Это видео недоступно.
Сожалеем об этом.
The 4680 Current Collectors Explained // A Temporary Crutch?
HTML-код
- Опубликовано: 6 авг 2024
- What do the 4680 current collectors do? There’s been speculation about what function they serve due to their intricate design. Today I'll walk you through that and why they might be a temporary crutch.
Patreon: / thelimitingfactor
Paypal: www.paypal.com/paypalme/limit...
Bitcoin: 3B3UXCZsnoPZxG6vYQ6npsF4TW8f5kK4LH
Teespring: teespring.com/stores/the-limi...
Reddit: / tlfbatteries
Twitter: / limitingthe
Timeline
0:00 Introduction
1:27 'Current Collector'
2:00 The Positive Terminal // Copper Plate
4:11 The Negative Terminal // Aluminum Plate
6:35 Other Functions of the Current Collectors
7:22 The Design is Evolving // Cleanerwatt Leak
9:14 The Best Part is No Part
10:56 Summary
Intro Music by Dyalla: Homer Said
What a great study and explanation, Jordan, I love this!
Excellent analysis. I may have missed it but in addition to everything you discussed they accommodate the difference in thermal growth of the steel can and the Al film. There is a very small but constant working of the length difference across the length of the cell that uses the "hinge" you accurately call out.
Your explanation of the mechanical elements of the current collector are spot on. I also agree with your rationale about electrical path preference and reasoning. From an optimization standpoint, I agree that the Gen 2 would likely move away from this CC design as it may have significantly dimensional variability which could impact performance should the design be sensitive to that. Best part is no part, and best process is no process!
Always glad to hear your thoughts Nickolas!
Thank you for doing manual closed captions!
Too easy on scripted videos buddy! Thanks for the thanks!
I just can't do it for talking head videos.
Splendid and fascinating as always. The Limiting Factor is one of the rarest and best channels on youtube.
🤠🙌
Min 8:12 - as English is not my native language (and I was not really listening) looking, the 2nd time I realized its a weld (as you said).
Good video and interesting to see how Tesla is still improving the cells but also the manufacturing processes.
Awesome, thanks for the always clear and concise description and explanation!
Sure thing Doug!
Thank you!
love your work brother!
Thanks man 😊
Battery class is now in session! Thanks Jordon!!
🤜🤛🤠
It's important to note that the ENTIRE sheet of copper that the graphite is printed on inside the jellyroll, including the laser cut tabs for the "tabless" electrode, as well as the entire aluminum sheet for the cathode electrode, are current collectors. It is not just the thing that is welded onto the tabs.
Technically correct, but that's not Tesla's internal lingo.
@@thelimitingfactor I’d like clarification please. How does Tesla refer to broad metal foil substrate for plating electrode active material & transferring current (& heat)? BMFSPEAMTC(&H)?😊
@@iandavies4853 Tabless electrode?
With the second generation of the 4680 design, it looks like laser welding is coming of age into mass production. Using laser welds to make continuous leak free joins of the can to the bottom is a critical step. Welding the tabs to the bottom from the outside of the bottom is an interesting application.
Has anyone considered that a Made-in-Texas battery just needed a hat? 🤠
One more excellent video! Thank you so much Jordan!
You're most welcome buddy!
I'd be really interested to see how leading competitors are making pouch cells, and how they compare in complexity and cost of production. To me, the Tesla design looks much less compex to make with very few steps and little handling between those. All of the videos I've seen for pouch cell manufacturing shows complex machines and a lot of steps.
Much harder to fold than to spool!
Some prismatics have shown a spooled foil wound on two pins attached to a bar and spun to make the foil into a racetrack shape. Then the racetrack is inserted into a rectangular can to complete the cell with some compression to minimize the empty space in the center. Some pouch cells are stacks of cut foils with tabs sticking out so all the cathodes can be electrically joined at one end and all the anodes at the other end. I think I also saw some pouch cells with a Z fold stack, but it's been so long ago I may not be recalling correctly. Look for videos on RUclips of cell assembly.
To me it looks like thermal expansion is the main thing they are trying to deal with. The joins in the collectors are all arched like they were deliberately designed with flex in mind, they won’t make good springs being thin foil. I think the new design they are instead relying on the electrodes themselves to take up any flex due to heating. It would also not surprise me if the turned around the joins on the positive collector to obtain a smaller current path and/or simplified it.
As always - very well done!
Thanks Michael!
Thank you for current collection deep dive
You're most welcome!
Page 9:00
Schematic diagram depicts no welding run between electrode and current collector.
Associate photo contradicts that with laser welding on 6 leaves of the collector.
That current collector (on which electrodes are welded to) must be made of hardened sheet metal to remove chance for open circuit under inertial shock.
As ‘tab less electrodes’ are made of annealed aluminum foils (dictated by manufacturability) which will be further deformed into internal open circuit (if not welded) and hence electric spark, electric spark in proximity with heated electrolyte is a fire risk. Welding or soldering will bypass the chance for such open circuit.
I stated above in the same year of the Battery Day using 4680 anode as example.
Have to see how much fatigue at the welds occurs in thermal cycling. Luckily the tabs have some flex. But cycling tests required. I really like the current collector designs.
I just realized that my chemistry 101 is all gone after listening to this episode. 😞
😂
Worth mentioning that the positive is refering to electron deficit. Batteries go against the standard convention and use negative to denote an excess of electrons (like we do in physics). So the anode is negatively charged, meaning it's got an excess of electrons, and the cathode vice versa.
Technically batteries can be in two states, charging and discharging. But usually ppl refer to the discharging state and use electron flow instead of conventional current. This is an important fact.
🙌 Yeah it's a mess, lol
@@thelimitingfactor its beyond a mess. I've been into electronics for more than a decade, and still manage to confuse myself.
The true art of compromise. Probably a triumph of design. Thanks.
good job😄
re welding .... SpaceX uses Friction stir welding ( which isn't welding actually ) it's entanglement mixing in the plastic phase, ... yes like blacksmith hammer welding they will have used that using the pressure and friction to amalgamate the sheets no soldering or welding needed
Lots of people use friction stir welding. I don't see why it's worth mentioning here though.
I guess another option for the end tab caps could be to reduce the chance of internal shorting if the cells get too hot and during expansion/contraction cycles. The cell expands and contracts during disharge and charge, and also during temperature changes. You dont want the expansion to put too much pressure on the end tab caps against the jelly roll, which might lead to shorting of the copper/aluminium ijelly role againt the end tab caps. since there are gaps in the end caps due to complex leaf type structure expansion can be axially rather than longitudonally.
Silicone in the carbon anode reduces the expansion during charge/discharge cycles, carbon delamination is a weak point in lithium cells, not sure if its been wholly overcome as yet.
Yup that is exactly the purpose of these assemblies, it is to maintain good electrical contact while giving the cell guts a strain relief for thermal expansion. They use similar designs for extreme power rectifier diodes.❤
Actually, the use of copper or aluminium is based on whether the metal is electropositive or electronegative, not their reactivity to a specific voltage. The lithium cell is two cells in series, with the use of copper, then carbon and manganese dioxide on to second cell with aluminium of the second cell. Each cell has a voltage of 1.2 volts nominal, which is the same nominal voltage of lead acid cells, which are 6 in series.
I probably could have worded that better and said stability in the anodic and cathodic environment.
Is it possible they were using the current collectors to compensate for "spring force"? with all those overlapping tabs, there has to be some Goldilocks zone of compression, no? too hard and you start to crush the jelly roll, too soft and you don't have enough pressure to guarantee all the tabs touch the end of the can. The only reason i kind of doubt this is the collectors don't seem strong enough to offer much compression resistance.
The current collectors would certainly take up tolerance variation in the jelly roll length. Only need that on one end though. Simplifying the welds and making a reliable electrical connection I think are the two big reasons for the current collectors.
There is also the issue of vibration over the life of the cell in a car. Any vibration over years that could degrade a coating or produce metal particles through abrasion makes a million mile battery problematic. Of course not a problem with stationary storage.
How thick is the aluminum cathode collector foil, do we know?
@6:08 the sides of the battery are isolated because the coolant lines would short between multiple cells.
no coolant lines at the bottom actually, 4680s are in fact side-cooled, and the Coolant Line is not electrically conductive afaik
(and i thought they would be bottom-cooled for a long time, too)
Yes, sides of can carry current, but must be coated to prevent electrical contact with serpentine cooling bandolier.
Thanks!
Thanks Clark!
just watched the video , right I had better explain the shape of the sheets, the issue is if you operate at high current you will get current hogging where the lowest resistance path will carry the most current this leads to failure, the cuts and shaping essentially segments the battery roll into separate segments as the resistance is cumulative and it all needs to be taken into account even a solid block of copper has resistance and has to be balanced, the cuts and shape allows balancing by the positive temp coefficient of the copper you are talking microhms but IT ALL ADDS UP, just take the shape and work it out the actual resistance then push 100 Amos through it and see .... microhms matter the fact this is outside the roll also matters,
🤗👍A BIG THANKS TO YOU AND EVERYONE ELSE JORDAN 😁
We appreciate CLEANER WATTS VIDEO (which we already watched) AS WELL
KEEP UP THE GOOD WORK KEEPING US INFORMED 🤔💚💚💚
🔥
The current collectors are identical to the ones used in hockey puck rectifiers and SCRs. The design is to accomidate for thermal expansion while giving good electrical conductivity upwards of a kiloampare.❤
Cool! Thanks for the input
FYi Tesla has a supply agreement with Syrah. Vidalia, LA is on the Mississippi River about 450 miles from Giga Austin. Oct 20 (Reuters) - Australia's Syrah Resources on Thursday announced plans to expand its Louisiana graphite anode plant fourfold after being selected for a $219.8 million grant . . . In July, Syrah, which mines graphite in Mozambique, signed a $102.1 million loan deal with the U.S. Energy Department for the ongoing construction of its 11,250 tonne per year Vidalia active anode material (AAM) plant and now plans a further upgrade.
The burn tracks on the bottom of the 2nd gen cell might be a soldering operation.
Great video Jordan! I have been wondering about the details of the design since your first video on the insides of the 4680 cell, and this was a very good explanation of those design choices.
One however I couldn't figure out: Why are the collectors transfering current through the longest electron path instead of the shortest. The only reason I can guess at is that it would concentrate the thermal generation on the shortest electron path if done so. However a shorter path would mean a lower resistance. Maybe it's an entirely different reason?
With the speed that electrons travel, what looks to us like a long distance isn’t actually that meaningful. On a computer chip where you’re managing trillions of operations a second, those distances really do matter, but in a battery you’re either charging or discharging, not rapidly switching between the two. The extra millionth of a second it takes to go through the “longer” path doesn’t actually matter.
Which would more water flow through: a skinny short hose or a long thick hose?
I know from experience that it's a long thick hose 😁
Correct, the longer path won''t make a difference. The more important thing is manufacturability, so that's what they designe to.
@@timwildauer5063 But the whole point of the 4680's tabless design is to reduces the internal electrical resistance of the battery cell, by shortening and widening the electron path.
If you have two copper wires with the same cross section, but one is twice as long as the other. Then the resistance of the longer wire will be double that of the shorter wire. And having the same amount of current going through those wires will result in twice the total heat generation of the longer wire. The speed of electrons doesn't impact this.
Batteries are mostly thermally limited. The 4680 cell size was made viable by the lower internal resistance of the battery via the tabless design.
Before the 4680 cell the batteri cells used was much thinner (18650 & 21700) because it makes them easier to cool. If the batteries generate more heat than can be cooled by the cooling system in the battery pack, then engineers must limit the charge/discharge current of the battery cells to avoid high temperatures.
@@thelimitingfactor I could totally see manufacturability being the deciding factor for the design.
Though It still seems strange to me that they would research in a revolutionary new battery design and then apparently hamper it by using a current collector which maximizes electron path.
If you connected two terminals at some random points onto a flat copper plate. Most heat would be generated between the two points.
Therefore, I still think that having a more spread-out heat generation seems to be the most likely reason for the designs of the collectors, to my mind.
Most of Teslas cell design choices seems laser focused on manufacturability, energy density and power output of the cells.
Re Welding (Soldering). As I have been investigating CuSi3 soldering with a welding machine (for old vehicles) instead of welding its getting interesting what materials are connected together. Maybe Tesla uses this at the carossery where where there is steel and aluminum put together.
Re battery, we have a steel can, copper and aluminum. And it seems all the different materials a put together through laser welding (as no welding/soldering material is added). Correct?
The materials are correct except for the can, which is most likely nickel plated high carbon steel
if the joint is gas free it doesn't matter if it's differential metals you only get a bit of thermocouple generation no corrosion, steel is a poor conductor compared to copper incidentally so is aluminium
@@kwinterburn Steel and aluminum are worse electrical conductors, but the thickness of the steel can will easily makes up the difference.
It almost seems that due to the fact that Tesla has oversupply of 2170/LFP for 2022 (and maybe part of 2023), they have chosen to iterate over the design of the 4680 so when they truly ramp up they'll have a superior design which would allow for faster volume production and cheaper cells. It makes sense for Tesla to start low volume production of MY in Austin to get as much learning from 4680 design and chemistry, while not ramping up for as long as they can to allow iterating the design before finalizing the production line. Seeing this I'm actually happy Tesla is pacing the 4680 ramp up and not ramping volume - finding the right balance (quick ramp now vs having to redo everything in a few months) is crucial.
They are still working on "modifications" of the battery construction/manufacturing. At present it is my understanding that they have a "Dry battery cathode" but do not have a "Dry battery electrode" yet.
Both are electrodes. They dont have the anode yet.
Both are electrodes, but they don't have the cathode yet 😀
Will there not be galvanic corrosion when copper collector plate is welded to steel can from inside ?
How about also using the current collector plate as a cell level protection fuse, like the small wire does in earlier 1865 and 2170 designs.
Definitely a good point and if I dig deeper and do another video, I'll look into that!
What if they sprayed the cathode material on a surface that the material didn't stick to? Then run the aluminum sheet over the carbon and press the two together. In pickling steel, they form a Lupe in a pit while holding the ends together so it can be welded, making the process continuously feed into the acid. The rolling process is what causes the problem. In forming Quarts countertops, they press the power into shape, but they use a sheet of paper on top of the power in this case, that is not needed because the sheet replaces the paper. If anyone could make this idea work, it would be the minds at Tesla. Small stockholders just trying to help.
333Wh/KG is better but really we need the SE up higher around 700Wh/KG. Tesla need to speak to J B Goodenough before it's too late.
It's be a while before we see those kinds of energy densities. Lab bench to factory can be a tough fight.
Last person to consult!
ruclips.net/video/kR8CESrigEg/видео.html
Tesla has contacts in all these labs (I'm sure they do because I do).
You won't see the glass battery anytime soon.
Re: Battery Tech Optimization.
These are the EV equivalent: pistons, rings, rods, crank, cams, valves, etc., parts found in archaic ICE power systems. Hotrodders take note. Though you won’t be able to access the battery internal tech, could building “hybrid” packs that use techs that are optimal for various phases of momentum (launch, acceleration, top end, cruise, brake/recovery) be key?
Bespoke cells. That's an interesting concept.
NIO is using a hybrid pack design right now... LFP and NCM cells all in one big pack, no modules. Presumably their BMS can manage the different charge/discharge rates of those different cells... Which would be the difficult part for hottrodders to get right... EV hotrodders are going to have to be hackers. As we have seen with Tesla and other companies, software tweaks optimizing battery management and motor control can make for as big bumps in power and range as physical modifications.
I have no qualifications so if what I say is stupid that is why. I suggest that both collectors be silver plated copper. This would make the battery better. Yes it would cost more. But not much more.
How well does copper weld to steel?
Nickel plated high carbon steel - no clue! Good question
@Jordan Gesige: Recently Charged EVs magazine held an online engineering conference and has a presentation by Dr Wasim Sarwar, Head of Research & Advanced Engineering at Rimac Technology. In this presentation he has several interesting slides showing relative performance in Energy Density and Volumetric Density of several BEVs. He also covers battery form factors, chemistries, and other topics relevant to BEV battery pack designs. It's an interesting presentation by a very knowledgeable person in the industry. Tried to contact you directly but failed to find an email address, so I'm just throwing this out in the comments section hoping you will find it. I have a link to the presentation but don't wish to share it here. What's the best way to make contact to share this?
Did Munro get your 4680 cell to you yet? And if so are there plans for you to send it off for another analysis?
Yup and yup. When data is available, it will be released to Patrons first.
why not just stamp a seamless bottom can same as a pepsi can for example and insert the jellyroll into it positive on top?
This solves no issues. Whether you call it the top or bottom, one end is already seamless... they're deep drawn
And THAT is how you make a plumbus
😂 Rick and Morty FTW
Is it a current collector... Or a tabless tabulator?
🤣🤣🤣
Turbo encabulator, obviously
I'm no expert, but Tesla's rate of progress and iteration seems kind of intimidating.
Coffe cup lid.
😁
"Current Collector" is a bad nave for the part. Reminds me of Transistors. Should be called "Summing Cap".. Current is coming from multiple sources (you could say) and flowing through the "Summing Cap" and exiting through soldered or welded connection. YOur Welcome
obiviously "nave" is "name".. I type in haste
Current collector is the common term used in cell manufacturing.
Merci !
It's strange to me that they used a current collector on the negative terminal for gen 1. Not sure why they felt they needed one there to begin with.
Welding different thickness metal parts is difficult. Welding different materials together is difficult. Doing both at the same time by heating the thicker metal is bound to take a long time to dial in.
@@martylawson1638 Any opinion on the above spin welding speculation?
@@Crunch_dGH Spin or friction welding is great for big strong welds, but it's probably too messy for a battery case. The jelly roll is quite delicate. In the teardown video, the foils acted like they were much thinner than kitchen aluminum foil so the flash from friction welding would tear it up. I'd guess that the can is laser welded on the same machine that does the can to jelly roll weld. Laser welding is extremely precise and super fast. So it's perfect for high speed production of difficult welds.
I love this, but still do not see the alleged advantages of the 4680. In the field and on the road , particularly the Model Y from Gig Texas there’s been no real advantages. It was supposed to create a lighter and more energy dense pack, but it hasn’t.
Thanks Alex! What we're seeing so far is V1. It should improve over time.
My concern is actually the DBE production system on the cathode side.
1) For the same amount of jellyroll (same kWh) in a pack, the 4680 configuration uses about 40%(?) Less steel for the "cans" in total.
.
2) Fewer cells leaving the factory, or conversely, more kWh per cell. If one production speed limitation was the physical number of cans that could be processed, this reduces the effect of that limitation.
.
3) As explained at length elsewhere on the channel, reduction of internal resistance largely eliminates heat build up, which is the enemy of cells. (Not to mention, "Heat" is "energy", so that's energy not leaving the cell as current)
The greater the current leaving a cell without causing heat, the more efficient the cell, the pack and therefore the vehicle.
.
4) (Speculation)
A 4680 is easier to glue into a structural pack..... Stronger?
.
5) Far fewer connections in the pack and much simplified monitoring of pack condition.
.
And probably other things I've missed(?)
Have you watched the Sandy munro teardown of a 4680 model y pack by chance? The whole entire pack, with carpet, seats, brackets, moveable center console etc only weighs 1,100# now. And makes up the entire floor of a dual large single casting model y with structural pack.
They also show it uses less than 1/4 the amount of welds per pack as a previous 2170 model y pack with modules requires.
So it's a whole lot more than just reduced weight or increases in energy density as advantages. It's an entire manufacturing revolution.
Where the castings and 4680 structural pack are just pieces to the overall jig saw puzzle of mass manufacturing modern evs at insane scale.
Reduced assembly line injuries to from stress and muscle fatigue are another benefit nobody seems to care much about but munro also mentioned. Something Joe justice has covered when comparing assembling the 4680 model y vs older tesla model's. 👍🏻
Also elon and drew have stated multiple times this year the 4680s will rivial anything cell wise available by the end of 2022. And be superior by spring 2023 to anything available as a cylindrical cell from any other ev cell maker.
During battery day in 2020 elon had said 24-36 month's realistically before they'd iterated enough to realize most of the gains they discussed during battery day. So they honestly seem to be right on track there. 2023 with semi/ct&4680 model y will be interesting to see&may be the gen2 roadster. Vs catl and byd with their lfp&lmfp cell to pack&cell to chassis designs hitting mass manufacturing. 🤔
As long as Texas and Fremont are making the same models, Tesla will keep the specs as close to the same as possible, this includes weight. Otherwise, customers will quickly begin selecting for the 'advanced' version and most problematically, rejecting a version they see as less advanced. Tesla must ensure that no matter what goes 'under the hood', if it's named the same it must preform the same.
I suspect that a number of people have already been 'holding out' for a Texas built vehicle because they assumed they would be getting a better preforming car and now they're disappointed that there's no real difference in the customer experience. When Tesla begins making products that could only work with the 4680, that's when the differences will begin to become clear.
I think the primary benefit for now is the cost. Tesla wants to get the 4680 cells at scale and at low cost before they focus on making them perform better than other cells currently available. Also, I wouldn’t be surprised if they’ve changed plans a little bit regarding which products receives the 4680s first. I think semi and cybertruck are first priority for 4680 cells since they have enough supply of other cells for model y. Semi and cybertruck require 4680 cells but the model y does not.