Anonymous
5/28/2025, 7:09:22 AM
No.2920199
>>2920191
>l nickel/steel strips
The problem is the properties of the Ni and Fe are both oxidized, annealed and hardened where the metals are joined and the result can be quite brittle, and likewise, if there's any Fe in there, extra corrosion in wet environments. This is a general weld problem that's happened forever on all kinds of welds. We see "popping" out of cell terminal welds where a little hole right around the weld point on the buss strip or the cell endcap.
> rare for a single cell to fail by itself, and it's even more rare for it to fail and not take the entire parallel group with it (either immediately or over time).
It's the exact opposite. It's almost always a single cell, even excluding situations like the famous DeWalt pack fiasco that had a design fault that ran the monitoring of the pack off a single cell and drained it. Cells are optimized for different things. Usually you want cells designed to tend towards failing "open" like Y1 class capacitors for safety reasons. Unfused cells that fail closed/shorted is an unmitigated disaster.
If a cell fails open, there's no issue. As was mentioned somewhere earlier, it's common to have fusable links, although aluminum is a really bad choice for high vibration environments, too after the infamous Trans-am fiasco in the late 60s, early 70s.
>>2920130
Yeah, LiFePo4 is probably a better choice, and extremely popular especially for /rcg/ use, I don't think anyone has to "sell" us on it at this point.
>l nickel/steel strips
The problem is the properties of the Ni and Fe are both oxidized, annealed and hardened where the metals are joined and the result can be quite brittle, and likewise, if there's any Fe in there, extra corrosion in wet environments. This is a general weld problem that's happened forever on all kinds of welds. We see "popping" out of cell terminal welds where a little hole right around the weld point on the buss strip or the cell endcap.
> rare for a single cell to fail by itself, and it's even more rare for it to fail and not take the entire parallel group with it (either immediately or over time).
It's the exact opposite. It's almost always a single cell, even excluding situations like the famous DeWalt pack fiasco that had a design fault that ran the monitoring of the pack off a single cell and drained it. Cells are optimized for different things. Usually you want cells designed to tend towards failing "open" like Y1 class capacitors for safety reasons. Unfused cells that fail closed/shorted is an unmitigated disaster.
If a cell fails open, there's no issue. As was mentioned somewhere earlier, it's common to have fusable links, although aluminum is a really bad choice for high vibration environments, too after the infamous Trans-am fiasco in the late 60s, early 70s.
>>2920130
Yeah, LiFePo4 is probably a better choice, and extremely popular especially for /rcg/ use, I don't think anyone has to "sell" us on it at this point.