Fraunhofer Research Institution for Battery Cell Production FFB 
SiPaFeb
Research project, "SiPaFeb - Ensuring particle-free production of battery cells"
The goal of all activities in this project is to develop a software and hardware system for lithium-ion battery cell manufacturing in an industrial setting, which displays the particle contamination level of the production line inline and in real time, so that the cell manufacturer can take measures to minimize particle contamination before contaminated cells are produced or even shipped to the customer.
To build this demonstrator, the following research questions must be answered:
1. Identification of the most relevant particle sources: “Which processes/other sources emit potentially critical particles?”
2. Demonstration of the distinguishability of particles from different sources: “How can particles be assigned to the various
particle sources based on geometric/other parameters?”
3. Development and validation of the measurement concept and infrastructure (demonstrator setup): “What must a measurement concept and the associated infrastructure look like to identify and localize critical particle contamination as early as possible?”
Demonstrator description:
The demonstrator to be built consists of three parts:
1. Inline sensor technology that monitors process parameters influencing particle contamination in real time
2. IT infrastructure that uses AI to predict contamination of the produced battery cells based on real-time information
3. Instructions for regularly performing offline measurements to recalibrate the prediction
The goal is to ensure that the cleanliness of the produced battery cells and thus their safety can be guaranteed through the implementation of the demonstrator. The demonstrator will be designed so that it can be implemented in any battery cell series production line (after adaptation to the specific plant design). Therefore, attention is paid to the suitability of the components used for large-scale production. To demonstrate the universal applicability of the measurement concept and the demonstrator, it will be implemented at both EAS and Fraunhofer FFB PreFab. Extensive preparatory work is required to establish the aforementioned sub-areas.
Work Plan
The entire project is divided into a total of seven work packages:
In WP 1, an initial cleanliness audit will be conducted in the EAS production facility and at Fraunhofer FFB PreFab. The results of these cleanliness audits will serve as the basis for conducting measurements and for selecting and installing sensor systems. Objective: Potentially relevant particle sources are identified, and a measurement strategy for all particle sources is defined.
In parallel, in WP 2, aiXbrain and Eurogard will establish the digital infrastructure, including interfaces, to enable the processing of all generated data in the subsequent work packages. Objective: Data from all relevant data sources (see WP 3) can be processed and analyzed.
In WP 3, measurement data is collected in four parallel work packages at the EAS production facilities and at FFB PreFab: All data is centrally aggregated in WP 4 by Eurogard to enable subsequent evaluation and AI-supported analysis of the various data sources. Objective: All recorded data is stored centrally so that it can be evaluated collectively.
- MeaStream supports the installation and operation of plant-integrated inline sensor technology for real-time monitoring of particle contamination
- Cleancontrolling examines components for particle contamination in its in-house cleanliness laboratories to establish a correlation between particles in the environment and particles in the product
- EAS and Fraunhofer document events in the production process to account for factors such as setup procedures or shift changes.
- Objective: The measurement concept defined in WP1 has been implemented at both EAS and FFB PreFab, and data is being continuously collected.
In AP 5, the collected data is evaluated and examined for correlations, for example between inline sensor data and component data. This is done manually by Cleancontrolling and Fraunhofer, as well as AI-supported by aiXbrain. Objective: Correlations between data sources can be established manually and automatically, so that the status of particle contamination in production can be displayed in real time using inline sensor technology.
In WP 6, all activities related to measuring sustainability criteria are consolidated. Among other things, a life cycle assessment is conducted here. Objective: All implemented measures are analyzed and evaluated in terms of their impact on sustainability.
WP 7 includes not only project management but also all activities related to the publication of results. Additionally, based on the project results, a training concept on the topic of “Technical Cleanliness in Battery Cell Manufacturing” will be developed, which will become part of the
“European Battery Cell Learning Lab” (ELLB). Objective: At least two publications on the project results will be produced. In addition, the ELLB will offer a specialized course on the topic of “Technical Cleanliness in Battery Cell Manufacturing.”
Utilization of results
To make the abstract description of the project activities a bit more concrete, here are four examples of the research findings:
- The metal particles generated by slitting increase continuously over time until the particle density drops abruptly. This behavior repeats at periodic intervals (“sawtooth curve”). A comparison with the metadata shows that the scheduled replacement of the blade causes the drop in the curve, which is why the replacement interval is being increased.
- The particles generated during slitting have a characteristic shape, which is why a function is added to the microscopy software to predict, based on the geometry of individual
- particles and with a certain degree of probability, whether a particle originates from the slitter or not.
- By comparing data from particle traps and inline sensors, it can be empirically confirmed that both methods can be used to determine the number of particles present in production.
- Switching from particle traps to inline sensors enables real-time monitoring of production, whereas analysis using particle traps requires waiting up to two weeks for a result.
- During shift changes, the particle load in the ambient air increases significantly. An examination of employees’ clothing reveals that the standards currently maintained in battery cell production are insufficient to prevent contamination of the production process. The examples mentioned are only potential project outcomes. Of course, the actual project results cannot be predicted in advance
Application:
- Product for inline particle monitoring.
- Gain in knowledge regarding technical cleanliness.
- Insights that can be incorporated into industrial projects.
- Improvement in product quality at a major cell supplier.
- Two publications.