Tubular Li-Ion Battery Design Feasibility

[jjsylvestre]

Hello,

I'm brand new to EV's and building battery packs so I would like to get some input from you guys on my battery design and where you see this possibly failing. Attached is an image of a split open view of the pack and the complete pack. There are a total of 12 cells summing to 48v. Each cell is composed of 12 18650 Li-Ion batteries that I got from here http://www.ebay.com/itm/251569014683?_trksid=p2057872.m2749.l2649&ssPageName=STRK:MEBIDX:IT

Each cell of 12 batteries is connected in series to the next using a aluminum plate (half circle). Each half cylinder is separated by a nonconductive membrane. Note: the top plate will be interconnected (in the Cad file, it is not shown connected). Here is what I am worried about failing- connecting the cells. I was not planning to do any soldering, spot welding, etc. What I had in mind was that the weight of the batteries would be enough to provide a solid connection between the battery terminals and the aluminum plate.

Do you see issues with this?

 

[IdleMind]

Great minds think alike, I've been working on something similar and just posted my work, then I saw your thread. I think you need more than just weight to get reliable contact force. I will be thinking about your idea. And don't let anyone tell you that febble also think alike.

 

[Hillhater]

Find a better source of cells. That one is telling lies. ,

 

[jjsylvestre]

Yeah I figured those technical specs were really exaggerated. The batteries are brand new so I just hope they're somewhat decent. Its too late because I already ordered them

I was thinking of soldering small springs onto the terminals of the batteries. If I did that though, I might as well just solder the batteries straight to the Al plate.

 

[Hillhater]

Before you assemble a large pack, you need to test and capacity match all cells....or expect early disappointment !
I fear you chances of finding matched cells from that source are not good.
And the capacity will certainly not meet expectations
You may well be wasting a lot of time and effort because of dud cells.
Oh, and you do understand that soldering onto the cells is considered a bad idea ?
...and that soldering onto aluminium is not easy !

 

[jjsylvestre]

Before you assemble a large pack, you need to test and capacity match all cells....or expect early disappointment !

How do you "capacity match" cells? I am not familiar with any methods to test the capacity of a battery.

And the capacity will certainly not meet expectations

Yeah, I'm expecting 1000-2000maH, which is plenty for my needs

Oh, and you do understand that soldering onto the cells is considered a bad idea ?
...and that soldering onto aluminium is not easy !

Yes, I understand it is not advised to solder directly to Lithium Ion cells, but with the equipment I have, I don't see a viable alternative except for some sort of mechanical connection like a spring.

 

[okashira]

For the love of god, do not build an ebike pack from those cells.
Using those cells is about as ill advised as relying on gravity for an electrical contact.

Other then those two glaring issues the design could be good.

 

[Hillhater]

How do you "capacity match" cells? I am not familiar with any methods to test the capacity of a battery.

Start reading....
http://endless-sphere.com/forums/viewtopic.php?f=14&t=53017&start=25#p880083

 

[okashira]

capacity matching is the wrong answer here. When you receive those cells, please dispose of them safely or return them for a refund. Do not make a pack or post anything about those cells. Lithium ion cells and 18650's already have a bad enough name public-ally.

 

[liveforphysics]

Yeah I figured those technical specs were really exaggerated. The batteries are brand new so I just hope they're somewhat decent. Its too late because I already ordered them

I was thinking of soldering small springs onto the terminals of the batteries. If I did that though, I might as well just solder the batteries straight to the Al plate.

Most of the cells I've seen that claimed over 4Ah had under 2Ah.

 

[TheBeastie]

Based on that price I can't help but predict they will be something similar to these.
Actually what I find surprising on a lot of the dubious brand 18650s is even there 1ah or 2ah guaranteed versions are just a rip off anyway compared to just getting a proper samsung cell, generally anyway.
[youtube]eOshOXcSkDA[/youtube]

Your cells might be good for a testing prototype battery pack which is what your probably want to start off with anyway.

 

[Syonyk]

Most of the cells I've seen that claimed over 4Ah had under 2Ah.

The highest "honest" capacity I've seen measured for a standard 18650 is about 3200-3300mah. I consider anything over 3500mah unless proven (usually by the candlepowerforum folks) to be an utter lie, and nothing but single-cell recycled rubbish suited to selling to ecig types who will buy anything, as long as it's cheap and says a big number. It's fine, as long as it's the only battery in a device, but they're unwise to parallel and downright hazardous to put in serial.

 

[jjsylvestre]

Alright everyone, I get that these cells specs are not what is listed. Even if these batteries were 1000maH, I'd still have a 12 maH battery

capacity matching is the wrong answer here. When you receive those cells, please dispose of them safely or return them for a refund. Do not make a pack or post anything about those cells. Lithium ion cells and 18650's already have a bad enough name public-ally.

This helps me in no way at all. I'm not going to throw away $150. Please make constructive posts.

The highest "honest" capacity I've seen measured for a standard 18650 is about 3200-3300mah. I consider anything over 3500mah unless proven (usually by the candlepowerforum folks) to be an utter lie, and nothing but single-cell recycled rubbish suited to selling to ecig types who will buy anything, as long as it's cheap and says a big number. It's fine, as long as it's the only battery in a device, but they're unwise to parallel and downright hazardous to put in serial.

Thank you. This is much more what I was looking for.

Based on the responses, I'm going to see if I can make a return.

 

[Syonyk]

Alright everyone, I get that these cells specs are not what is listed. Even if these batteries were 1000maH, I'd still have a 12 maH battery

I hope you mean 12AH. And if you want that capacity, there are cheaper ways to obtain it when dealing with honest cells.

The big problem here, and the reason you're getting some slightly horrified responses, is that you are literally relying on luck to keep your pack from (at best) dying, or at worst bursting into aggressive flame. And you're starting with a source known for not being lucky.

If you're lucky, all the cells are the same chemistry, and more or less the same capacity. I wouldn't count on that for cells marketed as such.

If you're not testing capacity on the cells, you literally have no idea what you're getting and putting in the pack, and are relying on luck and your BMS (which I see no evidence of - so just pure luck) to keep the pack intact.

The worst case situation would be that you put your pack together, most of the parallel cells (your 12 18650 layers) have 16-20AH, and one has 8-10Ah. That type of variation is entirely possible if you don't know the cell capacities, because you could put a few nearly dead cells together. Unless you've tested capacity, "holds a charge" is a worthless metric.

What happens there is that your pack works perfectly fine for small amounts of power use. It might sag slightly more than expected, but without individual cell level monitoring, you have no way to see that one cell is dropping in voltage more than the others, if you're looking at aggregate pack voltage.

It gets worse, though, if you take a longer ride. The total pack capacity of that hypothetical bad pack is only 8-10AH - but you'll still have good voltage off the pack after that, because the other layers are holding up the dead layer, while draining it fully (below 2.7v), and continuing to push current through it. You've now got a truly dangerous situation, as you've got cells that are dead, still being driven, and being actively damaged, but you have no way of knowing this. I don't know if you can invert a lithium based chemistry or not (drop them below 0V), but you've certainly got a system that would try!

Recharge the pack after that longer than expected ride, and if you're lucky (again), it won't take a charge - that layer just won't pass much current, and you can figure something is wrong. If you aren't lucky, you're now charging deeply discharged lithium cells. At best, you have radically reduced capacity and are likely to damage those cells more. At worst, eventually, they have enough and catch fire after charging, putting out enough thermal energy to set the rest of the pack off. YouTube has some good videos of what this does.

This helps me in no way at all. I'm not going to throw away $150. Please make constructive posts.

He did suggest returning them for a refund...

It's possible to use those cells safely in something that uses one cell, or at most, uses a few in parallel, and has a voltage cutoff. If they're all the same chemistry. Which, sadly, is a bit of an if. Something like an ecig will probably be fine on them. But building a large, series pack with literally unknown cells is not a wise idea.