Tesla is relying on massive parallelism and not drawing much through each cell. How much current is going through that wire? Not much. Yet they still use fuses because it makes sense to do so.
Also, its a wire, not magic.
Nickel strip has four times the resistance of copper. If our spot welders could handle copper we would be using copper.
Ballpark ampacity for nickel strip is CSAx19/4
And fusing is not a cure-all. Cell characteristics, p-group size and esc setting all play a part in determining fuse size.
Do fuses generate heat? Absolutely. Are we cooking eggs on p-groups? No.
By estimates in my 10s4p pack, If had done this I would be dropping 8 volts at 40 Amps.
This is compared to the 4.5V I am dropping right now.
That is literally 25% less range.
And it should be trivial to determine real-world fusing current of either nickel strip or tinned copper at various voltages.
The issue is not fusing current, the issue is arc supression. When the wire breaks, you expect the current to stop, but parasitic inductance has other ideas.
A Ludicrous P100D is been seen to generate over 570kW during max acceleration, which at 400V gives you 1,425A. The battery pack is arranged in a 96S 86P configuration. 1425/86 gives a peak discharge per cell of 16.56A, all going through that wee aluminum wire.
No, we dont expect the current to stop.
No doubt. But 95٪ of the time drivers are not hammering away at their batteries like that.
Ideally for CLF, you want a fuse material that is reasonably electrically conductive, not very thermally conductive, and has a fairly low melting point. Copper does not fit two of these criteria very well, granted. But it’s really hard for the average hobbyist to get tin wire (fairly ideal), or weld aluminum wire (also fairly ideal.) Thus we turn to nickel and copper, because they’re easy to get and easy to work with.
Im planning on soldering aluminium wire from nickel tabs on the cell to nickel strips.
This discussion is interesting though, if I hadnt gotten as far as I have into the pack, I would probably ditch the idea.
Please check the math
Whats the conductor length? Total length of series connections?
I don’t understand the point of cell level fusing. if the esc or motor were to short and a big current rush were to happen id think a regular overall pack fuse at something like 200amps would be good but what does a cell level fuse protect from? it would have to be some kind of short within the pack right? id think just building the pack well and without a bunch of hair-like conductors all around seems safer.
if some speck of metal crosses into the positive and shorts a cell the cell level fuse isn’t going to help, if anything I can totally see one of those cell level fuses breaking off and being the short.
teslas have so many cells in the battery, its like 100x more cells and…100x more likely to have a short within the battery so makes more sense to have the cell fuses. maybe im missing something.
Cell level fusing saved probably my whole board, and even all the cells in the battery when I made a stupid mistake directly shorting the battery. Re-welded the 8 fuses(12s4p shorted first 4 in p group, and 1st four in p group on the backside) and that battery is now on third charge working wonderfully
the length of the conductor from the battery terminal to the bus.
I shorted my battery pack four times while assembling it xD
30 mile range every charge :).
I tried applying my past physics knowledge and made some charts regarding nickel strips and the current needed to melt them in a given amount of time. Use at your own risk (the numbers look right but I haven’t double checked) and maybe someone will find it useful. Keep in mind that the strip cooling from being in contact with the cell has NOT been taken into account.
My understanding is that CLF only really makes sense in 4-5p and larger packs. Basically if one of your cells fails short circuit (unlikely but it does happen) and you have no CLF, then all the other cells in the P group will dump current into that one cell, and it will likely go into thermal overload and cause bad things.
If however you do have CLF, then when one cell goes bad and the others dump current into it, the fuse will blow and remove the current path. Thus the bad cell only has its own stored energy generating heat, and not the addition of all the others in the P group - much less likely to reach full thermal runaway and explosion/fire etc.
Ideally you want your CLF to be rated higher than your maximum single-cell discharge current (duh), but low enough that the other cells in the P-group dumping into the short cell will blow the fuse quickly.
With my pack (20s7p of 18650PF cells), I used 24AWG tinned copper wire, 10mm long. The fusing current is well above the single-cell current, but the 6 other cells can easily supply enough to blow it.
@Gamer43 The entire fusing setup adds a calculated 2.4 miliohms to my battery pack internal resistance. At my full discharge current of 100A, that’s a voltage drop of 0.24 volts and a power dissipation of 24 watts across the entire pack.
Given the internal resistance of the cells in the pack is <35 milliohms each (<100 milliohms over the whole pack), that’s a ~2.4% increase - negligible.