The development of new materials with different properties and processing needs and requirements, along with the integration of new features, as sensors and mechanisms for self-healing, will require a significant rethink of cell design, including remanufacturing issues. The redesign of cell architecture is essential to drive both competitiveness and sustainability, while maintaining or even increasing the energy density.
The availability of a new generation of breakthrough battery materials will open a new world of opportunities for innovative battery technologies. However, these new battery technologies will need to face at least two main validation phases. First, they will need to prove their potential at the prototype level; second, the feasibility of their upscaling into industrial level will need to be assessed.
Nowadays cell design follows a trial and error approach, and is limited to some standard formats (cylindrical, pouch and prismatic hard-case), though there are some incipient activities in applying modeling to the cell design, which opens the possibility to explore new cell formats.
Battery manufacturing is a quite well-established art today. This is particularly true for lithium ion batteries, seen as the reference technology for present and the near future. The three main phases: electrode production, cell assembly, and cell finishing consider several steps, like mixing, coating and drying, slitting, calendaring, vacuum drying, electrolyte filling, etc. In spite of this well-organized sequence of steps, current trial and error approaches at design and manufacturing should be overcome, also including radically new eco-design criteria minimizing scrap, use of primary energy and producing zero or near to zero emissions. In this line, current multiphysics modelling can be of a great importance in battery design and manufacturing to:
- Accelerate new cell designs in terms of performance, efficiency and sustainability, coupling multiphysics models to advanced optimisation algorithms in the artificial intelligence (AI) framework, as well, as inverse cell design, which represents a crucial step for autonomous battery design optimization, as it connects the desired properties to specific cell configuration, electrode compositions and material structures as targets to synthesize, characterize, and test.
- Accelerate the optimization of existing and future manufacturing processes in terms of cell chemistry, manufacturing costs and sustainability/environmental impact, although more effort is needed to develop multiscale physicochemical computational platform of the full manufacturing process chain of lithium ion batteries.
All these impressive efforts together with the rapidly growing computational and algorithmic capabilities, in particular in the field of AI, call us to go even further. The computational simulation of cell design and manufacturing process of new generation of batteries, for example, integrating interfaces discovered through the BIG MAP concept and/or cells including sensing and self-healing functionalities, will certainly pose new exciting challenges for multiscale computational science.
In a future scenario, current trial and error approaches should be avoided and cells and manufacturing processes need to be ”smart”, giving them a digital identity creating a digital twin, which is a virtual counterpart to a physical object.
The advances needed for future cell design and manufacturing processes can be summarised as follows:
- Introduction of new functionalities, like self-healing materials/interfaces, sensors or other actuators, cell eco-design and alternative cell designs.
- Flexible manufacturing processes and flexible, high precision modelling tools for the optimization of processing and conditions and machine parameters. In this way, human labour, trial and error, and waste products will be minimized; development of real-time models for the processing of electrode pastes and the performance in the cell (digital twin for cell manufacturing).
- Development and validation of multiphysics and multiscale models on cell manufacturing processes capable to provide accurate understanding on each steps of the process.
In short term: This would be done starting from state-of-the art information, and focus will be the battery cell design methodology. This would include improvement of simulation tools – multiphysics models- with the goals of reducing the computational burden and implementation of current AI techniques through deep learning and machine learning methods for cell design..
In medium term: Input is expected to come from BIG, MAP, sensing, self-healing, recycling and other innovation areas that would be integrated into the process. Also, the methodology will be adapted to manufacturability of new battery technologies, with the launch and implementation of the AI driven methodology to manufacturing after the developments made at cell level design: Modelling --> AI --> Manufacturing including new techniques, as well as the creation of digital twin of a cell manufacturing process.
In long term: Full maturity of the methodology is expected, closing the loop by means of integration of the cell design and manufacturing design sub-loops as a fully autonomous system with interface from BIG-MAP. Parts of this methodology can be progressively made available to the industry before the full package is made available as a commodity to a new state of the art.