We often see stepped fast charge limits as shown in this BMW iX3 graph.
Often the result of a limited test regime applied by the cell supplier (in this case CATL) to establish the maximum charge current for the cell.
These limits are them applied by the BMS team and without more data they have to drop back to safe charge current limits between these points.
A lookup table with linear interpolation is not hard to do in the actual vehicle/ BMS, in fact it is even easier to program than a floor interpolation (round down approach) as seen above. Those measurement corner points where current steps down often come from expensive testing like half cell experiments, or simply trial and error:Hin Kwan Roland Wong
The stepped profile is usually remnant of an early stage profile design where it is easy to design a profile with 5-6 corner points, then program a cycler with just those 5-6 points where current changes, and run an aging test that will ‘validate’ the profile over 3-6 months.
After such a profile passes the aging test, however, nobody then knows if it is acceptable to join those points smoothly without introducing additional aging. You never know whether those shaded areas (see below) has any lithium plating risk unless you run another aging test over 6 months, which will cost a lot of time and money.
Unless of course one has a detailed and validated electrochemical model of the cell, which defeats the purpose then of such a crude trial and error approach, as it allows one to design a smooth optimized profile in the first place. But it’s unrealistic to have a good electrochemical model to begin with.
That improvement doesn’t sound like much, but that is just related to the extra exergy added which is quite a small area that is being filled in.
This is a very simple approximation. However, charge times are a really important purchase metric and improving that time by more than 4% is probably worth the effort in doing a more detailed analysis of the cell charge limits.