[galderdi]

Based on this, I calculated a volume of 0.962 L, and so I expected close to 2 kg per cell.


This is a very different set of dimensions, only about 3/4 of the length and 3/4 of the height (and similar thickness): 149 mm x 40 mm x 98 mm is 0.584 L, so 1.23 kg is entirely reasonable (2.1 kg/L) and the 100 kg and 20 kWh totals make sense. The total cell volume is only 47 L, an easy fit in the available volume and reasonably packaged in four rows like that, presumably either with each row staggered back from the one below it or in a rectangular box leaning back, to match the seatback angle. Of course interconnects, structure, enclosure, and any cooling system will bulk it up, but it should still fit.

I'm not sure why the supplier assumes only 144 V / 40 cells = 3.6 V/cell, since most lithium-ion chemistries other than LFP run up to 3.75 VNOM, but it might just be a difference in assumptions about voltage at 0% and 100% SoC due to different charging practices, and the resulting nominal voltage. Samsung SDI specifies 3.68 VNOM.

Those cells are so small that they could almost (but probably not quite) run in a stack down the console under the shifter and parking brake as Duncan suggests, although only about half the wheelbase is available and that would only accommodate only one-third of the required cells.

Oh crud, I can see what happened now. These latest details ARE the ones I used for my estimates.
When I responded a few days back and quoted 200x37x130 I was looking at the wrong row in my spreadsheet. They weren't the ones I used for the estimate. Hence my totals did not match the cell dimensions.
So sorry for the confusion and resulting effort.

The 200x37x130 cells were one of many other options I have considered to some degree. They were LifePo4 105ah (2.02kg) and it was looking more like a single set (45s1p) at about 91kg and 15kwh.

But again I don't really have a firm choice for cells yet. I am a good 12 months possibly 18 months away from needing them.

 

[brian_]

Just because it's interesting to think about... would a complete 3rd or 4th generation Smart ED battery pack fit, across the car and tilted up to match the angle of the seatback?
Smart gen 4 Battery NEW!! 17.6 kw

The overall width will be a little over a metre, so it's probably too wide, but it would be really convenient if it fit. Of course these are built of typical LG Chem cells, not LFP.

 

[brian_]

The 200x37x130 cells were one of many other options I have considered to some degree. They were LifePo4 105ah (2.02kg) and it was looking more like a single set (45s1p) at about 91kg and 15kwh.

Using these large cells singly, instead of pairs of smaller cells in parallel, would eliminate the option of reconfiguring later for twice the voltage. This is likely to be the direction you will need to take, because the market direction is toward larger cells in the case of prismatics. With any cell type any needed number of cells can be connected in parallel, with the more commonly used pouch cells that's (presumably automated) ultrasonic welding, while the prismatic cells are normally connected via threaded terminals. The threaded terminals that make DIY pack building easier make OEM construction heavier and more expensive, so in prismatics the manufacturer will prefer to use cells that have the needed capacity without paralleling.

With the minimum viable EV pack up to about 40 kWh, the cells need to have about 100 Ah capacity (like these, or the 94 Ah cells used in the BMW i3); smaller prismatic cells are only likely to be produced for plug-in hybrids, non-roadgoing applications, and non-automotive markets.

 

[brian_]

Oh crud, I can see what happened now. These latest details ARE the ones I used for my estimates.
When I responded a few days back and quoted 200x37x130 I was looking at the wrong row in my spreadsheet. They weren't the ones I used for the estimate. Hence my totals did not match the cell dimensions.
So sorry for the confusion and resulting effort.

That all makes perfect sense. No problem, and thanks for sorting that out.

 

[galderdi]

That pack looks like a good option except we don't have them here. It would be touch and go for the width. But I am tending to think it would fit. Unfortunately we are a third world country when it comes to EVs. The list is just starting to grow in the last two years or so but still not enough to make a significant impact on the salvage market. So really we have a few Teslas and Leafs. Then there is handful of i3 and MItsubishi Outlander and iMev. That is about it. Hence another reason why I'm not stressed right now. There really will be a different set of options in 12-18 months. There will likely be Jag, Hyundai and MG to add to the mix just off the top of my head. I have some local EV registration figures, I'll dig em up.

 

[galderdi]

Just listing the reasonable numbers as at Jan 14, in Queensland only and only cars (no motorbikes or trucks):
BMW i3 = 48
Hyundai IONIQ = 95
Hyundai Kona = 136
Jag ipace =50
Mini = 17
Mitsubishi IMiev = 39
Mitsubishi Minicab = 27
Nissan Leaf = 267
Renault ZOE = 22
Tesla = 1454

This is in a state with about 3,000,000 cars in total (again excluding trucks and motorbikes). So you can see our salvage market so far hasn't bee great for EV enthusiasts. Just to clarify "reasonable numbers" I mean I didn't bother with those with only 1 or 2 as they are mostly DIY EVs and aren't likely to impact the salvage market.

 

[reiderM]

we had this discussion during the design of galderi's previous project; however, I think the explanation for this choice has already been given:

The motor itself is fairly small and the differential can be easily unbolted. At least from what I remember. So in theory it would be very similar to your current motor approach.

Please correct me if I'm wrong though. This is just what I remember reading online.

 

[galderdi]

Yes the suggestion makes sense. But I already have a DC motor and controller. The leaf option is not readily available here. I am toying with the idea of importing an entire Leaf from Japan but I can't justify that expense and hassle at the moment.

There is no reason why I can't run with what I have for now and then replace it with a better option once I get past the rest of the project expense.

 

[reiderM]

Yes the suggestion makes sense. But I already have a DC motor and controller. The leaf option is not readily available here. I am toying with the idea of importing an entire Leaf from Japan but I can't justify that expense and hassle at the moment.

There is no reason why I can't run with what I have for now and then replace it with a better option once I get past the rest of the project expense.

Ah I didn't notice the AUS flag. Typical American thing of assuming you were American lol.

Yeah importing likely isn't going to be worth any potential cost savings of going with a leaf.

Good luck with your project, looking forward to seeing it completed!

 

[galderdi]

Ah I didn't notice the AUS flag. Typical American thing of assuming you were American lol.

Yeah importing likely isn't going to be worth any potential cost savings of going with a leaf.

Good luck with your project, looking forward to seeing it completed!

Well my household is half American. My wife is from Louisiana. All good. :-D

Importing the leaf becomes viable if I plan to use the motor and batteries. But it's not a huge saving. It's mostly attractive because of the availability. but all the import red tape is a turn off.

 

[sunworksco]

The new Tesla 46mm x 80mm tab-less batteries should be available from wrecks, as soon as they begin rolling down the road.
I’m planning to build an ultralight 550 Spyder, with a monocoque Alulight chassis. Alulight aluminum metal foam panels are 8mm thickness, with 1mm aluminum face plates, heat fusion bonded. This Alulight monocoque chassis is several times more rigid than a Lotus chassis. The panels can be welded, bent and Lotus method of epoxy/Ejot bonded.
There is room behind the engine bulkhead for the batteries. Nice thing about the new Tesla batteries, is that their power density is much greater and they charge faster, so you can use a smaller pack. I’m using suspension A-Arms, front and rear. I’m using the Spyder space frame as a template, for making the monocoque chassis.
Tesla's new battery cell features a "tabless" design, which the company claims will provide five times the energy, six times the power, and 16% more range compared to its old batterycell.

 

[galderdi]

The new Tesla 46mm x 80mm tab-less batteries should be available from wrecks, as soon as they begin rolling down the road.
I’m planning to build an ultralight 550 Spyder, with a monocoque Alulight chassis.
There is room ahead of the engine bulkhead for the batteries. Nice thing about the new Tesla batteries, is that their power density is much greater and they charge faster, so you can use a smaller pack. I’m using suspension A-Arms, front and rear. I’m using the Spyder space frame as a template, for making the monocoque chassis.
Tesla's new battery cell features a "tabless" design, which the company claims will provide five times the energy, six times the power, and 16% more range compared to its old batterycell. View attachment 121663 View attachment 121664

Thanks for the info. Do you have a thread for me to follow? The problem with that type of chassis for me is that it would require a full set of engineering tests so it would add an extra $5000 to my project. The advantage of the chassis I have chosen is that it has already been through those tests. So as long as I stick to the chassis design it saves me $5000. Over there you probably don't need to consider the same constraints. But I am super keen to see how your build progresses. What do you consider ultralight? What would the target weight be? Mine needs to be 550kg or 1,220 lbs.

Don't forget we are about 2 years behind the USA with the rollout of EVs including Tesla. We are the poor cousin.

 

[galderdi]

Something doesn't add up. You mention Alulight which appears to be an Aluminium based sandwich foam product. But the photos you posted show a tubular chassis. Do they make Alulight in tube form and even if they did I can't imagine it being suitable for a tube chassis? Are you planning on designing a bespoke chassis using Alulight panels? I am confused.

 

[sunworksco]

Thanks for the info. Do you have a thread for me to follow? The problem with that type of chassis for me is that it would require a full set of engineering tests so it would add an extra $5000 to my project. The advantage of the chassis I have chosen is that it has already been through those tests. So as long as I stick to the chassis design it saves me $5000. Over there you probably don't need to consider the same constraints. But I am super keen to see how your build progresses. What do you consider ultralight? What would the target weight be? Mine needs to be 550kg or 1,220 lbs.

Don't forget we are about 2 years behind the USA with the rollout of EVs including Tesla. We are the poor cousin.

Sorry, no build images, right now. I’m planning to collaborate with a Spyder kit maker and can’t share much more. The entire car will weigh 1,000Lbs. (453.592 kilograms).
I’m assuming that Australian built cars have to go through stringent engineering to receive a registration?
Your chassis is very nice. I like monocoque chassis’s, because they’re much more rigid and offer slightly more interior space. Also the Alulight construction is very simple, using wood tools to cut and shape. You can also use the Lotus/Alcoa panel bonding construction, with one-part epoxy and Ejot screw rivets.
Alulight tubing design is made. The space tube frame is used as a template, to build the Alulight panel chassis.

Page 1 of 707 for Automotive : Gardner Web

 

[galderdi]

Sorry, no build images, right now. I’m planning to collaborate with a Spyder kit maker and can’t share much more. The entire car will weigh 1,000Lbs. (453.592 kilograms).
I’m assuming that Australian built cars have to go through stringent engineering to receive a registration?
Your chassis is very nice. I like monocoque chassis’s, because they’re much more rigid and offer slightly more interior space. Also the Alulight construction is very simple, using wood tools to cut and shape. You can also use the Lotus/Alcoa panel bonding construction, with one-part epoxy and Ejot screw rivets.

Page 1 of 707 for Automotive : Gardner Web

That is an ambitious target. I will keep an eye out for your progress.
Yes building a bespoke chassis requires a raft of tests that would add more than $5000 to the build cost. This includes a torsional rigidity test and a handling and lane change test requiring a closed track and a professional test driver.

The torsion test must exceed the following: Torsional rigidity should be at least 4,000 Nm per degree over the wheelbase.
Here are some photos showing the torsion tests: Clubman Builders Resource - Australian Chassis Modifications

 

[sunworksco]

Very rigid....

 

[brian_]

... Unfortunately we are a third world country when it comes to EVs. The list is just starting to grow in the last two years or so but still not enough to make a significant impact on the salvage market...

Just listing the reasonable numbers as at Jan 14, in Queensland only and only cars (no motorbikes or trucks):
BMW i3 = 48
Hyundai IONIQ = 95
Hyundai Kona = 136
Jag ipace =50
Mini = 17
Mitsubishi IMiev = 39
Mitsubishi Minicab = 27
Nissan Leaf = 267
Renault ZOE = 22
Tesla = 1454

This is in a state with about 3,000,000 cars in total (again excluding trucks and motorbikes). So you can see our salvage market so far hasn't bee great for EV enthusiasts.

The province of Alberta, Canada, has about the same number of cars (with a slightly smaller population), with about 3,000 EVs... so EV adoption and the salvage situation are not much different here. For any Alberta EV builder getting components is likely to be an exercise in shipping from other provinces.

 

[brian_]

The new Tesla 46mm x 80mm tab-less batteries should be available from wrecks, as soon as they begin rolling down the road.
...
Tesla's new battery cell features a "tabless" design, which the company claims will provide five times the energy, six times the power, and 16% more range compared to its old batterycell.

They are not likely to be rolling down the road any time soon - this is one of Musk's random claims that has some connection to reality, but no real schedule. Real Tesla practice is to still use 18650 cells in the Model S and X, and 2170 cells in the Model 3 and Y... and all of the models which they have been promising for years but are not delivering (new Roadster, Cybertruck, Semi). From Autoweek:

"Musk indicated they're about three years away, actually putting them beyond the starting dates for the production of several new Tesla models. The Cybertruck is planned to enter production in late 2021, about the same time as the Tesla Semi. Both of these models, at least according to Tesla estimates, are expected to outpace the start of mass production of the tabless battery."​

The "tabless" design has value, but it is completely unreasonable to expect huge improvements. And what would "five times the energy" and "16% more range" mean, given that range depends only on energy? What it means is just that the cell is five times as large (so one-fifth as many fit in the pack and there is no improvement in total energy in the pack); of course the much larger cell can produce much more power, but again there will be many fewer of them in the pack so it's not a big improvement. The 16% range improvement (due to slightly better packing density or efficiency) is an unsubstantiated claim, but even if we accept that it is true, it's a minor advance as should be expected after years of development.

In the end, the tabless battery thing is just a distraction which is not relevant to this project, even if the battery isn't purchased for a couple of years. It is, on the other hand, an example of the ongoing battery improvements that are hoped for.

 

[galderdi]

The province of Alberta, Canada, has about the same number of cars (with a slightly smaller population), with about 3,000 EVs... so EV adoption and the salvage situation are not much different here. For any Alberta EV builder getting components is likely to be an exercise in shipping from other provinces.

All the states surrounding Queensland would have similar numbers. I would have been quite willing to source from another state but even they don't have much coming up. Unfortunately the nearest place to import any reasonable Evs is Japan (or maybe NZ) and we can only import via Surface shipping. The Air freight companies are unwilling to ship large quantities of Lithium batteries (knowingly). But it is slowly improving.

 

[sunworksco]

I’ve read many articles that say the CyberTruck will have the 4680 cells and the CyberTruck will be sold end of this year. I think Autoweek is somewhat biased.