since it would essentially be a lower density, fast charging cell on one side of the roll, and then one slower charging higher density on the other side. So why doesn't Tesla just make 2 different 4680s one with thin and other with thick electrode and mix them inside the pack, seems like a simpler idea if this doesn't work out.
@thelimitingfactor I know Tesla is all about efficiency and this would definitely be better and more future proof. But if this turns out to be impossible to achieve in some way, 2 battery approach seems like a good idea, at least for more premium models. BYD blade to me seems like a clear winner in almost all aspects for midrange and budget cars. I don't really see the future for 4680 in cars, but for trucks and semis it's a good option.
Increasing the energy density gets rid of the dead weight, so it'll be cheaper Or to be more specific, each cell will cost more, but it'll cost less per kilowatt hour
Don't you find it weird that there's no "test" of a 4680 battery anywhere on the internet, i mean people have them, you can buy them, so why hasn't anybody tested them and posted a video about it!?
What do you mean? I did and entire teardown with UC San Diego and posted on this channel. Multiple videos, including SEM, energy density, resistance, etc.
I'm not sure what you mean by "ready yet." They're in vehicles. As for the charging curve, that's a separate issue - actually several separate issues. But as I showed in the image in this video, its already charging better than the competition in the truck space. Overall, it's a middle of the road cell all around - people are just disappointed because they expected it to be the best on the market... But, performanc is only a side benefit of the 4680. The main purpose of the program is cost and scalability.
@@thelimitingfactor Also the 800v chargers haven't been rolled out/activated yet. That doesn't change the cell but it should allow total charging time for the Cybertruck to improve.
Bingo Ed. Lot of factors there. There seems to be a few influencers out there making premature or unsupported claims, and then everyone else just repeats what they say. My approach at the moment is wait and see. The charging curve could get better or worse as they get more data.
actually I suspect calendaring would be EASIER on a slightly thicker electrode, and my reasoning is that it would be less prone to damage from slightly uneven grain size by increasing roller distance. Where as a 5micron particle might damage a roller or film lamination with a 5micron lamination thickness and you avg large particle is 5micron, if you increase the coating to 7micron, you clearance allowance saves the rollers. In particular with the dry coating process I bet it's like many adhesion problems, to thin is harder to do than slightly thicker.
Good insight thanks!
Another wonderful video, thank you for that Jordan.
I think another thing to be thankful for is not becoming a hype machine like some channels or industry watchers. We know in the Tesla community, you will get some who pump any scrap of info as game changing and something we will get very soon. It is so nice to have a source of content that is not jumping on the hype train, but trying to look at things through logical glasses instead bright rose colored glasses.
Battery tech is a huge part of why the current growth in the EV market has been happening. There would be no Tesla without the battery tech moving forward. Thankfully, Tesla has done much beyond batteries giving them launch pad to become the large scale builder we see today.
I agree with the idea of keeping projects on the low side, because it is better to come in low and then get better results. As you noted, you would rather have come in low than to have pumped people up with idea of 20% gains in 6 months than say 10% is the high side and it will likely take a couple of years if all goes well. That seems to be the right tone to me and I appreciate you willing to do that. I am sure you would rack up a bunch more likes if you had just said Tesla battery tech is going to blow away the world this year.
Note sure if you will cover this or not, but I know Tesla uses virtual testing more and more for many aspects of the company, is there good virtual testing programs out there that would really help speed the process along. We hear about it for factories and recently the Joe Justice video of him talking about how AI is helping Tesla to design better vehicles and processes thanks to the combo of massive data collection by Tesla and the ever improving AI software. It might be interesting to hear about the software tools out there for helping and who has the best and some of the evidence of the impact of using it. Just a thought.
Keep up the great work Jordan.
I'm really glad to hear it!
lol, yes, my thumbnails and titles do limit my reach, but I also feel like I have the best subscribers.
Lots of smart people commenting and supporting the channel.
Yes, I have the unboxed process on my mind 😀
Asymmetric lamination is definitely going to be a thing of the future not because of managing cell swelling, but because different active materials have different energy densities. This is usually addressed through an n:p ratio. When making a cell in a lab, you need to know what the relative capacities are for both the cathode and anode to pair them close to 1:1. If you don't do this then the cell will fail quickly. When you do pair them effectively, I can guarantee that they will be different thicknesses, which is okay.
Thicker electrodes are challenging because current densities shoot through the roof when charging at normal rates. Current density dominates the rate at which materials can accept ions. When this happens, the capacity decreases substantially, but can be remedied by charging slowly and I mean extremely slow. You also run into to diffusion issues because you've increased the tortuous paths for the ions to travel and find a binding site. There are research areas looking into this such as porosity engineering and electrolyte optimization.
🙌 Great insights thanks! 🤠
THANKS JORDAN,AND ALL YOUR PATRONS 🤗💚💚💚
Asymmetric lamination may work out but given that adding Silicon is already used in 2170 and battery industry, seems like adding Silicon would be more straightforward way to improve 4680. Of course as Tesla is finding out, everything with batteries is harder than it looks for large scale production.
Depends on how nicely the silicon plays with the dry coating process.
Can't they do both? Is there a reason they can't silicon dope AND laminate asymmetrical? Seems like they just have more levers to pull.
@tomodakisama as you said - more levers to pull equals more complexity and more problems to resolve. That means more engineering required and money/staff/time limit what you can do
Thanks Jordan. Interesting subject. Your in-depth coverage is much appreciated.
Thanks Jordan! I assume that a prismatic cell can be formed with thicker coatings, due to not needing the electrodes to be wound into a spiral. (In simplistic terms, parts with thick paint don't like to be bent.) But this limitation may be in different directions for different materials. For example, graphite may be weak to compression, due to its porous nature, so the inner layer may be limited in thickness. On the other hand, metals and oxides may be stronger against compression, and can be made thicker on the inner layer. That may allow Tesla to increase coatings on the mechanically unconstrained side, to increase the active material, as you stated.
The limit of the coating thickness is the liquid slurry separating.
With that said, the points you make are relevant. There are many factors to consider with each form factor and cell design.
Perhaps the reason for not pursuing asymmetry is that bigger gain are easier by other aspects of the battery such as chemistry?
Depends on the improvement.
Going with a high nickel cathode is easier, but going with a high silicon anode is tough.
It's the type of thing where you don't know how difficult it is until you start trying to solve the problem, so there's no way to know as an outsider
Thank you Jordan. As first answer I was thinking of different diameters. But temperature makes more sense as this can be varied more easy to adjust thicknesses.
When I get it right the different thicknesses would create a kind of battery with 2 different C-rates. Most extrem in the same way when there is a battery and a capacitor in parallel.
But as we have the same voltage it's not possible to move energy from the 'big slower' buffer to the smaller faster and more flexible buffer. But I am not a specialist.
The voltages will be continuously variable during charge and discharge
@@thelimitingfactor "The voltages will be continuously variable during charge and discharge"
I don't think this is relevant. Not voltage change over time, but voltage would be equal across battery pack and capacitor pack.
great video as always
Many thanks for sharing this with us.
Love these videos, they're the entirety of my education on power cells
🤗JORDAN, WE ARE HOPING YOU ARE SUCCESSFUL 🤔💚💚💚
Thanks man!
Asymmetric coating is a common feature in cylindrical cells, with certain 18650 cells exhibiting a 1-2% thicker coating on the electrode side facing outward. This asymmetry, after rolling, enhances overall cell uniformity.
Interesting. What's kept them from going with larger ratios?
@@thelimitingfactor 1-2% is the optimal value, it ensures a more or less constant anode to cathode ratio in the 18650. Think of the jellyroll as if it consisted of increasingly larger anode and cathode rings. There must be less material per surface area inside the ring than outside for the ratio to be constant. However, in 18650 it does not matter much, and in 4680 it is completely negligible due to the larger diameter.
By the way, 4680 are already made with asymmetric coverage. Roughly 1:1.3, so it's not a rumor.
Proof in supplementary material of "Lithium-Ion Cells in Automotive Application: Tesla 4680 Cylindrical Cell Teardown and Characterization"
Example of asymmetrical coating in 18650 at Figure 3.C of "Aging of Tesla's 18650 Lithium-Ion Cells: Correlating Solid-Electrolyte-Interphase Evolution with Fading in Capacity and Power".
My comment with answer disappeared.
At 2:32 of the video you show a clip from Sky News. I think this is taken from our facility at UKBIC in the United Kingdom as I recognised the sleeving on the cylindrical can!
Ha! Nice 🤠
That would put the max range of the longest range Cybertruck with the extra pack above 500 miles.
Great vid and to see where things are going. So far the biggest disappointment with 4680's in the cars has been DC charging performance but good to see some of these improvements can help address charging performance.
Would love to see you do a vid about the charging performance of the Porsche Taycan. Curious to see how you think they accomplished this and if it will lead to accelerated degradation of the pack? And if we can expect to see any of the car's approach to charging (especially aggressive charge curve) with 4680s?
Porsche likely just adjusted things like material thicknesses and improved the cooling system.
That is, I don't think there's genuine breakthrough there, just engineering tade-offs (as you pointed out).
Without a teardown, there's no way to know what's going on there.
@@GridConnections Porsche owners will tolerate a battery pack replacement at 100k, as maintenance. Just tell them it's part of the "ownership experience" and they will line up
Porsches don't get to 100k until the 4th owner...
Hey ! Did you hear about the Panasonic deal with Novonix from a few days ago ? It was about time, but it finally seems they are on the way to be a succes 😀
Yes! Very cool 💯
Thank you
Thanks for the video. I'm missing something. The volume of a cell, and hence the amount of active material, is fixed. Seems to me that It doesn't matter if the electrodes are thick or thin, there's only so much active material that you can put in a cell and that the energy density is fixed (unless, or course, you change the composition). I'm assuming the density of the coating is the same for symmetrical and asymmetrical coatings. What am I missing?
When you increase the thickness of the coating, although the total amount of material remains fixed, the proportion of active material increases.
This is just the same as if you're looking at one set of layers.
If the active material layers are twice as thick as the metal foil layers...you get twice as much active material in the cell as metal foil.
If you have active material layers that are 4 times as thick as metal foil, you get 4 times as much active material as metal foil.
NOTE! That doesn't mean it scales linearly.
There's decreasing returns...
1/4 vs 1/2 is a bigger difference than 8/10 vs 9/10
@@thelimitingfactor Thanks, Jordan.
With asymmetric lamination it seems like the simpler way to equalize the compression density of the thick and thin sides would be to laminate one side at a time using a two stage application approach
You'd still have to squeeze both sides
@@thelimitingfactor It’s unlikely that both sides need the same exact amount of compression to have proper adhesion and desired density, if that is the case the side that requires or tolerates higher compression could be stage one then stage two can be done a a slightly lower pressure
I notice that you assume that the modification would be to thicken one side. Might they also be reducing the thickness to one side to allow faster charging?
At practical level going thinner isn't a limiting factor.
They already do thin out electodes today to improve rate.
The real benefits and issues all appear with increased thickness, and balancing power and energy density.
Laminate thinner near the middle and thicker around the edge to reduce mechanical stress
Thanks very much again for this deep dive 🙂👌
You're most welcome!
Tesla is not filling their battery boxes fully now. I think this would only be used for cost savings in production (less battery in a pack).
So unless the battery lasts longer I see no advantage to a purchaser at this point.
It's excellent that battery tech has a long way to go in efficiency.
Good video. Thank you for sticking with what's known.
I wonder if it wouldn’t be simpler to have a few banks of batteries with thinner electrodes that the BMS controls differently than trying to make asymmetric layers
That would break from lean principles - unecessary complexity to achieve the same end.
I was thinking this. Whatever benefits you can get from asymmetric coating need to be compared to what you could get by building your pack with two or more varieties of cell.
We could instead make "thick" cells and "thin" cells and mix them in the same pack. While I could be missing something, this seems to give the same electrical characteristics as the asymmetric cells. The obvious benefit is that you no longer need to worry about the asymmetric manufacturing challenges.
There are some other more speculative benefits: this gives you many more parameters to play with to optimize your battery pack for whatever. In the asymmetric cells, the only parameters you have to play with is two thicknesses. With mixed cells, you can vary the ratio of cell types, you can have more than two cell types, and if it proves useful, you can segregate the cell types so that the battery management system can deal with them differently.
My knowledge is insufficient to say that mixed cells always win over asymmetric cells, but absolutely the comparison needs to be made.
The increased complexity of inventory control for mixed cells needs to be balanced against the increased complexity of manufacturing and lower flexibility of asymmetric cells.
Yes of course I enjoy the video.., as a suggestion to look at reasons to invest in battery companies that have a high potential, as Tesla and possibly Talga do when the Intrest Rate interference +/-, with Investment, gets done with in the next cycle.
14:35) A-symmetry to optimize spinning foil into the can. Centrifugally, would a thicker outer layer conform to better use of available physical space? (Of course, I ignore Time, And Relative Distance in Space).
Well done! I think the lack of research is likely due to the fact that the average battery scientist is focused on material science and has less capacity to think about the physical engineering of the battery.
Good point!
I came across information about Taiwans Prologium creating a giga factory for solid state batteries. I would love to see a video about this topic. If their claims are true they would have the best batteries I have seen in my research.
In best part is no part style, you get each battery to have a high charge region and a high-density region. No need for super-capacitors, electrical management system, etc. Each battery plays its part
Bingo!
Nice. Your video describes very well the various nuances of asymetrical electrode designs and what Tesla is trying to accomplish.
I wish this could be practically implemented as it would then allow Tesla's R&D team to focus on the other big bottleneck of fast charging: improving electrolyte ionic pathways.
In conventional designs, ions can only move in a linear way. This causes problem when using thick electrode designs with fast charging as it means only the cathode particles in direct contact with the electrolyte get ideal discharge/charge performance.
Obviously, better electrolytes, thinner electrodes and higher temperatures can be used to increase the practical charging limit, but the base problem still remains.
One way that has been done in research and has likely been implemented by CATL in their bleeding edge LFP packs is using optimized parallel electrolyte electrode pathways; this allows for proper electrolyte wetting deeper in the electrode and greatly heightens the fast charging performance ceiling throughout the 10-90% SOC range without requiring active management.
I hope Tesla and their R&D teams are already at work on this so that charging finally stops being the bottleneck for EVs, so that the focus only becomes cost and energy density.
Thanks for the insights! Much appreciated. Do you have any references that I can look at for the parallel pathways?
@@thelimitingfactor @thelimitingfactor Yes. Hopefully the links don't get blocked this time (they did get blocked it seems).
This is the main paper I'm referencing. It's rather complex, but illustrates my talking points in much greater detail.
PMID: 36072550
The second paper I paraphrased:
DOI 10.1149/2.1171805jes
This one is something that I found back in 2022, but never thought that it would amount to anything serious that could be implemented in mass production:
Electrolyte Membranes with Biomimetic Lithium-Ion Channels (nanolett.0c01910).
With how much I comment at this point, I should just become a Patreon member lmao.
Excellent thank you!
It seems that solid batteries can be made stable by introducing a thicker jelly (...). Is this the case, many of us reading/watching 'that' paper/study and Joe's comments, or at least someone tested all the variations including this one and Joe picked this up by analogy? If I remember, the problem was the lithium to swollen.
There's a lot of shorthand in this comment and I can't parse what's being said and what the question is
@@thelimitingfactor True, my only reference, which anyone could have watched, is this one, ruclips.net/video/XiisqnpGy_E/видео.html .I associated the mismatch of both sides and the gel to complement the thinner asymmetric side. after all, controlling the swollen lithium (don't know how is used) and other factors seem interesting in both cases, the Uni study and any experiment with the 4680. Do not worry, I'm just an old man curious about all this and thanks for your attention.
Thank you!
Aysmmetric pressure is possible with different diameters of the rolls
As someone who turns off your very good videos about half way through because they go over my head.
I understood the second half of your video better then the first half.
I did not understand the display with material on both sides of the anode and catholic (copper/aluminum).
I dont recall seeing anything like this before.
Or is my brain going through a brainfart?
Good question!
A lot of people assume it's coated just on one side.
However, that's just to simplify it.
Typically, cathodes and anodes are coated on both sides.
So, it's not new, but usually the visual representation is simplified
@@thelimitingfactor
Do the ions go through the aluminum and copper foil ?
@@golfish8589 No, just the electrons. Watch my how a battery works video!
I am absolutely certain at this point they're going to be no less than three simultaneous high impact breakthroughs in the next year. Tesla is going to have to create a low mileage version of each model. Maybe one that can accept a later add-on of batteries
I'm looking forward to increased energy densities. I feel like Tesla and battery suppliers have been lacking in improvements the last couple of years.
Would this be easier if Tesla pursued a rectangular prism cell like the blade battery?
No, the main issue is the coating
I am curious why some sort of electrostatic dispersion techniques aren't used to put the anode/cathode materials on the foils. Using rollers seems clumsy.
Cohesion and adhesion
Makes sense. Electrostatic dispersion would apply powder in lower densities, precisely and uniformly and then rolled. I guess using laser printer tech is not good enough.@@thelimitingfactor
I don't agree at all with the general thesis. This is the same kind of argument that people use when they suggest that mixing supercapacitors with batteries (that have thick electrolytes) could increase performance and range. No one has done that because it is not cost effective or weight/volume effective.
There is an optimal thickness for the electrolyte and I think that any power gained by the thinner part would be negated by the fact that only half of the cell has this advantage and likewise the thicker for range part. However, what if this variation is just the slight difference one gets because one layer is on the inside of the foil and the other is the outside? When wound into the jelly roll, equal thickness when laminated when flat will become thicker on inside and thinner on outside. So maybe just enough to compensate?
Nah, different argument. Two different technologies.
But with asymmetric laminations. Isnt that difference going thru the jelly roll. But the 4680 cell is going across the width of the cell. NOT thru the entire length of the jellyroll like most other cells.? Like using the butterfly cap and Not the tabs. The current is going vertical Not lateral. This is a non issue going tabless and changing direction across the jellyroll..
I can't parse this.
Wouldn’t it be easier to do a double roll of thick and thin? This would have thick/thin battery’s but they should balance through the current collectors.
Not sure what this means
Making even thick and thin layers (easy to make) then wrap thick-thick-thin-thin it makes almost the same battery. Every fourth layer is a thick against thin making it slightly less gain, so maybe 5% only.
The facings have to align in terms of capacity. You can't pair thick with thin
Could you not achieve the same result by having 2-3 types of batteries in the battery pack? Low-High or even Low-Medium-High?
Yes, but that would increase the cost and complexity
@@thelimitingfactor
The way I see it, having 3 huge machines churning out 3 simple products that require no R&D would be cheaper than having 1 huge machine churning out 1 complex product that requires additional R&D.
Where is the fault in my reasoning? does the self balancing behavior seen at the cell level not translate directly to the pack level?
I assume I must have a faulty premise, seeing as this is not the industry standard by now.
Model Y has 830 4680 cells, would there not be some configuration of High Capacity Slow Charging / Low Capacity Fast Charging cells that'd be beneficial to gaining faster charging and higher capacity?
Besides the scientifics, we all snore when this videos become all technical.
The more technical the video is, the more paid supporters I get - who are the backbone off the channel.
That is, I'm not making tech porn for casual viewers.
I'm not an expert in this whatsoever, but this seems really similar to that one company that made a 750mile battery for Model S. If I remember correctly, they mixed 2 different batteries, one that charges really fast but has low density and other that charges slow and has a really good density. This seems really similar since it would essentially be a lower density, fast charging cell on one side of the roll, and then one slower charging higher density on the other side. So why doesn't Tesla just make 2 different 4680s one with thin and other with thick electrode and mix them inside the pack, seems like a simpler idea if this doesn't work out.
Because using two different batteries is more complex.
With this all the batteries are the same and coming off the same production line.
@thelimitingfactor I know Tesla is all about efficiency and this would definitely be better and more future proof. But if this turns out to be impossible to achieve in some way, 2 battery approach seems like a good idea, at least for more premium models. BYD blade to me seems like a clear winner in almost all aspects for midrange and budget cars. I don't really see the future for 4680 in cars, but for trucks and semis it's a good option.
What if three layers with the middle layer thicker for symmetry, to avoid degradation?
As I pointed out, we don't know whether it would increase or decrease cycle life, so it requires testing
But the cost/kwh? Would that increase if using more of the expensive materials?
Increasing the energy density gets rid of the dead weight, so it'll be cheaper
Or to be more specific, each cell will cost more, but it'll cost less per kilowatt hour
Thank you. On tangent, is anyone working on a "torus" shaped battery?
It's been mentioend but not pursued because it doesn't have any inherent advantages but would be more expensive to produce
Thank you very much.@@thelimitingfactor
Maybe it would be suitable on a massive scale.
Don't you find it weird that there's no "test" of a 4680 battery anywhere on the internet, i mean people have them, you can buy them, so why hasn't anybody tested them and posted a video about it!?
What do you mean? I did and entire teardown with UC San Diego and posted on this channel. Multiple videos, including SEM, energy density, resistance, etc.
@@thelimitingfactor I mean a proper cycle charge-discharge with a battery charger, that shows real world capacity...
Wouldn’t we have the same thing by using 50% thin
electrodes and 50% thick in the total 4680 mix?
Yes, but then you increase the complexity of your cell manufacturing.
By putting it into the same cell, it doesn't increase the complexity.
"diploid?"
Challenging manufacturing ..?
Does not compute 😁
I'm not sure what the question is
Me making the Philip J Fry suspicious face not knowing if I'm listening to AI voice or not.
lol, it's not AI
I've contemplating using it, but when/if I do I'll make a pinned comment
@@thelimitingfactor I know I saw your posts on X.
Ai should give us more alternatives. Maybe a better way altogether.
Promo sm
No clue what this means
One thing for certain is the 4680's aren't really ready yet as reported by various Cyber Truck owners they have a poor charging curve.
I'm not sure what you mean by "ready yet." They're in vehicles.
As for the charging curve, that's a separate issue - actually several separate issues. But as I showed in the image in this video, its already charging better than the competition in the truck space. Overall, it's a middle of the road cell all around - people are just disappointed because they expected it to be the best on the market...
But, performanc is only a side benefit of the 4680. The main purpose of the program is cost and scalability.
@@thelimitingfactor Also the 800v chargers haven't been rolled out/activated yet. That doesn't change the cell but it should allow total charging time for the Cybertruck to improve.
@@thelimitingfactor They haven't meet the expectations people have come to expect with Tesla charging rates, hence the ready yet
Bingo Ed. Lot of factors there. There seems to be a few influencers out there making premature or unsupported claims, and then everyone else just repeats what they say.
My approach at the moment is wait and see. The charging curve could get better or worse as they get more data.
@@StormyDog That's why they split the battery in two when charging.
"too good to be true" and "tesla" usually come hand by hand these days... 😅
They make impossible merely late