-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
4680 cell design, chassis integration & factory discussion for investors
- Thread starter Buckminster
- Start date
Amazing summary particularly the second half:
I think Tesla wouldn't be building the Austin and Berlin factories if they were not confident about the 4680 DBE cell production process.
I do wonder if one of the rotating mass problems relates to trying to push the speed boundaries at Kato Rd.
For the pilot line, Tesla wants to push the boundaries, and find the point where machines break, or yields suffer.
A rotating mass problem, probably relates to speed, acceleration, or deceleration.
The obvious temporary work around is, just run the Austin and Berlin lines a bit slower.
One rotating mass problem is probably the Plaid drive for the Roaster driving it with a higher peak RPM for an even better Torque curve.
A Roadrunner production rotating mass problem could be DBE and/or winding.
The problem could be lower yield at speed, or equipment breaking...
Or maybe the production related rotating mass problem doesn't relate to the Roadrunner line.
I have little doubt Tesla will eventually find and fix the problem, assuming there is a problem.
Even if there is a problem, I doubt it is significant enough to prevent a reasonable ramp of Austin and Berlin.
I do wonder if one of the rotating mass problems relates to trying to push the speed boundaries at Kato Rd.
For the pilot line, Tesla wants to push the boundaries, and find the point where machines break, or yields suffer.
A rotating mass problem, probably relates to speed, acceleration, or deceleration.
The obvious temporary work around is, just run the Austin and Berlin lines a bit slower.
One rotating mass problem is probably the Plaid drive for the Roaster driving it with a higher peak RPM for an even better Torque curve.
A Roadrunner production rotating mass problem could be DBE and/or winding.
The problem could be lower yield at speed, or equipment breaking...
Or maybe the production related rotating mass problem doesn't relate to the Roadrunner line.
I have little doubt Tesla will eventually find and fix the problem, assuming there is a problem.
Even if there is a problem, I doubt it is significant enough to prevent a reasonable ramp of Austin and Berlin.
Doggydogworld
Active Member
Tesla must proceed with Austin and Berlin despite 4680 problems. They are required to keep the growth narrative alive. And when asked about Berlin and 4680 in a conference call they specifically said they had "de-risked" the plant. They did not give details of the backup plans, so we can only speculate. Based on the original schedule I suspected 2170, but the (convenient) delays open up other possibilities like LG/Pana 4680 or something really different like CATL LFP prismatic as in the MIC 3 SR+.
I still say they can't get DBE to scale. Another possible backup plan could be non-DBE 4680. I doubt they can do conventional coating at Kato Rd, but could possibly buy pre-coated electrodes and do the tabless winding part themselves.
- Sounds like suppliers will only be making nickel 4680s for Tesla and are 2 years behind
- All 4680s are likely to be Nickel only
- No room for the manganese chemistry anymore
- Iron Phosphate will be a different form factor - Warren is talking about pouches
Doggydogworld
Active Member
Good grief, this guy is a non-stop facepalm. Anyway.....
Yes, 4680 will be nickel regardless of source. LFP 4680 never made sense. And don't buy the 2 years behind claim.
Or maybe nickel-manganese is the one they settle on.
Sorry, but Warren is clueless. BYD Blade is a structural pack based on prismatic LFP. It completely outclasses the LFP 3 SR+ pack (which, to be fair was a quick-to-market kludge that will be replaced by a lasting design at some point).
Last edited:
petit_bateau
Active Member
@dogydogworld
Be gentle with me .....
Q1. What is the problem with LFP 4680 ?
Q2. What causes prismatic LFP to win over pouch LFP ?
Q3. What is different about LFP that causes cylindrical tabless (end-rolled-tabbed, 4680-style) to lose out vs pouch vs prismatic ?
Q4. How does this affect 'structural' pack designs (unless it is as simple as rails must substitute to carry the loads, etc) ?
Q5. Implications for LFP in storage ?
Q6. It might be easier with a mfg process primer for pouch vs prismatic vs cylindrical .......... and these days I know quite a lot about calendering and spooling and winding and all sorts of rotary mfg of stuff ......
Thanks.
I am wondering about this precise point...
My understanding is cylindrical LFP can be easily made, some companies already make 18650 or 2170.
But LFP is more tolerant of prismatic formats, and does work well in a prismatic format.
I think Tesla is getting Prismatic LFP packs from CATL for Model 3.
I'm assuming Tesla will make LFP cells in house for multiple projects:-
- Model 3
- Model 2
- Powerwall
- Powerpack
- Megapack
- 12 V battery and others?
- 4680
- Cylindrical but different size (probably larger diameter)
- Prismatic
Edit: Point made above that BYD blade is a structural pack, It is definitely cell-to-pack,,, can we cite a source that state it is structural?
I do think a larger format prismatic might be a good fit for energy storage products... allowing them to be constructed from fewer cells.
Last edited:
Doggydogworld
Active Member
Cylindrical helps with stuff like electrode expansion and thermal runaway containment that LFP doesn't really have. There's not much reason to accept the downsides of cylindrical if you don't need the benefits.
Pouch vs. prismatic is less clear cut. You can pull up videos showing prismatic manufacturing. It's also a jelly roll, basically the outer layers of a huge cylindrical cell. They "squash" the jelly roll and put it in a flat rectangular case. Pouch is more like stacking sheets of paper than a continuous jelly roll.
I don't know if BYD's Blade is a traditional jelly roll prismatic design, however. It's a very unusual shape, and may borrow from the pouch approach. The shape was chosen so the cells could be the "cross-ribs" of a structural pack. They bond to the top and bottom pack surfaces which act as shear panels, similar to Tesla's design. It's kind of like corrugated cardboard. It's not clear to me how well Blade resists longitudinal bending, which is important in a car, so maybe they use beefy side rails? On the flip side, it seems Tesla does not bond the top and bottom of the 4680s to the pack surfaces as I first thought, but instead fills the space between the cells with glue/flame retardant. That'd be a crap-ton of glue, most of which is basically dead weight.
I've been saying for years all storage will be LFP. There's some kind of patent thing with LFP that only applies outside of China. It expires next year so supply should increase. At that point I figure Tesla will finally shift Powerwall to LFP. I think everyone will use prismatic or pouch for storage.
I doubt Tesla will manufacture LFP cells. As far as I can see they've focused 100% of their internal effort on nickel, and still have a ways to go there.
petit_bateau
Active Member
Thanks, that is helpful.
It would appear that there are a lot of parameters in play and so small advances in any one area can shunt the optimal outcome quite markedly. I really don't care which format and/or chemistry dominates provided that cell supply goes up as quickly as possible. Observing that Tesla appear to have designed to accept a variety of formats and chemistries is encouraging in this respect.
Once again, thanks.
Tam
Well-Known Member
Great for robots who do the assembly but how about when the pack needs to be serviced by humans?
In the past, just drop the pack down, now should the whole top part of the car be lifted up?
Tam
Well-Known Member
Even with the best Quality Assurance, some brand-new batteries might still have to be replaced/repaired during the 8-year/70%-capacity warranty period.
Road hazards could also cause damage to a battery pack.
The cheapest Model Y is $54,990 so even with a price of $22,000 battery replacement, there's still $32,990 equity left in the car which is quite a lot of money to throw away the entire car.
Tesla promised cheaper battery service cost at $5,000 to $7,000 (for Model 3):
Similar threads
- Article
