Holding and Compressing Expanding Lithium Cells

Warning: this is a Rabbit Hole topic, with a lot of conflicting data and opinions out there. Though, if one enjoys rabbit hole diving, this thread is a good start. It is peppered with some technical info from EVE, especially as reported by 'Dzl' and 'ghostwriter66'.


But before digging into the whats and hows, let's be clear on the priorities for this van's cell retention system: 1st is safety, 2nd is safety, and 3rd is... safety 😏
  1. In case of an accident / rollover, cells must stay put and not cannonball through the cabin
  2. In case of an accident / rollover, the cells' terminals should not short circuit with their surrounding
  3. During normal operation, the mechanical stress on the cells must be well within the supplier's specs
The EVE LF280 Rev E spec alludes to millimeter-ish expansion during charging, and increased longevity from 2500 cycles to 3500 if a ~300kg compressive  force is applied on the case when the cell is in a ~30% to 100% SOC.

 
Expansion: I did not expect that behavior. So, I couldn't help but want to check that with my own eyes !
Indeed, a 4 corner measurement between the fully discharged and the fully charged states on one of my LF280s confirmed expansion, by ~1.2mm. The discharging / charging rate was a low 40A / 0.14C. In the following pic the cell is under the height measurement plate (i.e. cell thickness):
LF280 LiFePo4 cell expansion: 1.2mm btwn ~0% and 100% SOC

 
~300kg clamping force on the broad side of the cell: this is supposed to slow down delamination in the 'jelly stack' over time, as the cell experiences charge / discharge cycles.
 
This said, c'est donner de la confiture aux cochons(*)... We don't care much about the longevity gain for this van ! Because 2500 cycles is already close to 7 years. And that is way too pessimistic in our case, as:
  • the highest charging current per cell will only be ~57A (120A Victron Multiplus charger + 50A solar / 3P). That's barely ~0.2C
  • highest discharge current per cell: ~83A (250A / 3P), i.e. 0.3C
  • cycling will be limited to a ~20-95% SOC range (LLVC well above 2.5V)
  • most daily cycles will not even reach 50% discharge
  • part of the year 2 packs will be used in parallel, cutting down the current per cell even more
Those very mild use conditions will push the cycle life well past 15 years. At which point the cell should exhibit only a 20% capacity loss over brand new. In other words, no need to lose sleep over it !

 
But wait... cells have to be firmly held in the battery box. So some form of clamping is needed anyway. And here lies a critical requirement: LF280 cells should NOT be rigidly clamped, on their broad side:
  • If initially clamped when in a 0 to ~30% SOC they'll need some room to expand. EVE recommends against rigid clamping LF280 with no give
  • Conversely, if initially clamped in a ~80 to 100% SOC they'll be skinnier when empty and the rigid enclosure / clamp won't be holding them firmly anymore. That's a no-no for mobile applications
The worst expansion scenario for this van will be 5 cells side by side, i.e. up to 6mm / 1/4"... ☠️  Rigid clamping is out !
 
For this reason, and to meet the safety goals above, I am going with some form of constant-ish force clamping. Increased cell longevity will just be a bonus, it is not a driver.
 
(*) 'it's like giving good jam to the pigs'
____________________________________

Comments

Post a Comment