Investor Engineering Discussions

[mongo]

I wonder if they could add an additional step to mill some of the larger particles down to a smaller size before the process starts?

It may also be possible to make harder rollers that are less likely to dent.

I'm sure they will solve this problem eventually, and perhaps they have already solved it.

Based on past experience with somewhat comparable products, the mix is already super fine grain and homogeneous. However, with the dry electrode process, the mix may still clump afterward, and this could get worse under pressure, like corn starch and minimal water.
They likely have a solution, tougher rollers, more roller stages, or ... Just need to get the updates equipment for that and the other 10% of the bottleneck processes.

 

[mongo]

Last comment on this (which will be so persuasive that all debaters instantly surrender):

Elon didn't say Tesla is "bonding the tops and bottoms of the cylinders to the plates." He said:

"we have a filler [between the cells] that is a structural adhesive as well as flame retardant. So, it effectively glues the cells to the top and bottom sheet"​

So the glue between the cylinders bonds them to the plates, effectively not directly.

Yah, they still have all the busbars and connections on top, can't do minimal gap can to plate adhesive.
Bottom of cell will need an insulator between it and the pack bottom. Also need a coolant layer. Will be interesting how they bond it all.

Top (cabin floor?)
Insulator
Bus bars and goo
Cells and goo
Insulator
Heat plate
Pack bottom
Somewhere in the bottom stack is probably an alignment feature.
To get fancy, the heat plate could be plastic to perform the insulator role and also have pockets for the cells. That would be a trade off for thermal resistance vs electrical isolation compared to aluminium heat plate and thin insulator sheet.

Optionally, they could run pins/ posts from the pack bottom to the top beween cells for strength and cell alignment. That allows cabin floor loads to react against the pack bottom which is already thickened for impact/ debris protection. With 80mm cells gapped by 5mm the void has an 18mm diameter. Plenty of room for a post and even a fastener.

Cell gap itself will be interesting since cells in the same group can touch allowing a density increase, but that probably has other drawbacks like uniformity.

 

[mongo]

Yah, they still have all the busbars and connections on top, can't do minimal gap can to plate adhesive.
Bottom of cell will need an insulator between it and the pack bottom. Also need a coolant layer. Will be interesting how they bond it all.

Top (cabin floor?)
Insulator
Bus bars and goo
Cells and goo
Insulator
Heat plate
Pack bottom
Somewhere in the bottom stack is probably an alignment feature.
To get fancy, the heat plate could be plastic to perform the insulator role and also have pockets for the cells. That would be a trade off for thermal resistance vs electrical isolation compared to aluminium heat plate and thin insulator sheet.

Optionally, they could run pins/ posts from the pack bottom to the top beween cells for strength and cell alignment. That allows cabin floor loads to react against the pack bottom which is already thickened for impact/ debris protection. With 80mm cells gapped by 5mm the void has an 18mm diameter. Plenty of room for a post and even a fastener.

Cell gap itself will be interesting since cells in the same group can touch allowing a density increase, but that probably has other drawbacks like uniformity.

Pack crossection from Giga Betlin tour h/t @UltimateX

https://twitter.com/x/status/1446781892729126915

Side mounted thermal channels every other row (every cell cooled.
Bus bars not shown.
Goo appears to be the insulator (top, bottom, sides).
Real assembly with seat supports (note cross beams for seats are separate from pack).

https://twitter.com/x/status/1446788492558749701

 

[JRP3]

Side cooling??? And I'm not seeing any adhesive in that picture? Maybe it was left out because this is a display pack?

 

[mongo]

Side cooling??? And I'm not seeing any adhesive in that picture? Maybe it was left out because this is a display pack?

The front crossection shows the goo. It's removed for clarity in the intact cell area.

 

[mongo]

Good Busbar pic from @22522's post.

 

[mongo]

After all Steer by Wire would be cheaper by ~$200 per vehicle,

How you you calculate 200 in savings?
Delete:
2 steering shaft, steel, ujoint
Firewall seal
Drive gear in steering rack
Force sensor in rack

Add:
2 x Absolute position sensor to rack
2 x Absolute position sensor to steering wheel
Fully redundant steering control ECU
Force feedback system to steering wheel
Mechanism to limit steering wheel rotation (or change to non clockspring link to airbag/ horn/ switches)

 

[jbcarioca]

How you you calculate 200 in savings?
Delete:
2 steering shaft, steel, ujoint
Firewall seal
Drive gear in steering rack
Force sensor in rack

Add:
2 x Absolute position sensor to rack
2 x Absolute position sensor to steering wheel
Fully redundant steering control ECU
Force feedback system to steering wheel
Mechanism to limit steering wheel rotation (or change to non clockspring link to airbag/ horn/ switches)

The calculation came from a OEM client which commissioned the study. Although theirs is proprietary they considered weight reduction as a material factor, and included the labor savings as a significant factor. Their overhead was included because of supplier management. They treated Steer by Wire as a one time capex hit followed by the reductions.

They had more parts than you list on both sides.
The study was done in 2016. I used their actual US$ amount.

Note: They did not do it. Three stated reasons: 1) Legal risk; 2) Manufacturing opposed it; 3) Their proposed method was to use a company known for supply of aircraft control systems. The management team did not think it was wise to “try to fix something that is not broken”.

Sorry for long response. Sorry also I cannot share the study.

I think Tesla can do this well, and perhaps have created greater benefits through ease of integration with other already automated systems. The other guys saw this as entirely independent if any other automation.

 

[mongo]

The calculation came from a OEM client which commissioned the study. Although theirs is proprietary they considered weight reduction as a material factor, and included the labor savings as a significant factor. Their overhead was included because of supplier management. They treated Steer by Wire as a one time capex hit followed by the reductions.

They had more parts than you list on both sides.
The study was done in 2016. I used their actual US$ amount.

Note: They did not do it. Three stated reasons: 1) Legal risk; 2) Manufacturing opposed it; 3) Their proposed method was to use a company known for supply of aircraft control systems. The management team did not think it was wise to “try to fix something that is not broken”.

Sorry for long response. Sorry also I cannot share the study.

I think Tesla can do this well, and perhaps have created greater benefits through ease of integration with other already automated systems. The other guys saw this as entirely independent if any other automation.

Thanks!
No need to divulge propriety info, I get it.
I'm guessing their starting point was a little different than Tesla's existing electric rack vs conventional power steering pump, belt, hoses, and fluid.

 

[jbcarioca]

Thanks!
No need to divulge propriety info, I get it.
I'm guessing their starting point was a little different than Tesla's existing electric rack vs conventional power steering pump, belt, hoses, and fluid.

Quick references include Nissan’s Super HICAS

HICAS - Wikipedia

en.wikipedia.org

 

[mongo]

A picture of the new 18650 battery pack in the Model S Plaid has surfaced on Reddit:

Tesla Model S Plaid battery pack shows that 18650 cell innovations are not over yet

A picture of the Tesla Model S Plaid’s 18650 battery pack with its modules visible was recently shared online. The image hinted that Tesla is still in the process of optimizing and innovating its 18650 battery technology, despite the cells being quite a bit long in the tooth. Tesla is constantly...

www.teslarati.com

Only 5 modules in the current pack vs 18 in the previous pack.

Neat
Previous large packs were 16 modules. Lower capacity (60kWh) were 14, with the front double stack set removed.

 

[Artful Dodger]

Neat
Previous large packs were 16 modules. Lower capacity (60kWh) were 14, with the front double stack set removed.

Interesting. So if I'm reading this right, the number of modules in the palladium pack might be responsible for the higher total pack voltage.

 

[nativewolf]

Since no one has brought this up, and this is a great thread, I thought I'd ping the posters for the thoughts on the seats mounted directly to the battery pack as shown during the giga berlin tour. To me that is sign that the castings have huge additional manufacturing efficiencies that are yet to be discovered/utilized.

 

[JRP3]

The battery pack doesn't look like a casting so probably not related.

 

[nativewolf]

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

If so, it may be a 4680 form factor but not necessarily with all the advanced chemistry they showed at battery day. Initial 4680s will have the tabless connectors, but possibly a conventional chemistry anode and cathode. Then later they would be upgraded with the dry electrode process, the...

teslamotorsclub.com

I was referring to the pictures of the pack with seats mounted directly onto the pack structure. The poster @Stretch2727 thought that this was pretty damn cool so i thought I'd see what other engineering sorts thought of it. As for me, I can imagine it cutting down assembly time tremendously. I can imagine lots of other savings as well other assembly savings, less error, etc.

 

[mongo]

Interesting. So if I'm reading this right, the number of modules in the palladium pack might be responsible for the higher total pack voltage.

Yeah, less modules allows finer control on pack voltage. Assuming all modules are the same, the total number of bricks would now only need to be a multiple of 5 instead of 16. Previous packs were 96s overall with each module having 6 bricks in series. Granularity of 4.2*16= 67.2V.
Five modules gives granularity of 4.2*5= 21V. Setting these up with 20 bricks each gets 100s for 4% more voltage.

Design constraint then becomes area or cell count. 100kWh old pack was 8,256 cells. Starting at 100s, each cell added/ removed from a brick is a delta total cell count of 100 cells, or about 1.2kWh.
Much easier to adjust size/weight/ capacity.

100s83p = 8,300 cells
100s82p = 8,200
105s78p = 8,192
105s79p = 8,295

 

[mongo]

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

If so, it may be a 4680 form factor but not necessarily with all the advanced chemistry they showed at battery day. Initial 4680s will have the tabless connectors, but possibly a conventional chemistry anode and cathode. Then later they would be upgraded with the dry electrode process, the...

teslamotorsclub.com

I was referring to the pictures of the pack with seats mounted directly onto the pack structure. The poster @Stretch2727 thought that this was pretty damn cool so i thought I'd see what other engineering sorts thought of it. As for me, I can imagine it cutting down assembly time tremendously. I can imagine lots of other savings as well other assembly savings, less error, etc.

Clarification: from.what I saw, the seats are not mounted directly to the pack (no pack lid to seat rail welds). The seat rails are welded to a separate piece of sheetmetal which spans the pack/car. This piece may then be glued to the pack.

Assembly wise, yeah it allows a lot of parts to be mounted to the pack as a subassembly reducing the number of steps on the body line. It also makes the installation of seats, floor ducts, carpet, and center console much easier since there is no body in the way. There are some dependencies like the penthouse connections under the rear seat which may limit the ability to preassemble. However, with the pack mounted to the drive units before the body marriage step, these connections may already be completed.

Correction, rear seat mounts to rear casting, not floor/ pack.

 

[jbcarioca]

Since no one has brought this up, and this is a great thread, I thought I'd ping the posters for the thoughts on the seats mounted directly to the battery pack as shown during the giga berlin tour. To me that is sign that the castings have huge additional manufacturing efficiencies that are yet to be discovered/utilized.

Just thinking logically it seems that they have placed the equivalent of the floor pan on top of the pack. That should save assembly time but probably at least a couple of robots that don't need to place seats. Now I wonder how much else might be done to increase preassembly.

Much of what people have been noticing in Grüneheide does suggest there si even more efficiency than we had imagined. That does add color to the Elon statement last year that 'Model Y built in Germany will be a different car than the one built in California'. I cannot find that quote right now thus '...'.

Thus far maybe none of us actually imagine just how consequential having two huge casting plus the third piece of the structural pack. The weight reduction alone should yield a long series of virtuous consequences.

Do any of us have the capacity to make logical estimates of those things?

On the surface it seems they will have:
-reduced Capex for robots and line,
-reduced labor costs,
-increased speeds,
-far increased body rigidity,
-reduction in warranty costs,
-and more.

The engineers among us could maybe do that,
by that I mean almost all of us in thread except me,
I only counted beans and a few other trivial things.

 

[mongo]

The engineers among us could maybe do that,
by that I mean almost all of us in thread except me,
I only counted beans and a few other trivial things.

Careful there, I almost had to give you a disagree.

 

[mongo]

Post with image from main thread:

Have to come back to the amazing Gigafactory Berlin-Brandenburg Factory Tour.
GA of the interior with the Structured Battery Pack looks like a cakewalk: Start with vent pipes and carpet, add center console and finally seats - you're done!
Every step is done outside the car.
View attachment 720141

Forgot about the foam spacer blocks in the floor. Warm feet!