Biggest Battery Fail Ever - A123 AMP20

[themadhatter106]

Ok so maybe it's not the biggest fail ever but please take this opportunity to point and laugh and also maybe give advice (if you feel so inclined).

Many cells on only one half of my battery pack got damaged (holes in the foil on the bottom) and leaked electrolyte as well. The acrylic that I was using to compress the cells cracked too but didn't puncture the cell as far as I can tell. I discovered this when I had not used my bike for months and wanted to use it, the BMS was cutting out. One cell on the end was at 0V but all of the other cell voltages were fine even the ones that turned out to have nicks in the foil.

When I took the pack off my bike it was making a hissing sound. "Well that can't be good!" It didn't catch on fire or even get hot; I'm thanking LiFePO4 for its ability to resist such catastrophes.

This is a split 20S A123 pack I made mounted with each half on either side of my front fork. I compressed the cells using acrylic, steel flat bar and hose clamps. The two half packs were covered in padding and held in a custom mount. Everything was fine for about a year until I had to change my mount because I was changing to a new fork. There are two differences past this point, the new mount has the battery not perpendicular to the ground but at an angle instead, the rake angle of the fork. Also, On one half I noticed that the Styrofoam insulation I used as padding was starting to degrade and I didn't have enough to cut more for both half packs so Instead I used soft mail packaging foam of the same thickness.

It's odd that one half was undamaged and the other half was completely frocked up! I think the Styrofoam must be better at protecting the battery than the soft foam padding. Although, the soft foam was still intact with no holes. Also I think having the pack at an angle to the ground may have put more stress on one side of the pack or caused the pack to rub more over bumps.

Once I replace the botched cells I want to solve this problem.

I am thinking of still using hose clamps but using a different material which won't crack. I want to use 1/4" aluminum since it will provide the most even distribution of force without needing really big and thick plastic plates. The only problem is that it's conductive so I will want some type of insulation for safety. Maybe a sheet of rubber so that when the pack swells with each charge cycle the rubber can squash down a little to accommodate the thicker cells? I could make the aluminum extend a bit below the bottom of the cells to prevent them from getting damaged there.

I also don't have any insulating material between the cells. I should probably add some since I see some spots towards the middle in between some of the cells in the pack which are disconcerting. Any suggestions?

 

[davec]

personally i don't think AMP20 was meant for the hobby market
i like the format easier than paralleling tons of tiny cells and is more space efficient
but engineering a proper case for it , is tricky.. i think the proper way to go is to try to source the gear from a123 / platters /compression bands and work with that....

the good news is there's finally documentation
http://www.formula-hybrid.org/wp-content/uploads/A123_AMP20_battery_Design_guide.pdf

 

[themadhatter106]

Hi,

Thanks for your reply.

Yes, you are right about the hose clamps tending to bow the material. At first I planned to just use acrylic but realized that it was bowing badly. Then I went to the 1/8" steel bar reinforcements to prevent the bowing, that helped a lot.

I think that a whole sheet of 1/4" aluminum would be better than the acrylic with the steel reinforcements. I could even use the 1/8" steel reinforcements on the aluminum to further stop it from bowing if needed. I can't think of any better way to clamp the pack down other than using hose clamps. I imagine that using nuts and bolts on the edge wouldn't be much better since I would need to make the aluminum sheet bigger than the cells which I think would cause more bowing.

 

[liveforphysics]

Corrosion started anywhere on the outer pouch foils means game over and time to recycle them.

 

[999zip999]

I have the factory end compression plates they weigh a ton. So they just sit on my shelf I have 920 Cycles on my two 12s packs as 24s. Just like new. The pack is heavy as it is.

 

[cell_man]

Compressing big cells like this is very difficult to do. I tried and for sure big hose clamps do not cut it, not unless the plates are very stiff and IMPO, you really need to have some sort of a profile on the plate, for the band to sit into (which means the plates need to be thick). Plates just tend to bow and all the pressure is on the side, nothing in the middle (using bands).

Reading the various requirements in that A123 document... Well that is a lot to do and consider for a DIY build. It also does not lend itself too well, to smaller packs, as you still need to have these special end plates etc, but now you only have 10 cells or so to compress, not 30 or 50, as you might have in a larger module. Then the case gets so heavy/big, compared to the cells contained within. It just gets out of hand, if you really try to fulfill all the requirements. They are great cells though, loads of current and that flat discharge curve is so nice.

I had thought about using bent, spring steel clamps, that would be clamped down until the are flat. This would tend to put the pressure onto the mid point of the clamp, which would better exert an even force over the plate/cell. I even went so far as to source suitable steel, bend it and run some tests. In the end, it was just too much for me at the time.

Thankfully, we moved away from the big 20Ah A123 cells to the A123 26650, then 18650 Samsung and life became simpler

 

[amberwolf]

It's odd that one half was undamaged and the other half was completely frocked up! I think the Styrofoam must be better at protecting the battery than the soft foam padding. Although, the soft foam was still intact with no holes.

Styrofoam in general can be fairly uncompressible, compared to any kind of soft open-cell foam. The more compressible it is, the less good it does in absorbing shock and vibration, probably because it has already compressed a fair bit just from the weight of the pack itself.

I've used a variety of padding materials to secure things on my cargo bikes, and the only ones that actually worked are the less compressible types.

Styrofoam worked well enough in the cargo pods for general purpose noise absorption, heat insulation, and bump padding, but it degrades and crushes permanently once suffiicient weight or impact occurs.

Mousepads, made from thin closed-cell foam, did very well to pad my EIG battery pack inside a 50cal ammocan, and it's been going for a few years now on the bike and the trike.

Some blue closed-cell foam works ok, too, but it compresses a lot, and like styrofoam it degrades the same way under same condtions.

Exercise mats, the really thin ones, also work ok like the mousepads did--I usually just layered them 3-4 layers thick, giving around 1/2" of padding.

For certain things there are very dense closed cell foams I've saved bits of from various scrapped items, almost always black, but it's always tiny pieces maybe an inch or two long, 1/2-1" thick. It'd be a really good option if I had larger pieces, but I've never looked to see where I could buy it new (I figure it will cost too much, especially fi it's from online due to shipping).

All the open-cell foams I tried really didn't do anything to help, but was generally always completely intact.

 

[themadhatter106]

I had thought about using bent, spring steel clamps, that would be clamped down until the are flat. This would tend to put the pressure onto the mid point of the clamp, which would better exert an even force over the plate/cell. I even went so far as to source suitable steel, bend it and run some tests. In the end, it was just too much for me at the time.

This sounds like a good method but bending the steel into a proper spring sounds difficult. The other thing I am considering is that the amount of bowing is proportional to the clamping force. This could be one reason not to overdo the amount of clamping force.

I know that you used shrink tubing on your A123 AMP20 packs, do you know how much pressure the shrink tube exerted? It couldn't have been much and your packs seemed to hold up fine. Maybe at the currents we draw you really don't need much pressure? Looking at the graph in the document it appears that too little pressure shortens the life of a cell more than too much pressure. Although with too little force you have a big problem. The cells can shift around over bumps and eventually break off a tab. I had this happen to one of my outer cells a while ago.

This is why I am thinking of using a thin sheet of neoprene rubber in between each cell as an insulator and for extra grip to prevent shifting with a small clamping force.

Styrofoam in general can be fairly uncompressible, compared to any kind of soft open-cell foam. The more compressible it is, the less good it does in absorbing shock and vibration, probably because it has already compressed a fair bit just from the weight of the pack itself.

Good to know. Yes, the main disadvantage of the Styrofoam is that it needs to be periodically replaced. Eventually it just gets compressed and beat up. I didn't know about the exercise mats or mouse pads. That's a good idea. I'll need to take a look online too at various closed cell foams to see what my options are.

 

[cell_man]

Yes, i had those kind of thoughts about how much to bend the steel, cross section etc. I don't recall what i did exactly, but i do recall having some approx figures for the force exerted and the total area, to get a certain pressure. Frankly, I didn't get that far with it. I'm 1 guy, with pretty limited time/resources at the time and it just got a bit too much, to be practical, plua the realisation, that by the time you did all that the energy density wouldn't be great....

There was no awareness of the requirement to have this high pressure when i got these cells, it was more than i signed up for TBH, when even finding a reasonable way to connect them in series, is already kind of tricky.

 

[Punx0r]

What outer casing was there? All the cells look damaged on the bottom edge so, assuming they were stood upright it could be them bouncing up and down on the bottom of the outer box, or maybe instead water got in there and corroded them.

 

[Doctorbass]

Just buld a sinple aluminum box with 3/32" thick sheet and it will work...

Here is my 16s2p built 4 years ago still working perfectly and cells remain in excellent shape, no real compression, just use a uniform foam layer and it will spread soft pressure on these:

https://endless-sphere.com/forums/viewtopic.php?p=709069&display_history=true

I always use aluminum box for lipo and all poutch cells... when done properly it protech and keep cells in original shape and avoid them to poff..

Doc

 

[999zip999]

It looks like they've been sitting in the bottom of a boat in salt water. There's gotta be something more there. I mean it took some time to get rust like that

 

[miro13car]

like Doctorbass I built simple Alu box from thick enough aluminium plates.
These plates if thick enough will squeeze 6cells really good, I havnt tried more than 6.
I have two subpacks 6 cells each.
I put white packing thin foam between each cell. they are /cells/ happy that way.
Thousends of kilometers made on my two A123 20Ah 12S packs testify to quality of those pouches.
One of the pack I made of HongKong A123 QC rejects is still going strong - over 1000 cycles.
BUT , BUT I do not discharge more than 15Ah - never.
If you try to squeeze like 12 A123 pouches - thats another story.
example shown in pictures - very bad way to compress pouch cells.

 

[999zip999]

Where did these cells come from ? Are they half tab ?

 

[liveforphysics]

The PE film layer on the outside of the aluminum pouch laminate will have pinholes. If you let water wick between cells at different potentials with no insulation layers between them, you will experience electrolysis driven through those pinholes until it leaks electrolyte and ends up looking like the corrosion seen on your pack.

 

[themadhatter106]

These are short tab cells from OSN Power. Some of the cells had some really weird corrosion or something on the edge.

View attachment 2



Some of the damage on the bottom.





I thoroughly looked at all of the cells in the other half of the pack and they look good, like new. Maybe the electrolyte drying in between the cells might have had something to do with the corrosion?