Interesting point! After I released this video, I had a researcher tell me that the 2170 cell cans fail at 30 bar rather than 95 bar. So, it seems there's a lot of conflicting information out there.
@thelimitingfactor In the speira white paper we scaled the thickness of 21700 and 18650 cell cans using Barlows formula and got very similar thicknesses like the tesla cans (we checked the similar strength via some quick Solidworks simulations). This is where the ~95 bar yield stress pressure at 20°C for the sidewalls is derived from. The can sidewalls need to remain strucuturally intact to ensure venting through the bottom or top ventig mechanism (as you explained in the video) --> In reality venting happens around ~20 bar for the smaller cells and the can usually fails around the bended area at the top or bottom edges. Also it is important to note the cans have pretty high thickness tolerances due to the production process and can failiure will happen at hightended temperatures --> Even the same cell might behave very differently regarding can failure and venting during safety tests but it is vital the sidewalls stay intact. I think this is why we see the increased can thickness of 46XX cells. Also check out this presentation regarding the topic: www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwiL1KOywfv7AhWN8LsIHY7VBvIQFnoECAsQAQ&url=https%3A%2F%2Fntrs.nasa.gov%2Fapi%2Fcitations%2F20180004170%2Fdownloads%2F20180004170.pdf&usg=AOvVaw2vD851rM5K4-puVMdTx7uq Cheers :)
Oh mighty UA-cam gods. Bless this video with all the promotion is needs. This is better than any paid content I watch and Jordan shares it all for free. And props to those of you who do pay him.
Comparing absolute charging power of different pack sizes is like saying that you can get more kilowatts at the supercharger if you plug in two cars instead of one.
Even though absolute charging power isn't relevant for the discussion of the battery pack's engineering, it is (when weighted with overall vehicle efficency) the ONLY relevant charging statistic for consumers. I can charge a capacitor from 0-100% in less than a second, but it doesn't matter if a pack made from capacitors only gets you 2 meters. And if I had a battery that took 2 days to go from 0-80%, that wouldn't matter, if I only ever needed to charge from 10.1% to 10.2%, because the pack technically has a million km of range when fully charged.
Always enjoy watching your content. The image for this video is also fantastic 😂 Thank you for all the work you do to bring this information to the public, Jordan!
Tesla battery packs typically include the onboard charger and DC/DC converter in their mass. I'm not aware of any other manufacturers that have this situation. So the pack level density difference is even greater.
Exactly right. Plus the 4680 is the structural load floor of all 4680 model y teslas now. And the one munro weighed recently? Still had the seats&rails, center console, carpeting, sound deadening foam and brackets all mounted to it. Yet it weighed just 1100# or so. 🤔
Energy density is not as important as people think given the ongoing improvements in regenerative braking. Cost per kWh remains the key metric as the initial cost of an EV compared with an ICE car is still the limiting factor in the sales of EVs.
Agree with the reason why they are using 2170 in the Semi. Also wondering when they will move Semi production to Austin. Maybe when the 4680 production is big enough to support CT and Semi. Don't expect it to happen before 2024.
4680 was designed to be part of a structural battery pack. It seems unlikely that this feature is useful in the Semi, so we may not see 4680’s in the Semi anytime soon. The added weight of the ‘can’ would not be offset by reduced weight elsewhere.
@@claykarmel7720 good point, although I assume that a stiff chassis is also appealing for a semi. Instead of it being dead weight it can replace some heavy steel beams. And don't forget they need to produce 6x less, pack 6x, weld 6x less etc etc when using the bigger 4680s.
A module batterypack can be an advantage, though. Bjørn Nyland recently reported on a Tesla model S that stranded its owner in Italy. A local EV-Hub repair shop in Italy repaired the faulty module and the owner could drive home to Norway...The price was a few hundred dollars afair.
Thank you Jordan, a most excellent update. I suspect Elon chose the 4680 format not so much due to the form factor having a high volumetric efficiency as for the scalability potential, especially in using their DBE process to achieve massive throughput of materials in their production. Ultimate MIBO machine.... materials in batteries out.
I think the 4680 form factor is only now possible as a result of the tables architecture which reduces internal resistance. It would have generated too much internal heat without tables innovation.
Hi Jordan, firstly, thanks for great detailed videos. The buffer on the outside edges of the battery pack, is there for a 'third' reason.... The more weight is tucked in at the bottom 'centre' of the car, the less likely the car will roll over, and better it will handle the turns... Less centrifuge g-force...
This was incredibly informative! It's obvious that you put in countless hours researching your topic. I'm excited to see ev battery technology progress in the coming years. I'm curious to see how the current 4680 and catl LFP technology will coexist with solid state batteries, if and when they ever reach mass production
Any time you make radical design changes your first version is conservative. Even if one out of a million battery packs explodes due to a fault, it will be all over the news. So the first packs are probably conservatively design, and probably have added sensors embedded looking for problems and performance. Once they have real world data they will iterate and optimize. Look at the Falcon rockets as an example.
Love you videos. I think model Y is taking full advantage of the structural pack and I hope all vehicles moving forward will be able to accommodate a variety of different packs. All packs are a structural component, just the 4680 doesn't require additional steel to be that structure.
Not those packs rubber mounted to frame rails - F-150, hummer. Cybertruck will rely on structural pack & exoskeleton, castings to eliminate need for ladder frame. Point is the cans & foam do the work of internal (to pack) structural beams, while also acting as pressure vessel (and transferring current to top of cells, heat transfer duties). Don’t obsess over one function.
@@iandavies4853 you are correct but on a different topic. I'm talking about the model-Y body that utilizes a 4680 "structural pack" and also uses 2170 packs with additional internal structural members that are not the cells or foam. It's my fault for saying "all battery packs are structural" my flashlight does not use a structural battery pack, my apologies.
@@WentzCraft OK, point taken, discussion restricted to Y. Yet Musk has made claims of exceptional rigidity of foam honeycomb pack allowing further reduction elsewhere in the car. Plus there’s better integration with F&R castings. Tesla just pays more attention to this stuff.
Great video Jordan! Did you see Talga's latest announcement yesterday, an 40% increase in energy capacity with 9% Talnode Si (which contains 50% silicon) Would love another deep dive into Talga.
Regarding the pack, it's easy to see there's space, literally, to volumetric energy density improvement, not so sure about the gravimetric, that is main point discussed.
Great video as always Jordan. Another valid point is that many are saying Tesla is not delivering the revolutionary products they said they would at Battery Day. Lower costs, higher energy density, lower weight, scale. But they forget that TEsla never said all of this would be fulfilled overnight. it was in pieces towards 2026-30. However what they did say was the 100 GWh output target for 2022, which people can rightfully criticize Tesla for not achieving (like you mention the scaling of the DBE cathode).
I could definitely see Tesla letting those dead spaces and foam around the outside stay for fire and crash protection (though calculating the potential capacity of they where used is a decent idea if comparing to other packs that don't have those same features)
@thelimitingfactor I’d be really interested to see you do a deep dive into the potential max C rate of 4680. I think it’s being missed that C rate will be the stand out feature for 4680. The C rate will dictate the max charge speeds for all the future vehicles once 4680 is widely implemented. If it’s a choice between two vehicles and one of them can charge in half the time it’s going to have obvious consumer advantage. My personal feeling is a C rate of 4-5 is probably fairly likely and that would give a 150kWh Cybertruk a charge rate of up too 750kW which would be needed to give a 10-80% charge in about 25min
Very nice video, thank you! One point I disagree with is at 6:15 where the gap between batteries is said to be usable. The robot handling the four blocks of 207 cells needs to be able to reach the bottom (using a flat "hook", much like we use our fingers when carrying a large box) to lift the pack and place it into the pack. Once placed, it needs to get out, so the space is required. It may be possible to change the manufacturing process (i.e. magnetic pickup), but that carries risk (short across current collectors, drop on power failure, etc). It seems much easier to change the battery chemistry, which they plan to do anyways. This also avoids increased mass which can invalidate crash tests (which is the exact problem they ran into with the Roadster). EDIT: I stand corrected, it's already using a magnetic placement that eliminates the hook (which is only used for safety): ua-cam.com/video/fiwUE_2JhvY/v-deo.html
Another solid video. Need more members of the media to follow this channel so they can sound like they know what they are talking about and hopefully get some of the FUD out of the stories we see out there. I appreciate your effort to look at things from all angles. Too many take one point of view and will cement themselves in that position even if the facts shows flaws in their point of view. I 100% agree Tesla is using their young battery tech that will likely have greater levels of improvement left in it, where some of the other battery tech is semi-mature, so not as likely to have as much room for improvements ahead of it. The pace that Tesla improves things is going to mean Tesla will be likely be in strong position with battery tech at least through 2030. I also agree with the idea that at this stage of the EV market, ICE builders are the competition. When one executive with BYD basically said ICE was the competition when asked about Tesla, it made sense. I did not take it as BYD not wanting to acknowledge Tesla. Considering they have praised Tesla, do busines with Tesla, I do not think they are trying to avoid talking about Tesla. Once we see 80% of the annual new vehicle sales being EVs, then it will be time to start looking at EV versus EV for market share. Until then, just need to see more EVs becoming obvious choices over ICE models so that more people will pick EV over ICE because it truly appeals to them. As much as improving air quality, better for the environment, gets away from depending on OPEC and others, should be the reason a consumer buys a vehicle, it is cost to own, cost to operate, and ease of ownership that drives most choices. I do believe EVs will prove to be much easier to own and operate than any ICE option in the near future. We already see that in some categories of vehicles. The FUD that has built up over the past three decades from under funded or limited tech EVs, needs to be removed from the mainstream consumers minds. When that happens, that 80% of the annual new vehicle sales being EVs will happen. With CATL and BYD bringing batteries with sodium in the mix, it will be interesting to see how this plays out with Tesla batteries. Would not surprise me if Tesla buys some of the batteries for use in China or Germany. With the US incentive programs, it will likely make less sense until CATL or BYD gets US battery factories making these . I find it amazing how people keep predicting how hard Tesla will fall, when the current iCE king, Toyota, dumped hundreds of millions into hydrogen R&D and production that they will never truly see profits from, and their first batch of EVs had major issues, as in the company bought back the first batch it was so bad. I am not sure how long it will take Toyota to recover from wheels falling off their EVs image. I know those vehicles were built for them and they slapped the Toyota name on them, but most news stories just had headlines of Toyota EVs wheels fall off. Japan can be thankful Panasonic and Sony have gone after EV related tech, because the legacy Japanese builders went after hydrogen hard. Thanks to the way battery tech has evolved, hydrogen will likely never become mainstream. We may see the end of Japan as a dominate vehicle manufacturer by 2030. That should be about the time Tesla is the biggest volume builder in the world for vehicles.
@@jasonbishop991 BYD has already started delivering in Australia and New Zealand. I believe parts of Northern Europe have gotten their first shipments as well. If I had to guess when and where in the US, they already have a bus factory in the Los Angeles area, so that will likely be where they start bring in EVs. Rumors of a possible US EV car factory coming, but just rumors.
What is the opposite of FUD?? That’s the problem with all the Tesla fanboy and girl 👧 channels! Meanwhile all the Tesla bag 💼 holders cling to the idea 💡 that Musk is going to make them rich 🤑!
I suspect that both gravimetric and volumetric energy density in the Semi pack will be significantly better than a smaller pack using the same cells, as there is more opportunity for pack optimisations in the much larger pack.
when a pack is used as structure, it will be less energy dense. they removed the excess structure, but that doesn't mean they have removed the volume of structure
I also took delivery of our 2023 Y Austin built 4680 recently. Info is so scarce on these. I added a few UA-cam videos on my channel to help. I plan to do a charging test to help gather more data. If anyone has any request I'll do what I can to accommodate.
Just watching the video, from a production engineering perspective I would just say Tesla has created a usable producible product with few real risks, saying it's not good enough is a ridiculous point of view for some of the commentators, it's the draft 1 product, only an idiot would produce in large volume a flagship product with unnecessary risks, so they used and specified a bank of materials guaranteed to work as long as it reaches the sold capacity NOW for production NOW, they are not in the kite flying business, they need products operating in the wild to analyse NOW then you add the risks and reduce the cost, a business does not usually commit suicide in public, dry coating is a risk issue, structural pack is a risk issue, can thickness is a red herring that's obviously to ease risks at the start, a 0.25 mm can would be fine for the pack once you have some actual data which they have very little now that's probably more to do with handling in production than anything else, metrics may be interesting, but in reality the usability of the product is uppermost, Tesla is on its own path , as Elon said the chemistry was ridiculous series of digging a hole and filling it in making the wrong product, they are making the product from the materials that are optimum and making the optimum materials at source, removing all the margins, that is the point, Tesla will be the battery maker in total
CATL Cell to Pack with M3P will blow everyone away on a cell and pack level energy density score. Tesla China-made Model Y and Model 3 Standard Range vehicles should be getting these new packs soon!!
Jordan. Slightly off topic. But I think important. Tesla Semi overnight charging on a "lesser" unit (V3)? . Essentially an "instant network".....? . 4-5 hours? Utilising capacity not generally used overnight? . Discuss
You will probably say this soon, but pack-level density isn’t what matters with a structural pack. It is a structural pack, meaning it provides structure that would otherwise be provided by the frame of the vehicle. You would have to account for this weight savings in the vehicle frame/body itself to get to a real comparison.
My guess as to why they don't use 4680 in the semi, is that those cells can be used for structural packs, and the semi doesn't benefit from a structural pack. Maybe I should elaborate. A structural pack would have high rigidity and relatively low strength. Perfect for normal cars where the strength is enough. A semi needs to be be able to flex as it goes over rough terrain, and then a rigid pack would just crack as you can't expect it to be nearly as strong as a semi truck frame.
Tesla semi has enough travel on air suspension to squat down to fit turntable under trailer when hooking up. I wonder if modern semis still need frame flex / articulation like in olden times? (From ignorance- not a rhetorical question.) Front suspension should be the indicator.
The can does not need much strength for structure. In a box beam, which is what the battery pack is, all the stress is on the upper and lower skins. The cans just need to connect these two surfaces. Think of honeycomb fill between airplane aluminum panels.
I also don't like the charging graphs to compare charging speeds, exactly due to the size and therefore power difference. imho it's best to plot C-Rates over %SOC, because this is independent from the battery size!
Cutting cost first. The whole point is to meet the demand for production and then cut costs. Has nothing to do with having the best battery. It works, it works well, and life should be good. They dont need to be 50% of everyone else. They just need to have enough batteries to be able to sell cars. As other manufacturers speed up EV production, so will costs from demand. Battery supply has been a limiting factor and will be again in the future.
I’m confused 😂 I put the deposit on the model Y 4680 car yesterday, but I’m not sure if I should change to a long range 2170 Model Y. 🤔. . I like the structural pack and as Tesla ramps 4680 , it’s seems like 2170 is still good or if not better?
Greetings: The resistivity (and heat generation) in the copper strip increases with the lenth of the foil in older types of Tesla batteries. The 2170 battery strip is only ~ 800 mm long, the curent from the strip end runs through the resistance of the strip, producing waste heat. 4680 is ~4x longer. so if the same design would be used, the increased resistivity would result in about ~4x more energy losses as heat. Making the connection from the edge eliminates the heat evolution problem of long copper electrode strips. This allows for use of 6 micron foil, (maybe jess in the future), instead of 8 microns. The limitation is how the machines would be able to wind the thinner yet copper foils. The design of 4680 allows with advantage to use thinner copper foil, both in charging and discharging the battery. The speed of Tesla automatic winding is an absolute technical miracle of production efficiency. I can not so far even imagine, how can the flat cell design come anywhere near in production cost, comparing kWh to kWh of capacity. It would be most interesting to disassemble and analyze GM, WW, BYD and CATL batteries.
@@iandavies4853 I have some calculations I can share, but not yet in an open forum. Have you actually seen the 46x120 battery design? Is it available? What batteries disassembled design is available, such as Munro's 4680 Tesla?
If today's Nevada $3.6B Gigafactory announcement for 4680 and Semi assembly means they have either or both ready for mass production, that is big. Would consider the DBE process is moving full steam ahead, no?
IMO the small amount of void space between the 4680 packs is necessary to avoid rupture of many cans at the same time in the event of a major side collision. I do not see the volume being useful in the future.
Great video. It’s like you are head of technical division at Tesla and Elon asked you to do a video of the entire process Tesla went through. Great job again. Also I believe Tesla’s current concern seems to be safety and speed of chargin and not so much as energy density. Tesla would want its batteries quickly absorb energy without affecting battery quality. The talk of Tesla Semis and cyber trucks more often gives away their priorities.
Greetings:Tesla battery production process appears to be very highly automated. Winding the anode and cathode strips ~ 80mm x ~3,500mm alminum foil base and copper foil base apears to me as an exceptionally efficient approach. Ibelieve that most Tesla batteries utilize 8 microns coper foil (~17 g/cel?) or 6 microns (~15 g/cell) microns copper foil. It would be interesting to compare battery production processes of GM, WW, BYD and CCTL with Tesla. (That I am Tesla fan is obvious from our T3, TY, TS-Plaid, T-Roof, two preordered T-Trucks , as well as 5 T-Anejo Tequillas.) 🤩
Unfortunately, we can only really guess at the processes of all these companies. They're super secretive. Thanks for the support Jiri! Shout out coming in the next video 😀 However, I do think it might be worth comparing production processes at some point to show the benefits of cylindrical.
Close packing increases (not reduces) local can stress from internal can pressure. The foam mitigates very little of the additional stress because the stiffness is much less than aluminum.
@@DarylOster Is that because the cells are cylindrical and therefore there is only limited contact between neighboring cells, causing increased stress on the portions of the cell that aren't touching a neighbor?
@@andrewbuck5016 yes, the hoop stress from pressure is a tension load on the wall of the can. Any constraint limiting the slight diameter expansion from any internal pressure will introduce concentrated bending loads in addition to the uniform hoop stress. Design safety factors allow for such additional stresses. A scratch in a can is another type of 'stress riser' that is mitigated by use of a design safety factor.
I'd say, a little crow is not bad,.. Every now and then; the whole, "Caesar is not god" thing,.. It's when crow is all you are eating, and calling it steak, that I got any issue,.. Great episode!👍🏻😎
Something not factored by detractors of the 4680 is manufacturing time and complexity. There are 4-5x less cells, less welds, less fuses and less production using the 4680. Manufacturing speed at this point of expansion is probably more important than energy density.
Still not sure how you are getting such a low pack cost per kWh considering the current market pricing of Lithium Carbonate Equivalent. What is your assumed kg per kWh? I use .8kg/kWh, but could be higher.
Great video. The problem with the 4680 development is that it is that many details are in full public view along with many of the challenges that are being solved in the process. Other manufacturers are not involved in the development of the actual cells and are spending their time using innuendo to confuse consumers into believing that their battery tech is superior. In the short term this may yield results in the consumer buying realm. The challenge is that Tesla has to push development foreword at breakneck speed to avoid appearing inferior. The public is keenly aware that perception is 9/10ths of the law. This channel and others push the fact that Tesla is in fact the leader in battery tech even though they are relatively newcomers to the game in Liew of Panasonic LG Chem and others.
5:56. Hi Jordan. I listened to this several times, still no sure if I understand the statement. What I can say is that the 4680 equipped Model Ys are almost a different vehicle to the 2170 equipped Model Ys. 4680 Model Ys have a rear casting, a front casting, the structural battery pack, and the structural changes along the sills to attach the structural battery pack. 2170 Model Ys have a rear casting, the old-timey stamped steel and welded front end, a floor pan, and a bolt in battery pack. Likely some minor differences between the rear castings also.
Are you trying to derive that from the images on screen? Because that was two years ago and as I've said in the past, they might be able to make a structure that accepts both.
@@thelimitingfactor Partly, I imagine those designs might still be current. The original design above, and the Structural Battery Pack below. From drone flights around the factory often the 2 styles of Body In White are seen. Here for example, ua-cam.com/video/daUeYvFomA0/v-deo.html Also from counting the stacks of front and rear castings outside of the Casting Plant, it’s clear that many more rear castings are being made than front castings. I haven’t done a count lately, but the ratio is about 12 or 13 to 1. Unless they consistently hide the front castings inside somewhere, it’s a good indication that the majority of the vehicles do not have a front casting, therefore no structural battery pack. I was a little surprised when it first became clear. There’s all sorts of follow ons, more complicated production line, or two separate production lines. Stamping out steel components for the front end, plus stamping the various other components that would be unique to the non-structural battery pack Model Y. The space the extra processes take up. Are they making both types of battery pack on site, or shipping in the 2170 packs? Sort of surprised that a 2170 structural battery pack isn’t in production. Indicates to me it either isn’t practical, or they are confident the issues (whatever they are) with the 4680s are in the process of being sorted out, and all Austin production will end up being 4680. I had assumed, as I think most had that all Model Y production at Austin would be with the 4680 Structural Battery Pack. Since body parts were first seen in bins down by the Stamping Plant 12 months ago, it became clear that’s not the case. ua-cam.com/video/a65t8Jdyurc/v-deo.html
@@thelimitingfactor Hi Jordan. Further shots of the different bodies. Good shot here in Jeff’s Roberts’ November 27 video. 3 on the left are 2170 bodies. ua-cam.com/video/ZDnDzTKs91c/v-deo.html Better shot here in Jeff’s September 13 video. Shows the sheet-metal firewall. Model Ys with the cast front end don’t have a firewall. ua-cam.com/video/sxeFdyhyb-k/v-deo.html Joe Tegtmeyer’s 31 August video. Pair of 2170 bodies, and a 4680 to the right. ua-cam.com/video/GrjqQbJBesQ/v-deo.html Munro video from 2 years ago showing the firewall in the 2170 equipped Model Y. ua-cam.com/video/v3BFfPYvh28/v-deo.html Munro video from around August showing the lack of a firewall in the cast front end equipped Model Y. ua-cam.com/video/WNWYk4DdT_E/v-deo.html I see your name in Joe’s list of Patreons. Joe’s pictures from August 31 have some good shots. Second group, 25/30. And the additional 4 showing Model Ys in a recycling trailer.
@@rogerstarkey5390 if you fill it full of cells with no foam. Will it be stronger or weaker? Before they designed the structural pack. There were metal beams between the cells. Now they don't. Where is the strength coming from? The foam is very strong. You can see this in the difficulty getting individual cells out.
The “disappointment” is actually a good thing, this means Tesla has lots of room for growth, we are looking at immature new cells and they are holding up to the mature cells. We won’t have to go far before they pull far ahead. But by far the biggest issue other automakers will face is production volume and in that regard Tesla is many years ahead.
Disappointment is never a good thing outside of recalibrating expectations to something more realistic. Basically we were sold a bill of goods on this battery. The technology is not there. Tesla could not execute on the promise. There really isn’t anything good about that as far as I’m concerned.
Disagree on one point. Semi and Cyber to be Profitable, must have 4680 with about 300wh/kg. The density and cost savings need to be delivered. They delivered what they had a higher cost to Pepsi, but they need a much better battery to produce en masse.
Question: if the can thickness needs to scale with diameter for hoop stress reasons, I assume that is for an unloaded can. If it is eventually used as a structural pack, do the added loads require a thicker can to contain things? I could see additional loads requiring more thickness, but how much are those loads in the same direction as the hoop stress?
Great question! To answer that question for 100% we'd need modelling or actual pack analysis. My understanding is that hoop stress is uniform stress radially from inside the cell, whereas pack rigidity would be more dependent on a twisting motion that compresses the cell unevenly in the vertical axis. The vertical axis is 80mm long, so it has a lot more 'thickness'>>> Much like if you stand on the edge or flat part of a 2X4, the rigidity is my higher on edge.
Suggestion. The "6 pack test" Apply lateral force to a top edge.while doing so, apply similar to the opposite bottom edge. The cans slide against each other as the angle changes. The 4680 cells have foam preventing the slip/ angle change. They have a top plate doing the same. Going nowhere.
@@rogerstarkey5390, I agree. The question is how does the force distribute and does it add to the hoop stress anywhere. It seems plausible that the shear force would distribute uniformly, spread across the can. Those shear forces would all be parallel. There would be a spot on any given cross section where the shear force and hoop stress would be aligned and additive.
I believe the pack is structural, but that may or may not include the cells. In other words, an empty pack or one filled with foam may be just as structural as one filled with cells no matter if those cells are 4680 or 2170 or even 18650.
Hi Jordan, greetings from the U.K. It may well be that Tesla’s 4680 might not be the miracle cell we thought it might be, but I’m not too concerned about this. What we do know is that cell technology/chemistry is going to evolve hugely over the next decade, who will come up with the killer technology is unknown at this time, but Tesla has the ability to be far more agile in adapting to future cell technologies than many other auto makers, with the possible exception of BYD. We are still in the early days of EV cell chemistry, one thing is certain; future cells will be better in every way, and Tesla will pivot to that technology as appropriate.
Or, it may be that the sum of all parts has potential to *be* the miracle? . The aim is to provide a versatile cell, more than sufficient for purpose yet able to advance with technology, scalable to in excess of 3TWh, produced globally(?) in small (modular?) factories. . It seems they need the cathode DBE, then the other elements are quite possible. Even if they get the cathode system to 35% of the envisaged rate, they can simply install 3 sections to supply the system. Still smaller and more efficient than the old way. If (when) they *do* crack the problem, they have spare capacity.
"cell technology/chemistry is going to evolve hugely over the next decade..." Ummm, that is hardly the case. EVs account for a trivial number of uses of Li-technology batteries worldwide, and capacity, lifetime and efficiency have been steadily improved over 20+ years for myriad applications. Tesla is _incredibly_ late to this party. The gains of Li-ion technology will be small and incremental, inevitably leading to a relative dead-end simply due to the materials involved. Li-S has promise to be significant, but it's a really long way off. Solid-state is just a spin on cell safety and stability, with a little bonus of energy density.
No we don’t know that at all. Batteries are not new and a big leap in the technology within 10 years is not something a rational person who understands battery technology would just pencil in as some sort of inevitability.
I think the biggest issue is the claims that Tesla ie Elon made about the 4680. It is cheaper and faster charging. It’s not more volumetrically dense than the 2170.
There is something to be said for production speed. If they can produce 3x as many 4680, with similar density profile, better thermal management and charging speeds.. That's not nothing. People are hung up on density as the be all, end all. Tesla needs as many batteries as they can get, as fast as possible. There's probably not going to be some amazing breakthrough in range people are hoping for. But there doesn't need to be if they can produce way more than they need, faster and at a lower cost than anyone else.
Hello! So must admit that I haven't seen the entire video. Maybe you have already taken this point up... in which case sorry. But wasn't one of the main points of the 4680 pack that it is a structural battery pack? I.e., it is supposed to act as part of the structure of the vehicle frame. Thus, part of the point of the pack is to remove weight from the vehicle frame? Then looking at the energy density of the battery itself becomes a little too simplistic, it must be viewed as part of the vehicle frame. Do you know how much weight can be removed from the vehicle frame due to the new 4680 packs? How does this compare between the different battery packs you have assessed (is it even possible to say)? What matters is the total weight of the vehicle (battery pack + vehicle frame).
Ok I see that you talked about some of this. Bear in mind that there is significant uncertainty in how the different battery packs will be integrated into the vehicle frame and how much weight can be removed from the frame - resulting in differences in "actual energy density" (not energy divided by battery weight but by total structure weight).
Recap: Tesla is like a marriage and although high battery energy density is sexy, looking hot doesn’t make a happy, successful, long-lasting marriage. 😉
Who thinks "Ultium" is any good? Bolt fires much? Old pouch, non-structural, non-LFP tech? It is a previous generation (and generous to even call it that, vs. non-structural 2170 tech). Plus, this completely misses the point. Tesla (and maybe BYD) has shown the ability to switch battery tech (chemistry, form factor, etc.) fast and flexibly (shout out to BMS software). Legacy auto sucks at software and battery flexibility. If better tech is commercialized (sodium?), Tesla will be the most able to incorporate that new tech quickly.
Interesting point! After I released this video, I had a researcher tell me that the 2170 cell cans fail at 30 bar rather than 95 bar. So, it seems there's a lot of conflicting information out there.
No worries..
Awesome video.... Greetings Andrew Wicket.🥰
@thelimitingfactor
In the speira white paper we scaled the thickness of 21700 and 18650 cell cans using Barlows formula and got very similar thicknesses like the tesla cans (we checked the similar strength via some quick Solidworks simulations).
This is where the ~95 bar yield stress pressure at 20°C for the sidewalls is derived from. The can sidewalls need to remain strucuturally intact to ensure venting through the bottom or top ventig mechanism (as you explained in the video) --> In reality venting happens around ~20 bar for the smaller cells and the can usually fails around the bended area at the top or bottom edges.
Also it is important to note the cans have pretty high thickness tolerances due to the production process and can failiure will happen at hightended temperatures --> Even the same cell might behave very differently regarding can failure and venting during safety tests but it is vital the sidewalls stay intact. I think this is why we see the increased can thickness of 46XX cells.
Also check out this presentation regarding the topic: www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwiL1KOywfv7AhWN8LsIHY7VBvIQFnoECAsQAQ&url=https%3A%2F%2Fntrs.nasa.gov%2Fapi%2Fcitations%2F20180004170%2Fdownloads%2F20180004170.pdf&usg=AOvVaw2vD851rM5K4-puVMdTx7uq
Cheers :)
I wonder if the foam contributes any bursting strength to the sides of the cell cans?
thank you for not using too many abbreviations or acronyms, helps me as a beginner
Sure thing!
Oh mighty UA-cam gods. Bless this video with all the promotion is needs. This is better than any paid content I watch and Jordan shares it all for free. And props to those of you who do pay him.
🤣🤜🤛
Well done. I wish others would take on their criticism. You did a professional job.
With one video you answered all my concerns. Bravo and thank you for your work 👏
You're still my go to person for battery info Jordan. Keep up the great work!
Comparing absolute charging power of different pack sizes is like saying that you can get more kilowatts at the supercharger if you plug in two cars instead of one.
Comparing C-rate would be best
@@MaykThewessen best would be average C - Rate until 80%. Peak only isn't that important
Even though absolute charging power isn't relevant for the discussion of the battery pack's engineering, it is (when weighted with overall vehicle efficency) the ONLY relevant charging statistic for consumers. I can charge a capacitor from 0-100% in less than a second, but it doesn't matter if a pack made from capacitors only gets you 2 meters. And if I had a battery that took 2 days to go from 0-80%, that wouldn't matter, if I only ever needed to charge from 10.1% to 10.2%, because the pack technically has a million km of range when fully charged.
I love the contrast in speed. Somebody with an expensive spoon and a dark paste, and an ultra fast battery line. I think I see the bottle neck.😁
Always enjoy watching your content. The image for this video is also fantastic 😂
Thank you for all the work you do to bring this information to the public, Jordan!
Thanks!
You're most welcome Laura!
Tesla battery packs typically include the onboard charger and DC/DC converter in their mass. I'm not aware of any other manufacturers that have this situation. So the pack level density difference is even greater.
Exactly right. Plus the 4680 is the structural load floor of all 4680 model y teslas now. And the one munro weighed recently?
Still had the seats&rails, center console, carpeting, sound deadening foam and brackets all mounted to it. Yet it weighed just 1100# or so. 🤔
Energy density is not as important as people think given the ongoing improvements in regenerative braking.
Cost per kWh remains the key metric as the initial cost of an EV compared with an ICE car is still the limiting factor in the sales of EVs.
Agreed but that’s not the claim Tesla, ie Elon made when rolling out the 4680.
Agree with the reason why they are using 2170 in the Semi. Also wondering when they will move Semi production to Austin. Maybe when the 4680 production is big enough to support CT and Semi. Don't expect it to happen before 2024.
4680 was designed to be part of a structural battery pack. It seems unlikely that this feature is useful in the Semi, so we may not see 4680’s in the Semi anytime soon. The added weight of the ‘can’ would not be offset by reduced weight elsewhere.
@@claykarmel7720 good point, although I assume that a stiff chassis is also appealing for a semi. Instead of it being dead weight it can replace some heavy steel beams. And don't forget they need to produce 6x less, pack 6x, weld 6x less etc etc when using the bigger 4680s.
THANKS JORDAN 🤗 AND YOUR SUPPORTERS FOR DOING THIS FOR US
WE APPRECIATE YOUR HUMILITY AND RESPECT YOUR HONESTY 🤗💚💚💚
A module batterypack can be an advantage, though. Bjørn Nyland recently reported on a Tesla model S that stranded its owner in Italy. A local EV-Hub repair shop in Italy repaired the faulty module and the owner could drive home to Norway...The price was a few hundred dollars afair.
Thank u! Needed something to fall asleep to 😊
As usual, nice to ear from you Jordan 😎❤️👍
I love all the receipts you have. Everything makes sense
Wow I learned a lot thanks. Hope they figure out the dry cathode !
Thank you Jordan, a most excellent update. I suspect Elon chose the 4680 format not so much due to the form factor having a high volumetric efficiency as for the scalability potential, especially in using their DBE process to achieve massive throughput of materials in their production. Ultimate MIBO machine.... materials in batteries out.
I think the 4680 form factor is only now possible as a result of the tables architecture which reduces internal resistance. It would have generated too much internal heat without tables innovation.
Thank you for the great video!
Hi Jordan, firstly, thanks for great detailed videos. The buffer on the outside edges of the battery pack, is there for a 'third' reason.... The more weight is tucked in at the bottom 'centre' of the car, the less likely the car will roll over, and better it will handle the turns... Less centrifuge g-force...
Thanks man! Yeah, it certainly helps with the centre of gravity and polar moment of inertia.
This was incredibly informative! It's obvious that you put in countless hours researching your topic. I'm excited to see ev battery technology progress in the coming years. I'm curious to see how the current 4680 and catl LFP technology will coexist with solid state batteries, if and when they ever reach mass production
😊
As it turns out Semi production gets its own factory at Giga Nevada, not Austin. At least for now
Preferable!
Great to hear from you - always a highlight!!!
Any time you make radical design changes your first version is conservative. Even if one out of a million battery packs explodes due to a fault, it will be all over the news. So the first packs are probably conservatively design, and probably have added sensors embedded looking for problems and performance. Once they have real world data they will iterate and optimize. Look at the Falcon rockets as an example.
Amen!
Extremely interesting, the idea of a B+ structural pack going into a Cybertruck gives me some optimism.
Love you videos. I think model Y is taking full advantage of the structural pack and I hope all vehicles moving forward will be able to accommodate a variety of different packs. All packs are a structural component, just the 4680 doesn't require additional steel to be that structure.
Not full advantage, otherwise it would be better.
@@TecnamTwin so you think that the Y has extra body structure to compensate for the non 4680 structural packs?
Not those packs rubber mounted to frame rails - F-150, hummer.
Cybertruck will rely on structural pack & exoskeleton, castings to eliminate need for ladder frame.
Point is the cans & foam do the work of internal (to pack) structural beams, while also acting as pressure vessel (and transferring current to top of cells, heat transfer duties). Don’t obsess over one function.
@@iandavies4853 you are correct but on a different topic. I'm talking about the model-Y body that utilizes a 4680 "structural pack" and also uses 2170 packs with additional internal structural members that are not the cells or foam. It's my fault for saying "all battery packs are structural" my flashlight does not use a structural battery pack, my apologies.
@@WentzCraft OK, point taken, discussion restricted to Y. Yet Musk has made claims of exceptional rigidity of foam honeycomb pack allowing further reduction elsewhere in the car. Plus there’s better integration with F&R castings. Tesla just pays more attention to this stuff.
Great video Jordan! Did you see Talga's latest announcement yesterday, an 40% increase in energy capacity with 9% Talnode Si (which contains 50% silicon) Would love another deep dive into Talga.
Awesome video.... Greetings Andrew Wicket.🥰
This was great. I listen to a few different Tesla podcasts/channels, and I have been craving deeper technical dives. Thank you!
Soooooo much thanks... Really enjoy your 🧠
😊🤜🤛
Awesome video! So thorough and clean.
Thanks Eric! Looking forward to sharing the next one 😁
@@thelimitingfactor
Next week, Humble pie?
😉👍🖖
Regarding the pack, it's easy to see there's space, literally, to volumetric energy density improvement, not so sure about the gravimetric, that is main point discussed.
Thank you. I was a bit late picking this up. My thoughts in one word - RESPECT.
Thanks, Jordan!
Great video as always Jordan. Another valid point is that many are saying Tesla is not delivering the revolutionary products they said they would at Battery Day. Lower costs, higher energy density, lower weight, scale. But they forget that TEsla never said all of this would be fulfilled overnight. it was in pieces towards 2026-30. However what they did say was the 100 GWh output target for 2022, which people can rightfully criticize Tesla for not achieving (like you mention the scaling of the DBE cathode).
I could definitely see Tesla letting those dead spaces and foam around the outside stay for fire and crash protection (though calculating the potential capacity of they where used is a decent idea if comparing to other packs that don't have those same features)
Thank you!
this channel is much undervalued!!
excellent video, always appreciate your inputs.
Thanks for this update. Eating crow sucks, but it builds character! And who knew that science and engineering are hard?
😂 Amen!
Enjoyed your deep dive.
Nice video, thanks Jordan!
The best 4680 video to date. Thanks for all the research kind sir
@thelimitingfactor I’d be really interested to see you do a deep dive into the potential max C rate of 4680. I think it’s being missed that C rate will be the stand out feature for 4680.
The C rate will dictate the max charge speeds for all the future vehicles once 4680 is widely implemented.
If it’s a choice between two vehicles and one of them can charge in half the time it’s going to have obvious consumer advantage.
My personal feeling is a C rate of 4-5 is probably fairly likely and that would give a 150kWh Cybertruk a charge rate of up too 750kW which would be needed to give a 10-80% charge in about 25min
Very nice video, thank you! One point I disagree with is at 6:15 where the gap between batteries is said to be usable. The robot handling the four blocks of 207 cells needs to be able to reach the bottom (using a flat "hook", much like we use our fingers when carrying a large box) to lift the pack and place it into the pack. Once placed, it needs to get out, so the space is required. It may be possible to change the manufacturing process (i.e. magnetic pickup), but that carries risk (short across current collectors, drop on power failure, etc). It seems much easier to change the battery chemistry, which they plan to do anyways. This also avoids increased mass which can invalidate crash tests (which is the exact problem they ran into with the Roadster).
EDIT: I stand corrected, it's already using a magnetic placement that eliminates the hook (which is only used for safety): ua-cam.com/video/fiwUE_2JhvY/v-deo.html
Another solid video. Need more members of the media to follow this channel so they can sound like they know what they are talking about and hopefully get some of the FUD out of the stories we see out there. I appreciate your effort to look at things from all angles. Too many take one point of view and will cement themselves in that position even if the facts shows flaws in their point of view.
I 100% agree Tesla is using their young battery tech that will likely have greater levels of improvement left in it, where some of the other battery tech is semi-mature, so not as likely to have as much room for improvements ahead of it. The pace that Tesla improves things is going to mean Tesla will be likely be in strong position with battery tech at least through 2030.
I also agree with the idea that at this stage of the EV market, ICE builders are the competition. When one executive with BYD basically said ICE was the competition when asked about Tesla, it made sense. I did not take it as BYD not wanting to acknowledge Tesla. Considering they have praised Tesla, do busines with Tesla, I do not think they are trying to avoid talking about Tesla. Once we see 80% of the annual new vehicle sales being EVs, then it will be time to start looking at EV versus EV for market share. Until then, just need to see more EVs becoming obvious choices over ICE models so that more people will pick EV over ICE because it truly appeals to them. As much as improving air quality, better for the environment, gets away from depending on OPEC and others, should be the reason a consumer buys a vehicle, it is cost to own, cost to operate, and ease of ownership that drives most choices. I do believe EVs will prove to be much easier to own and operate than any ICE option in the near future. We already see that in some categories of vehicles. The FUD that has built up over the past three decades from under funded or limited tech EVs, needs to be removed from the mainstream consumers minds. When that happens, that 80% of the annual new vehicle sales being EVs will happen.
With CATL and BYD bringing batteries with sodium in the mix, it will be interesting to see how this plays out with Tesla batteries. Would not surprise me if Tesla buys some of the batteries for use in China or Germany. With the US incentive programs, it will likely make less sense until CATL or BYD gets US battery factories making these .
I find it amazing how people keep predicting how hard Tesla will fall, when the current iCE king, Toyota, dumped hundreds of millions into hydrogen R&D and production that they will never truly see profits from, and their first batch of EVs had major issues, as in the company bought back the first batch it was so bad. I am not sure how long it will take Toyota to recover from wheels falling off their EVs image. I know those vehicles were built for them and they slapped the Toyota name on them, but most news stories just had headlines of Toyota EVs wheels fall off. Japan can be thankful Panasonic and Sony have gone after EV related tech, because the legacy Japanese builders went after hydrogen hard. Thanks to the way battery tech has evolved, hydrogen will likely never become mainstream. We may see the end of Japan as a dominate vehicle manufacturer by 2030. That should be about the time Tesla is the biggest volume builder in the world for vehicles.
Thanks for the thoughtful, readable, and well-organized comment!
where do you see byd in next few years. they seem to be scaling fast with good battery tech.
@@jasonbishop991 BYD has already started delivering in Australia and New Zealand. I believe parts of Northern Europe have gotten their first shipments as well.
If I had to guess when and where in the US, they already have a bus factory in the Los Angeles area, so that will likely be where they start bring in EVs. Rumors of a possible US EV car factory coming, but just rumors.
What is the opposite of FUD??
That’s the problem with all the Tesla fanboy and girl 👧 channels!
Meanwhile all the Tesla bag 💼 holders cling to the idea 💡 that Musk is going to make them rich 🤑!
@@richardwolf6269agreed. Truth is not very common or popular lately. Elon is part of the problem so the media has a legitimate reason for FUD.
Awesome analysis and presentation as always
thank you
I suspect that both gravimetric and volumetric energy density in the Semi pack will be significantly better than a smaller pack using the same cells, as there is more opportunity for pack optimisations in the much larger pack.
when a pack is used as structure, it will be less energy dense. they removed the excess structure, but that doesn't mean they have removed the volume of structure
I also took delivery of our 2023 Y Austin built 4680 recently. Info is so scarce on these. I added a few UA-cam videos on my channel to help. I plan to do a charging test to help gather more data. If anyone has any request I'll do what I can to accommodate.
Just watching the video, from a production engineering perspective I would just say Tesla has created a usable producible product with few real risks, saying it's not good enough is a ridiculous point of view for some of the commentators, it's the draft 1 product, only an idiot would produce in large volume a flagship product with unnecessary risks, so they used and specified a bank of materials guaranteed to work as long as it reaches the sold capacity NOW for production NOW, they are not in the kite flying business, they need products operating in the wild to analyse NOW then you add the risks and reduce the cost, a business does not usually commit suicide in public, dry coating is a risk issue, structural pack is a risk issue, can thickness is a red herring that's obviously to ease risks at the start, a 0.25 mm can would be fine for the pack once you have some actual data which they have very little now that's probably more to do with handling in production than anything else, metrics may be interesting, but in reality the usability of the product is uppermost, Tesla is on its own path , as Elon said the chemistry was ridiculous series of digging a hole and filling it in making the wrong product, they are making the product from the materials that are optimum and making the optimum materials at source, removing all the margins, that is the point, Tesla will be the battery maker in total
Bingo!
CATL Cell to Pack with M3P will blow everyone away on a cell and pack level energy density score. Tesla China-made Model Y and Model 3 Standard Range vehicles should be getting these new packs soon!!
You didn't watch the last video then?
Jordan.
Slightly off topic. But I think important.
Tesla Semi overnight charging on a "lesser" unit (V3)?
.
Essentially an "instant network".....?
.
4-5 hours? Utilising capacity not generally used overnight?
.
Discuss
Next video 😁
You will probably say this soon, but pack-level density isn’t what matters with a structural pack. It is a structural pack, meaning it provides structure that would otherwise be provided by the frame of the vehicle. You would have to account for this weight savings in the vehicle frame/body itself to get to a real comparison.
Lesson learned today
Stop getting lost in the sauce when your eating a full meal
My guess as to why they don't use 4680 in the semi, is that those cells can be used for structural packs, and the semi doesn't benefit from a structural pack.
Maybe I should elaborate. A structural pack would have high rigidity and relatively low strength. Perfect for normal cars where the strength is enough. A semi needs to be be able to flex as it goes over rough terrain, and then a rigid pack would just crack as you can't expect it to be nearly as strong as a semi truck frame.
Tesla semi has enough travel on air suspension to squat down to fit turntable under trailer when hooking up. I wonder if modern semis still need frame flex / articulation like in olden times? (From ignorance- not a rhetorical question.) Front suspension should be the indicator.
If the semi can't benefit from a structural pack (I don't see how it could) I wonder if Tesla will make a Semi 4680 with a thinner can thickness? 🤔
Ty
The can does not need much strength for structure. In a box beam, which is what the battery pack is, all the stress is on the upper and lower skins. The cans just need to connect these two surfaces. Think of honeycomb fill between airplane aluminum panels.
I see the possible IRA battery/pack manufacturing CRs being a prime factor in reduced space and capital costs
I also don't like the charging graphs to compare charging speeds, exactly due to the size and therefore power difference. imho it's best to plot C-Rates over %SOC, because this is independent from the battery size!
Love your work Jordan! What are your thoughts on the upcoming sodium ion batteries? Australia is doing some good work in the area.
Cutting cost first. The whole point is to meet the demand for production and then cut costs. Has nothing to do with having the best battery. It works, it works well, and life should be good. They dont need to be 50% of everyone else. They just need to have enough batteries to be able to sell cars. As other manufacturers speed up EV production, so will costs from demand. Battery supply has been a limiting factor and will be again in the future.
One 4680 aspect I did not hear about is repairability. It appears from the Munro teardown and analysis that it is not serviceable in any way.
Correct, and Munro explained why that's better.
Unless Munro is going to pay for replacement batteries out of warranty it doesn't matter. Repairability is a big part of sustainability.
@@yolkcapitol They don't repair packs, they replace.
I’m confused 😂 I put the deposit on the model Y 4680 car yesterday, but I’m not sure if I should change to a long range 2170 Model Y. 🤔. . I like the structural pack and as Tesla ramps 4680 , it’s seems like 2170 is still good or if not better?
I'd focus in what kind of range and performance you need rather than what cell is in it.
Greetings: The resistivity (and heat generation) in the copper strip increases with the lenth of the foil in older types of Tesla batteries. The 2170 battery strip is only ~ 800 mm long, the curent from the strip end runs through the resistance of the strip, producing waste heat. 4680 is ~4x longer. so if the same design would be used, the increased resistivity would result in about ~4x more energy losses as heat. Making the connection from the edge eliminates the heat evolution problem of long copper electrode strips. This allows for use of 6 micron foil, (maybe jess in the future), instead of 8 microns. The limitation is how the machines would be able to wind the thinner yet copper foils. The design of 4680 allows with advantage to use thinner copper foil, both in charging and discharging the battery.
The speed of Tesla automatic winding is an absolute technical miracle of production efficiency. I can not so far even imagine, how can the flat cell design come anywhere near in production cost, comparing kWh to kWh of capacity. It would be most interesting to disassemble and analyze GM, WW, BYD and CATL batteries.
Note BMW is trying 46x120 as well, (also 4695 4680). Probably need extra volume given inefficient drivetrain, wanting to one-up Tesla.
@@iandavies4853 I have some calculations I can share, but not yet in an open forum. Have you actually seen the 46x120 battery design? Is it available? What batteries disassembled design is available, such as Munro's 4680 Tesla?
Great job
If today's Nevada $3.6B Gigafactory announcement for 4680 and Semi assembly means they have either or both ready for mass production, that is big. Would consider the DBE process is moving full steam ahead, no?
IMO the small amount of void space between the 4680 packs is necessary to avoid rupture of many cans at the same time in the event of a major side collision. I do not see the volume being useful in the future.
If that were true then all previous packs wouldn't be so close packed and the other cells in this pack wouldn't be so close packed, etc.
Great video. It’s like you are head of technical division at Tesla and Elon asked you to do a video of the entire process Tesla went through. Great job again.
Also I believe Tesla’s current concern seems to be safety and speed of chargin and not so much as energy density. Tesla would want its batteries quickly absorb energy without affecting battery quality. The talk of Tesla Semis and cyber trucks more often gives away their priorities.
Greetings:Tesla battery production process appears to be very highly automated. Winding the anode and cathode strips ~ 80mm x ~3,500mm alminum foil base and copper foil base apears to me as an exceptionally efficient approach. Ibelieve that most Tesla batteries utilize 8 microns coper foil (~17 g/cel?) or 6 microns (~15 g/cell) microns copper foil.
It would be interesting to compare battery production processes of GM, WW, BYD and CCTL with Tesla. (That I am Tesla fan is obvious from our T3, TY, TS-Plaid, T-Roof, two preordered
T-Trucks , as well as 5 T-Anejo Tequillas.) 🤩
Unfortunately, we can only really guess at the processes of all these companies. They're super secretive. Thanks for the support Jiri! Shout out coming in the next video 😀
However, I do think it might be worth comparing production processes at some point to show the benefits of cylindrical.
Keep those videos flowing💯💯💯
good news☺️
I notice and appreciate how EVtube 😉contributors are able to leverage each other’s research 👍
It seems like the fact that the cells are tightly packed in would allow neighboring cells to help counteract hoop stress, no?
That's a really interesting point!
Close packing increases (not reduces) local can stress from internal can pressure. The foam mitigates very little of the additional stress because the stiffness is much less than aluminum.
@@DarylOster Is that because the cells are cylindrical and therefore there is only limited contact between neighboring cells, causing increased stress on the portions of the cell that aren't touching a neighbor?
@@andrewbuck5016 yes, the hoop stress from pressure is a tension load on the wall of the can. Any constraint limiting the slight diameter expansion from any internal pressure will introduce concentrated bending loads in addition to the uniform hoop stress. Design safety factors allow for such additional stresses. A scratch in a can is another type of 'stress riser' that is mitigated by use of a design safety factor.
I'd say, a little crow is not bad,.. Every now and then; the whole, "Caesar is not god" thing,.. It's when crow is all you are eating, and calling it steak, that I got any issue,.. Great episode!👍🏻😎
♥️
4680 is a bet made for the next 30 years.
What is that supposed to mean?
Isn't there no silicon in the anode of 4680 at this time?
Something not factored by detractors of the 4680 is manufacturing time and complexity. There are 4-5x less cells, less welds, less fuses and less production using the 4680. Manufacturing speed at this point of expansion is probably more important than energy density.
Still not sure how you are getting such a low pack cost per kWh considering the current market pricing of Lithium Carbonate Equivalent. What is your assumed kg per kWh? I use .8kg/kWh, but could be higher.
Great video. The problem with the 4680 development is that it is that many details are in full public view along with many of the challenges that are being solved in the process. Other manufacturers are not involved in the development of the actual cells and are spending their time using innuendo to confuse consumers into believing that their battery tech is superior. In the short term this may yield results in the consumer buying realm. The challenge is that Tesla has to push development foreword at breakneck speed to avoid appearing inferior. The public is keenly aware that perception is 9/10ths of the law. This channel and others push the fact that Tesla is in fact the leader in battery tech even though they are relatively newcomers to the game in Liew of Panasonic LG Chem and others.
Correct. Ultium has more marketing pizzaz than 4680. Did someone say Tesla doesn't advertise, therefore no marketing? Hmmmm....
Jordan, have you heard of XFC batteries? They use nano particle silicone.
Yeah
5:56. Hi Jordan. I listened to this several times, still no sure if I understand the statement. What I can say is that the 4680 equipped Model Ys are almost a different vehicle to the 2170 equipped Model Ys.
4680 Model Ys have a rear casting, a front casting, the structural battery pack, and the structural changes along the sills to attach the structural battery pack.
2170 Model Ys have a rear casting, the old-timey stamped steel and welded front end, a floor pan, and a bolt in battery pack.
Likely some minor differences between the rear castings also.
Are you trying to derive that from the images on screen?
Because that was two years ago and as I've said in the past, they might be able to make a structure that accepts both.
@@thelimitingfactor Partly, I imagine those designs might still be current. The original design above, and the Structural Battery Pack below.
From drone flights around the factory often the 2 styles of Body In White are seen. Here for example, ua-cam.com/video/daUeYvFomA0/v-deo.html
Also from counting the stacks of front and rear castings outside of the Casting Plant, it’s clear that many more rear castings are being made than front castings. I haven’t done a count lately, but the ratio is about 12 or 13 to 1. Unless they consistently hide the front castings inside somewhere, it’s a good indication that the majority of the vehicles do not have a front casting, therefore no structural battery pack.
I was a little surprised when it first became clear. There’s all sorts of follow ons, more complicated production line, or two separate production lines. Stamping out steel components for the front end, plus stamping the various other components that would be unique to the non-structural battery pack Model Y. The space the extra processes take up. Are they making both types of battery pack on site, or shipping in the 2170 packs?
Sort of surprised that a 2170 structural battery pack isn’t in production. Indicates to me it either isn’t practical, or they are confident the issues (whatever they are) with the 4680s are in the process of being sorted out, and all Austin production will end up being 4680.
I had assumed, as I think most had that all Model Y production at Austin would be with the 4680 Structural Battery Pack. Since body parts were first seen in bins down by the Stamping Plant 12 months ago, it became clear that’s not the case. ua-cam.com/video/a65t8Jdyurc/v-deo.html
@@thelimitingfactor Hi Jordan. Further shots of the different bodies.
Good shot here in Jeff’s Roberts’ November 27 video. 3 on the left are 2170 bodies. ua-cam.com/video/ZDnDzTKs91c/v-deo.html
Better shot here in Jeff’s September 13 video. Shows the sheet-metal firewall. Model Ys with the cast front end don’t have a firewall. ua-cam.com/video/sxeFdyhyb-k/v-deo.html
Joe Tegtmeyer’s 31 August video. Pair of 2170 bodies, and a 4680 to the right. ua-cam.com/video/GrjqQbJBesQ/v-deo.html
Munro video from 2 years ago showing the firewall in the 2170 equipped Model Y. ua-cam.com/video/v3BFfPYvh28/v-deo.html
Munro video from around August showing the lack of a firewall in the cast front end equipped Model Y. ua-cam.com/video/WNWYk4DdT_E/v-deo.html
I see your name in Joe’s list of Patreons. Joe’s pictures from August 31 have some good shots. Second group, 25/30. And the additional 4 showing Model Ys in a recycling trailer.
Removing the 1.5 inch of foam and sliding in more cells will weaken the structural strength of the pack.
Just my opinion
Hmmmm
Metal honeycomb instead of foam?
Weaker?
(In terms of torsion)
Nah.
@@rogerstarkey5390 if you fill it full of cells with no foam. Will it be stronger or weaker?
Before they designed the structural pack. There were metal beams between the cells. Now they don't. Where is the strength coming from?
The foam is very strong. You can see this in the difficulty getting individual cells out.
I don't think the foam is contributing as much as the cells to the strength of the structure.
@@TheEvilmooseofdoom what part of the structure is making it stronger now then it was before?
Didn't they remove a number metal crossbars?
The “disappointment” is actually a good thing, this means Tesla has lots of room for growth, we are looking at immature new cells and they are holding up to the mature cells. We won’t have to go far before they pull far ahead. But by far the biggest issue other automakers will face is production volume and in that regard Tesla is many years ahead.
Disappointment is never a good thing outside of recalibrating expectations to something more realistic.
Basically we were sold a bill of goods on this battery. The technology is not there. Tesla could not execute on the promise. There really isn’t anything good about that as far as I’m concerned.
Disagree on one point. Semi and Cyber to be Profitable, must have 4680 with about 300wh/kg. The density and cost savings need to be delivered. They delivered what they had a higher cost to Pepsi, but they need a much better battery to produce en masse.
Profitable is about production costs and mark up, not energy density. I think your reasoning is flawed. What was the cost per unit to Pepsi?
Fun fact: They compared two cars, one with larger cells and one with smaller cells. They had the same weight... but, the 4680 had less capacity.
superb, hardcore as always. Love it!!!!
😊
Question: if the can thickness needs to scale with diameter for hoop stress reasons, I assume that is for an unloaded can. If it is eventually used as a structural pack, do the added loads require a thicker can to contain things? I could see additional loads requiring more thickness, but how much are those loads in the same direction as the hoop stress?
Great question! To answer that question for 100% we'd need modelling or actual pack analysis.
My understanding is that hoop stress is uniform stress radially from inside the cell, whereas pack rigidity would be more dependent on a twisting motion that compresses the cell unevenly in the vertical axis. The vertical axis is 80mm long, so it has a lot more 'thickness'>>>
Much like if you stand on the edge or flat part of a 2X4, the rigidity is my higher on edge.
Suggestion.
The "6 pack test"
Apply lateral force to a top edge.while doing so, apply similar to the opposite bottom edge.
The cans slide against each other as the angle changes.
The 4680 cells have foam preventing the slip/ angle change.
They have a top plate doing the same.
Going nowhere.
@@rogerstarkey5390, I agree. The question is how does the force distribute and does it add to the hoop stress anywhere. It seems plausible that the shear force would distribute uniformly, spread across the can. Those shear forces would all be parallel. There would be a spot on any given cross section where the shear force and hoop stress would be aligned and additive.
I believe the pack is structural, but that may or may not include the cells. In other words, an empty pack or one filled with foam may be just as structural as one filled with cells no matter if those cells are 4680 or 2170 or even 18650.
Jordan is the real slim shady
Even when sitting down.
😁
Hi Jordan, greetings from the U.K. It may well be that Tesla’s 4680 might not be the miracle cell we thought it might be, but I’m not too concerned about this. What we do know is that cell technology/chemistry is going to evolve hugely over the next decade, who will come up with the killer technology is unknown at this time, but Tesla has the ability to be far more agile in adapting to future cell technologies than many other auto makers, with the possible exception of BYD. We are still in the early days of EV cell chemistry, one thing is certain; future cells will be better in every way, and Tesla will pivot to that technology as appropriate.
Amen! The 4680 doesn't lock them into a dead end
Or, it may be that the sum of all parts has potential to *be* the miracle?
.
The aim is to provide a versatile cell, more than sufficient for purpose yet able to advance with technology, scalable to in excess of 3TWh, produced globally(?) in small (modular?) factories.
.
It seems they need the cathode DBE, then the other elements are quite possible.
Even if they get the cathode system to 35% of the envisaged rate, they can simply install 3 sections to supply the system. Still smaller and more efficient than the old way.
If (when) they *do* crack the problem, they have spare capacity.
"cell technology/chemistry is going to evolve hugely over the next decade..." Ummm, that is hardly the case. EVs account for a trivial number of uses of Li-technology batteries worldwide, and capacity, lifetime and efficiency have been steadily improved over 20+ years for myriad applications. Tesla is _incredibly_ late to this party. The gains of Li-ion technology will be small and incremental, inevitably leading to a relative dead-end simply due to the materials involved. Li-S has promise to be significant, but it's a really long way off. Solid-state is just a spin on cell safety and stability, with a little bonus of energy density.
No we don’t know that at all. Batteries are not new and a big leap in the technology within 10 years is not something a rational person who understands battery technology would just pencil in as some sort of inevitability.
I think the biggest issue is the claims that Tesla ie Elon made about the 4680. It is cheaper and faster charging. It’s not more volumetrically dense than the 2170.
There is something to be said for production speed. If they can produce 3x as many 4680, with similar density profile, better thermal management and charging speeds.. That's not nothing. People are hung up on density as the be all, end all. Tesla needs as many batteries as they can get, as fast as possible. There's probably not going to be some amazing breakthrough in range people are hoping for. But there doesn't need to be if they can produce way more than they need, faster and at a lower cost than anyone else.
Elon has said the 4680 packs are a C implementation of an A architecture, replacing an A implementation of a B architecture.
Hello! So must admit that I haven't seen the entire video. Maybe you have already taken this point up... in which case sorry.
But wasn't one of the main points of the 4680 pack that it is a structural battery pack? I.e., it is supposed to act as part of the structure of the vehicle frame. Thus, part of the point of the pack is to remove weight from the vehicle frame? Then looking at the energy density of the battery itself becomes a little too simplistic, it must be viewed as part of the vehicle frame. Do you know how much weight can be removed from the vehicle frame due to the new 4680 packs? How does this compare between the different battery packs you have assessed (is it even possible to say)? What matters is the total weight of the vehicle (battery pack + vehicle frame).
Ok I see that you talked about some of this. Bear in mind that there is significant uncertainty in how the different battery packs will be integrated into the vehicle frame and how much weight can be removed from the frame - resulting in differences in "actual energy density" (not energy divided by battery weight but by total structure weight).
Recap: Tesla is like a marriage and although high battery energy density is sexy, looking hot doesn’t make a happy, successful, long-lasting marriage. 😉
Who thinks "Ultium" is any good? Bolt fires much? Old pouch, non-structural, non-LFP tech? It is a previous generation (and generous to even call it that, vs. non-structural 2170 tech). Plus, this completely misses the point. Tesla (and maybe BYD) has shown the ability to switch battery tech (chemistry, form factor, etc.) fast and flexibly (shout out to BMS software). Legacy auto sucks at software and battery flexibility. If better tech is commercialized (sodium?), Tesla will be the most able to incorporate that new tech quickly.