P42A pack damage - WTF happened?

rich · September 11, 2021, 10:46pm

I need some help please to identify what went wrong with my 3 month old 12s2p pack (maybe 20 cycles). I used it on my leightweight dirt jump MTB so the impacts/vibrations are quite hard. I charge with 8A and discharge usually between 30A and 60A (sags about 5V) and seldom short 80A peaks (6-7V voltage sag). The battery was spot welded with 0.2mm nickel with kweld at 42.5J. One layer, serial connection width about 30mm. (More info and battery building pics in my board build thread for those who are interested)

Prehistory what happened 1 day before:

The troubles probably started when I lost my battery during jumping although the heat shrink tube of the battery was firmly glued and additionally secured with duct tape. Apparently the heat shrink tube tore during the landing and the battery flew away. I guess the battery landed on the side because the JST port of the BMS is damaged. There are small stones everywhere and hard packed dirt.

At home I checked cell voltages and visual damages. I opened the heat shrink tube of some cells with marks a bit but found minor scratches only.

Most damaged was the BMS

.

Then I charged all cells from about 4V to 4.2V with a balance charger and checked cell voltages afterwards. I used a thicker heat shrink tube and a lot of tape and glued it back in place.

Now the main problem:

I was jumping and had fun in dirt park and drained the battery until the end (usually about 42V = 3.5V/cell because of massive sag) but this time it felt a tad shorter.

At home I wanted to charge again but recognized this melted hole. There is fishpaper and several layers of kapton and fabric tape in between so the heat must have been incredibly high to create a hole there. Checked cell voltages, all at 3.7V except group 11 at 2.9V

It took a while until I realized that I could lift the nickel with a knife and that the spot welds on both negative group cells somehow vaporized.

Then I measured the single cells of group 11

What’s going on here ?
Does anyone have any idea what happened here?

I’m quite sure that some damage happened when the battery dropped but what?
The obvious reason would be that the spot welds broke but I am not sure for 2 reasons. First the other welds are still strong, it took some force with the pliers (rolled) to remove the nickel further. On the serial connection on the other side I even damaged a negative pole, too much pulling force, now there is a small buckle (no idea how you salvage cells guys).

The cells are glued with a strong glue similar to sikaflex. In theory it could flex a tad, who knows.
BUT the second reason and what baffles me the most is that all the spotwelds vanished. What I mean is that after removing the nickel there is sharp material left (and sticking out) on the cells and on the nickel strip where the spotwelds were. But on the 2 negative poles of group 11 there is no material, the surfaces are very smooth like polished (also the nickel strip). Here you can see and compare what I mean

Aftermath mockup, the cell with 2.95V (group 11) is not far from the spot where the BMS is damaged so there was definitely some impact.

Thanks for reading
Any ideas?

soketsu · September 11, 2021, 11:37pm

this is a very typical end result of abused packs. Frequent 5A sag on lipo is fine but not on li-ion. P42A sure can output 30A per cell but that doesn’t mean it would stay healthy on such sustained high amp discharge.

TheBoardGarage · September 12, 2021, 1:20am

Well, I’m not certain, but could this have something to do with it?

It looks like you essentially had no padding or protection on it, and it suffered a lot of abuse (both electrically and mechanically) before some point of the pack gave up and led to failures elsewhere.

A13XR3 · September 12, 2021, 3:32am

I kinda enjoy failure analysis so here’s my take:

Everyone setup is going to be different, but the kweld manual recommends 50J and I weld at 58J for .2 thickness nickel. So I’m going to call that insufficient.

PVC shrink is very brittle, it will tear pretty easily if there’s and cut or hole to start which may have been small and unnoticed then resulting in the full failure on this jump.

This is why having multiple spotwelds is important, because all the current is flowing through those dots. In this case it’s likely you had a single weld point, maybe two, still holding on to the cell carrying that entire load.

The nickel popping off in such a way proves to me that you didn’t have sufficient penetration on the spot welds. They did not vaporize, they were not actually there to begin with.

The cell has broken free, the higher voltage one that is. So it was not draining. You were pulling full current and rapidly draining the single cell that was still attached. You stopped just in time before it could have been a big issue.

None of these welds are sufficient, not even close really. The tiny dots left behind are only slightly better than the nothingness. You should have to tear the nickel off destroying the strip in a way that looks like wolverine removed it, they should never pop off by prying.

I have faced this issue myself when I was first experimenting with .2 nickel and with P42A specifically. Whatever alloy comp they use on the cans is more difficult to weld to, especially the negative side. This is why testing you welds is so important. I would assume you can remove every piece of nickel on that pack that easily and retire the packs until further repair or remediation can be done.

TheBoardGarage · September 12, 2021, 3:38am

Sheesh, man. That’s some dark shit right there.

AviatorEsk8er · September 12, 2021, 3:39am

Looks like my Sunkko does stronger welds than these . Good writeup Alex!

jack.luis · September 12, 2021, 4:00am

Fuck yeah man. Got a 1600a truck battery for my welder bc of this

marketthrowaway · September 12, 2021, 4:00am

Sunkkos are notoriously shit and the kweld is well known as the best available spot welder for our purposes.

pls don’t bring your shit takes everywhere.

AviatorEsk8er · September 12, 2021, 4:03am

T’was a joke. I’m well aware of what a Kweld is.

sk8l8r · September 12, 2021, 9:12am

Can’t help but wonder if NESE based batteries might be better for high physical abuse packs. A 12s2p wouldn’t be too much plastic - last time I pulled cells from one I was amazed at how hard it was to get them out again, I think I would trust those over welds and silicon or whatever.

https://18650.lt

rich · September 12, 2021, 2:43pm

Thanks for the replies guys!
It was late night and I fell asleep.

Before I reply/discuss other interesting aspects and posts I want to thank first @A13XR3 for opening my eyes! After reading your post I did more tests and unfortunately I have to agree with everything you say.

I was really under the misconception that the battery is very solidly built. I mean it is solid but as you say not the spot welds On the left side removed with pliers, right side lifted just with stanley knife.

At least I couldn’t rip off the nickel on positive poles and also the blade broke. Then I tried it a last time with more force and fuck…

I have to admit my spot welds are probably the shittiest on earth

So true!
Also crazy when considering I’m jumping since 3 months with this bomb between my legs.
As you say when the battery dropped the upper cell of group 11 broke free.

I was very scared in the beginning but it seems the nickel (and surface of the negative pole) got extremely hot and not the cell itself which I thought first. In dirt park I was jumping for about 40 minutes without breaks, draining up to 60A and more from like 2 single spot welds from a single P42A cell. Not to mention the up to -23A current from braking for 0.5-1.5 seconds. Crazy numbers, lucky me…

My I ask with which power source/battery?

This is the setup I use

I think one issue why the spot welds are shitty is the Lipo. It is 2 years old and noticeable weak. I had to re-charge it a lot even for this tiny pack. 2 years ago when the battery was fresh I did a 12s6p VTC6 with 0.2mm nickel welded at 35J.

I chose 35J because of the tests with a stainley blade before

I thought the burning marks on the bottom of the blade at 40-50J were too much.

After removing with pliers it looked like this. 40J was very hard to remove and 50J with excessive force only.

I have to say the nano-tech battery was much stronger so these 35J welds on VTC6 should be better

I use this battery since 2 years without any issues but should test/inspect the welds as well. Only problem I am not mentally ready for this.

But back to the P42A pack, those 42.5J welds look like 10-20J on stainley blade after removing the nickel (again new Lipo)

Crazy!

Need new Lipo, the 3s 5Ah nano-tech 60C-135C Lipo is out of stock but the nano-tech plus 70-140C is available. It is cheaper, looks cheaper but better specs? Strange

At least I have 24 test cells now where I can find the best Joule setting

I am just not sure if it is dangerous to use too much welding power and also what happens when I pull with too much force with pliers?

YeetMeat · September 12, 2021, 3:22pm

It won’t matter, the kweld uses a certain amount of energy per weld, hence the use of joules instead of a pulse length duration to weld. That being said, having a beefier battery means you need to charge less. Although on a welder that uses a pulse duration like a malectrics a beefier battery will make a big difference.

The difference in materials between the cell and blade are too different to get an idea of how strong of welds you need. Use the cell you will be welding on to test welds (ideally it will be discharged all the way)

You want all your welds to take this much effort to remove

Try not to do this when you’re taking nickel off a welded cell. Try peeling it off by getting some pliers on the nickel, and rolling the nickel over itself so it peels off. Sometimes when you pull the nickel off it’ll leave the cell kinda rounded off

ShutterShock · September 12, 2021, 3:33pm

Ya unfortunately that blade is not a very good representation

I used 54J for 0.2mm on p42a

Halbj613 · September 12, 2021, 3:36pm

i was using around 44J with p42a and got really nice welds on 0.2

glyphiks · September 12, 2021, 3:42pm

I have no science to back this up, but i believe it does matter.

I recently just changed batteries from lipos to an agm battery and I noticed a big difference in weld quality at the same settings. I even had to dial the settings down a bit with the new battery.

YeetMeat · September 12, 2021, 3:59pm

Tbh I can’t speak from experience, what I said was based on what I’ve read 🤷. I’ve always thought that it just draws however much power from the battery specified on the welder and that was the whole story. I suppose I’ve oversimplified it

I’d take your word over mine lol

rich · September 12, 2021, 4:34pm

True but in this case it was the group 11 with half capacity.

My guess was that it still lasts longer than lipos and I can build whatever size I want. A 12s3p is too heavy for my taste. I use very conservative 45A motor/battery settings each esc so no power horse at all. I have no continious discharge because I jump with this board therefore the drain is more like in waves. I made several logs with lipos before and decided to go for 12s2p.

Here an example of easy jumping with 12s2p P42A

here a bit more aggressive jumping with 12s2p P42A

Maybe the charge amps are a bit too much

True and normally I would agree with the padding. I always use/used padding from all sides inside the enclosure and preached padding. But also wondered how it is possible that they hold well in other applications without or with tiny padding (like some pre-built boards, e-bikes or whatever).

I built this board for jumping so it must be lightweight. It sounds stupid but an enclosure adds weight Also when I built it I prepared all long 12 AWG motor wires until I realized and measured how heavy they are and went with 14AWG I can’t remember the weight difference but it was a lot.

However 2 years ago I glued my lipos with the 2mm double sided self adhesive EPDM foam because I was in a hurry and it hold well for several months until I had to re-glue them. BTW the enclosure was 0.15mm thick silicone sheet. I thought I can do the same with the 12sp P42A and get rid of the velcro straps I always had for safety, bad idea.

It is
Also duct tape was a really stupid idea because it simply breaks off when pulled hard.

Damn that beast must weight a ton

could be interesting

Makes sense, I read at the time that it’s very close to Li-ion cells and I didn’t want to sacrifice a new cell. Lesson learned.

My poor english means exactly what you describe when I say pull (rolling with pliers)
Peeling it off sounds much better and is what I did.

This happened when peeling off the poor welds on 3 cells (without pulling), it’s hard to get on picture.

Thanks for the info

Yeah for sure it matters and the kweld doesn’t like all of the ultra-high discharge lipos. That’s why it’s the approved 3s 65C-135C nano-tech in the setup. If I remember right @Acido had troubles with a 70C graphene where the internal solder joints melted or something like this.

The description states that the 70C nano-tech plus is a further development and improvement of the 65C nano-tech so it should be alright to use I hope.

Xpress2016 · September 12, 2021, 4:36pm

Molicel P42a are particulary hard to spot on the negative side compare to other cells.
Maybe because it was a cell made for vap but not to be spotted.
At 55J on kweld all my Molicel packs are perfect.
But I only need 40 for other cells.

rich · September 12, 2021, 4:40pm

Do you (and others) use the same Joule settings for negative and positive poles of P42A?

Xpress2016 · September 12, 2021, 6:54pm

Yes because if it welds great on negative side, it will on positive.
And negative side is less fragile than positive.