Following on from Ford's battery pack, we know that once the initial design is finalized, we optimize it through various methods. In module design, parameter optimization involves a crucial issue: cell expansion. We've also discussed battery selection paths before.
This expansion causes the module to extend longitudinally, significantly impacting the insulation of the end plate and the entire module. This is particularly problematic when the module is pressurized.
Single fixing method: laser-welded end plate
Straps, long screws, etc.
The compression of the internal foam or insulation material also needs to be considered; estimating it or adding sensors later is always a bit troublesome.
This can be traced back to Boeing's GS Yuasa box shell, where the expansion of the monolith under various conditions caused a short circuit in the external structure, which in turn led to more failures.
Note: This is the most extreme case. Under the existing chemical system, such extreme conditions should not need to be considered. However, considering some extreme situations at the end of the lifespan, further misuse at the end of the lifespan may make safety incidents more likely to occur.
Therefore, strictly speaking, we need to understand the expansion of batteries under various conditions, and also how much difference exists between different optimized conditions and cycle experiments conducted in the laboratory at room temperature or high temperature. This is something GE, the University of Michigan, and Ford are doing, with the goal of breaking through the traditional (V, I&T three-parameter) design SOC threshold for applications by collecting detailed temperature and pressure data of batteries. Previously, only introductory materials from the initial stages of the project were available.
"ControlEnablingSolutionswithUltathinStrainandTemperatureSensorSystemforReducedBatteryLifeCycleCost"
The final report of 2015, obtained by chance, revealed that it had yielded considerable results.
Unlike now, where there are fewer and fewer temperature points, the test vehicle can be equipped with as many points as possible.
By achieving higher precision, laboratory techniques can be applied to automobiles.
The assembled pack looks like this.
To simplify things, embedding a pressure sensor, manufacturing some prototype packages, configuring them via BMS, and then testing them on a real vehicle to see the actual durability results might be a good option.
This is still achieved by modeling individual units and then performing pack testing.
First, build the model using clamps.
By reducing the number of individual units or opening a larger SOC window, I think adding a pressure sensor seems worthwhile. Incorporating a measurement channel into development testing is something worth trying.
