Because of their power density, lithium-ion batteries as used by electric vehicles (EV) are subject to strict quality monitoring. Industrial computed tomography (CT) increasingly is being used to detect defects and internal changes throughout a battery’s lifecycle, while CT-data analysis and visualization software provides functions that allow a deep look into the inner workings of energy storage devices. Methods from the field of artificial intelligence also are becoming increasingly important.
Although CT analysis cannot reveal the electrochemistry within the cell, it can illuminate the "mechanical" inner workings. A thermal runaway (internal fire) can have mechanical causes. Conversely, electrochemical processes can change mechanical conditions. The insights that CT can contribute are increasingly being used by research institutions and battery manufacturers. Both geometric measurements and material testing are now feasible through advanced software.
Monitoring anode overlap
Beginning in R&D, CT-data analysis can help test prototypes for their optimum structural designs. Wall spacing, seals and tolerances, chemical cell distributions and the housing each can be captured and quality-checked for their role in reliability and designed electrical output. Next, due to the very complex process of manufacturing these batteries, CT-based analyses make sense during production, for example, to remove defective components from the process chain at an early stage.
With a high-resolution CT scanner, irregularities in the layers of the electrode packages become visible. Delaminations are a typical phenomenon, as are localized foreign particles such as residues resulting from the cutting process and from welding during battery assembly. Foreign particles pose a risk of short-circuiting.
An important internal dimension to be monitored during manufacturing is anode overlap. The anode is always dimensioned so that it overlaps the cathode. This is to counteract lithium plating and possible damage to the cell. Lithium plating means that pure Li is deposited in the anode; it then is no longer available for the formation of ions. But a constant anode protrusion requires high precision in manufacturing. It is defined by the manufacturer and can be checked using CT analysis software. Finally, engineers can use CT for inspection and forensics in the after-sales phase to determine the cause of a device failure.
AI detects defects
When examining batteries using CT, quality engineers face a challenge: The interesting structures in the gray-scale images provided by the CT scanner often have very low contrasts. This is due to the low-density differences of some materials. In addition, the films and coatings of the cell packages are very thin and close together. It sometimes can be difficult to determine which irregularities can be interpreted as defects, scattered radiation or artifacts.
The central question: which voxel is a defect voxel, which is not? (A voxel is the smallest 3D element in the CT model, similar to pixels in 2D images.) Even experienced quality engineers come to different interpretations. The only thing left for them to do is to adjust or vary the scan parameters of their system accordingly and, if necessary, to pay special attention to certain regions of interest (ROI). This is the conventional approach – but it has its drawbacks. It depends on the operator and therefore is individual, but it requires additional time which has to be added to the scanning time.
Since the sometimes-filigreed structures of the cell interior require high-resolution scanners – and therefore long scanning times – things quickly become tight, especially when accompanying random sample inspections must be carried out within a specified cycle time. In such cases, artificial intelligence (AI) and its deep-learning subset proves to be particularly effective, particularly when examining cast-metal workpieces where similar tasks have to be performed, as in the case of battery cells.
The application of deep learning/neural networks (NN) in CT-based defect detection delivers very accurate and rapid results. The network needs something like a memory for this to occur; it must be "trained" with defect data. But where does this data come from? There are basically two ways to generate it: First, the simulation of artificial defect data based on the model of real defects. Special software is available for this purpose and the physical effects that occur during scanning can be imitated in this way. The result is an artificial but accurate pool of data. Secondly, the defect data also can be derived from actual components. In this case, the defects have to be detected manually. This approach requires a larger number of real objects.
The method that is most suitable must be decided on a case-by-case basis. Customers can commission the creation of individual NN programs. Here, algorithms are designed to exactly fit their problem. If need be, developers can provide a certain number of battery cells, both intact and containing critical faults, to refine the analysis.
When the trained NN is applied, the irregularities in the real scans are compared with the defect data "in memory." The NN then identifies similarities. It reliably answers the question of what a defect voxel is – and what it is not. An advantage for series-part inspection: The method is also very accurate at lower resolutions, as they occur in the case of shorter scan times. It also could allow inline inspection of battery cells in the long term.
CT analysis software is a time-saving and critical quality tool with a proven history of success from R&D to production inspection.
Pascal Pinter is product manager, Material R&D, at Volume Graphics.Continue reading »