Battery construction etc

[enjoythesilenc]

"The range of the Stark VARG is similar to a full tank of gas on a 450. The bike can complete a full heat at MXGP intensity, or up to 6 hours of trail riding. A full recharge takes 1-2 hours. There is 6kWh of battery capacity, and our battery system is the most energy dense in the world. The state of the art ‘flying V’ concept creates a very compact powertrain structure. contributing to the optimized centre of gravity. The battery system also includes a patent pending pressure relief system, which makes it IP69K fully waterproof. The Stark VARG is the first electric motorcycle to be the performance leader within its category. Deliveries start September 2022, reserve yours now for just $100 / €100: https://starkfuture.com #ChallengeYourself #StarkVARG Follow us on Instagram: https://www.instagram.com/starkfuture... Follow us on Facebook: https://www.facebook.com/starkfutureo... Follow us on Twitter: https://twitter.com/StarkFutureBike Contact us: support@starkfuture.com"

This is starting to grow on me. I wonder if Mr. Waas or our own fsfs is a fan of the Scorpions

Scorpions, Metallica, Megadeth Featured In 'Flying V' Documentary

Members of Scorpions, Metallica, Megadeth and Judas Priest are among the stars appearing in a new doc, Flying V, about the famous rock guitar

www.udiscovermusic.com

The video shows a half case with room for 200 cells. I read that the battery is 360 volts nominal. That gives us a cell configuration of 100S 4P. They also gives us the 6kWh capacity which might equate to 6000Wh/3.6V/400 cells= 4.167 Ah per cell

Here is a possible component, 4200mAh cell from molicel. At least the math adds up?

https://cdn.shopify.com/s/files/1/0697/3395/files/INR21700P42A-V3-80092.pdf?2425

 

[Oded]

What interest me most is if the individual cells can be replaced when going bad.
In the Alta we learnt the hard way that it's mission impossible.

 

[Rashid510]

What interest me most is if the individual cells can be replaced when going bad.
In the A lta we learnt the hard way that it's mission impossible.

To remove individual cells would require an engineering solution that would not be cost effective.

 

[enjoythesilenc]

Does the fact that each parallel group has only 4 members make this configuration less tolerant of a single weak cell within its ranks? Maybe the probability of having an actual cell failure is really low as long as it is robustly connected and protected. I haven't looked to see if that molicel candidate would be overstressed at 80 hp yet (or 1 hour charge time)

 

[mr.skywalker]

The molicel P42A is capable of really high output(45A) so based on 100s4p configuration, its capable of 64.8kw or about 86hp. Its the go to cell for surrons so not surprised if its what they use

 

[The T]

I think that you calculation is not as right as a manufacturer calculate.
A Manufacturer would calculate wit the max. Voltage of the system. (4.2 V per cell).
For this version the would have a bigger range of Cells.
The possible Cells would be Sony Murata VTC6a 21700, Samsung 40T 21700 and Molicell P42A. That would make the things easier.

So the more important and interesting information would be the connection of the Cell to the housing and the heat transfer at the "slippy fingers" methode.
Which material the use for isolation, and why they use magnesium instead of aluminium.

 

[Vextrek]

IP69 ingress protection is very impressive.

 

[Silenthorse]

To remove individual cells would require an engineering solution that would not be cost effective.

What makes you so sure? I've been licking the tops of 9v battrees since I was a kid. All Battrees is da same and ain't got nutin that komplikated Goin on inside em. My grandpa taught me all aboutm and he was one da best plumber's i ever knowed.

 

[enjoythesilenc]

Can the cells be connected in ways other than wirebonding? What's the current state of the art? Do cylinder cells want to swell and deflate as they are cycled?

 

[Vextrek]

Can the cells be connected in ways other than wirebonding? What's the current state of the art? Do cylinder cells want to swell and deflate as they are cycled?

We use a Sunstone Pulse Arc welder (P250i) to connect the Molicel P26A cells with .25mm thick cooper sheets. The pack performs very well with no heat issues. Using a di-lithium BMS with 10 thermistors and we routinely pull 600A of current. Disclaimer: Pack does not utilize cell-level fusing so risk of thermal runaway is greater.

 

[Chadx]

We use a Sunstone Pulse Arc welder (P250i) to connect the Molicel P26A cells with .25mm thick cooper sheets. The pack performs very well with no heat issues. Using a di-lithium BMS with 10 thermistors and we routinely pull 600A of current. Disclaimer: Pack does not utilize cell-level fusing so risk of thermal runaway is greater.

Gary,
Does your description apply to Varg battery, Kart battery, or both (same tech)? Assuming the last, but wanted confirmation. Thanks.

 

[Vextrek]

Gary,
Does your description apply to Varg battery, Kart battery, or both (same tech)? Assuming the last, but wanted confirmation. Thanks.

Thanks. Description applies to the go kart battery we are currently using as well as an electric motorcycle battery we are finishing now. Have not made one for an ALTA but would love to do it soon.

 

[enjoythesilenc]

Nice pics of your copper. I'm pretty sure Stark is connecting both + and - on the same end of the cell. The negative end is sitting in a pocket in the magnesium case so no connections possible on that end, i'm assuming.

 

[Rashid510]

Thanks. Description applies to the go kart battery we are currently using as well as an electric motorcycle battery we are finishing now. Have not made one for an ALTA but would love to do it soon.

So Alta modules and lots of High voltage modules utlize wirebonding or laser welding nowadays. Wirebonding isnt that hard to do, but the equipment and setup costs are the main pain points.

 

[enjoythesilenc]

To remove individual cells would require an engineering solution that would not be cost effective.

Assuming cells are sturdily connected, what rate of cell failure would you predict in a heavy use scenario like the Stark claims it can withstand? Are cells in the Stark more or less likely to be stressed than those in an Alta (or Tesla)? Were there many instances of Alta (or Tesla for that matter) cells actually failing in the course of normal operation?

 

[Matt]

Assuming cells are sturdily connected, what rate of cell failure would you predict in a heavy use scenario like the Stark claims it can withstand? Are cells in the Stark more or less likely to be stressed than those in an Alta (or Tesla)? Were there many instances of Alta (or Tesla for that matter) cells actually failing in the course of normal operation?

I am only aware of two packs with a cell failure vs an other failure in the pack such as a broken wire bond. Mine was one of them. I suspect the cell was a dud from the factory though. I am sure others may know of more.

 

[enjoythesilenc]

I had an idea about a connection scheme that might make individual cell replacement doable but I guess its a dead end if 1. premature cell failure rates are very low 2. anticipated improvement in cell performance in a given cell format is unlikely

2 failures from premature cell death out of 1000 Alta packs is .2%. Stark has 4/5 the numbers of cells so has less likelihood of being struck by a dud cell. Assuming the Stark and Alta cells have similar rates of being duds, 16 out of the 10000 Starks on order could potentially end up with a zombie cell. Obviously a lot of assumptions here but might be worth further speculation etc

 

[Rashid510]

I am only aware of two packs with a cell failure vs an other failure in the pack such as a broken wire bond. Mine was one of them. I suspect the cell was a dud from the factory though. I am sure others may know of more.

So in general Cells are sorted based on resistance and voltage. Most of the industry gets wrapped cells due to high order reqs. Alta had a supplier agreement to have pre-sorted, naked cells to make the process of manufacturing easier (it sucks to unwrap cells). If a cell is a dud its probably due to a bond lift which created a discharge path for the cell to drop below 3.5V. The testing we did before the bike allowed us to tease the issue, but unlike Teslas with 4000+ cells, when an Alta loses a cell it does create an issue.

Assuming cells are sturdily connected, what rate of cell failure would you predict in a heavy use scenario like the Stark claims it can withstand? Are cells in the Stark more or less likely to be stressed than those in an Alta (or Tesla)? Were there many instances of Alta (or Tesla for that matter) cells actually failing in the course of normal operation?

Based on the designs that have been shown...21700s installed straight to the pack, and potentially laser welded (maybe wirebonded) the cell fail rate is probably going to be in the industry standard of 1-2%. Unless they have a cell supplier whos not the best with keeping there ranges tight.

Cells failing is all based on the factory manufacturing and also what issues could arise in production. Most of the "A" grade cells go to the big suppliers (Power tools/Automotive/Vaccums) . Then "B" grade cells go to the smaller companies/vendors who make smaller EV products. Its really hard to get into the top level. Back in the Tesla days the failure rate was about 5% a month for bad cells from Panasonic..which then dropped to about 3%

 

[Matt]

So in general Cells are sorted based on resistance and voltage. Most of the industry gets wrapped cells due to high order reqs. Alta had a supplier agreement to have pre-sorted, naked cells to make the process of manufacturing easier (it sucks to unwrap cells). If a cell is a dud its probably due to a bond lift which created a discharge path for the cell to drop below 3.5V. The testing we did before the bike allowed us to tease the issue, but unlike Teslas with 4000+ cells, when an Alta loses a cell it does create an issue.

Interesting. Thanks for sharing. The dud cell in my pack still had all of the wire bonds secure with continuity between. I still don't know why there is a dud cell but it dragged the voltage of a p group down very low. It also dragged all of the p groups before that one in the sting down a bunch. I think it was p group 9 with the dud cell so 0-8 p groups were all lower than 10-20

 

[testus191]

I would like to see something along the lines of this. Do you think your Sunstone could do this Gary?