Dimension & CCD Inspection Defects in Li-ion Pouch Cell Manufacturing

A CCD inspection station uses industrial cameras and image-processing software to measure the size and look of every pouch cell. It checks distances and positions in a few hundred milliseconds, far faster and more repeatable than a person with a vernier caliper.

Original EngineerX process map: the 11 main steps of pouch cell manufacturing in the order used on this site; the highlighted step is the subject of this article.

In many lines this station comes near the end of the assembly section, after sealing, side cutting and barcode printing. The cell shape is then final for the purpose of handling, so any dimensional fault found here can be sorted out before the cell goes through electrolyte filling, formation and packing into modules.

Why this stage matters

Pouch cells are soft and have no rigid can to hold their size, so tolerance control depends entirely on the quality of the cutting, folding and sealing steps. If a dimension drifts, the cell may not fit the formation tray, the degassing fixture or the module frame, and automatic machines may jam or place it wrongly.

Some dimensions also work as indirect safety indicators. An uneven seal height may reveal poor sealing, and a tab spacing error may show that the jelly roll has moved inside the pouch. CCD inspection therefore protects the customer, but it also gives early feedback to the upstream machines.

How the process works

The cell arrives on a conveyor or tray and is placed in the same position every time, using stops or locating pins.
A sensor detects the cell and triggers the lighting and the cameras.
Controlled light, often from several angles, makes the edges, seals, tabs and marks stand out clearly.
The camera captures images of the areas to be inspected, such as the full outline, the seal regions, the tab area, the degassing side and the barcode.
The software finds reference edges and points, converts pixel distances to millimetres using the calibration data, and calculates each dimension.
Each value is compared to the limits in the recipe. A pass or fail result is displayed and stored with the cell barcode.
Failed cells are rejected to a separate bin, and the result is reported to the production system so that trends can be reviewed.

Key process parameters

Parameter	What it controls	How it is checked
Camera calibration	Conversion of pixels to real millimetres	Calibration plate or master sample at set intervals
Lighting intensity and angle	Contrast of edges and marks	Light check with a reference sample and review of image histogram
Cell positioning accuracy	Repeatability of measurements	Fixture wear check and repeated-measure test on one cell
Tolerance limits	Pass or fail decision per dimension	Loaded from the cell drawing and controlled by quality approval
Threshold and edge-detection settings	Reliable finding of edges and marks	Golden sample and known-defect sample tests
Trigger timing and conveyor speed	Sharp, correctly placed images	Image review and trial runs after any speed change

The actual tolerances depend on cell design, materials and equipment supplier. They must always be taken from the cell maker's specification and drawing, not from a general guide.

Typical defects, causes and corrective actions

1. Cell length out of specification

Causes: incorrect side-cutting position, blade misalignment, mechanical drift in the trimming station, or laminate stretching during sealing. Detection: CCD length measurement against the drawing. Corrective action: re-teach the cutting position, replace or realign the blade, and check the sealing tension.

2. Sealant height variation

Causes: uneven sealing pressure, sealing-bar misalignment, wrong jelly roll position, or laminate thickness variation. Detection: CCD seal-region measurement, supported by seal strength checks. Corrective action: level and align the seal bars, verify pressure, and review the incoming laminate. Impact is weak seal zones and leakage risk.

3. Tab distance mismatch

Causes: jelly roll shifted in the pouch, tab bending during handling, or poor tab positioning. Detection: CCD measurement of the spacing between positive and negative tabs. Corrective action: check the jelly roll placement, add guides to reduce tab bending and review the transfer steps.

4. Degassing side length error

The extra pouch region used later for gas removal is too long or too short. Causes: cutting position error or fold position error. Detection: CCD measurement of the side length. Corrective action: correct the cutter and fold settings, because a wrong length can cause trouble in the fixtures that hold the cell during formation and degassing.

5. Side corner damage

Causes: worn blade, contact during transfer, or high trimming pressure. Detection: CCD corner-shape check plus visual review of rejected images. Corrective action: change the blade, smooth the transfer path and lower the pressure. Cracked laminate weakens the pouch.

6. Degassing corner defect

Damage or poor shaping near the degassing corner. Causes: blade wear, tool misalignment or rough handling. Detection: CCD corner inspection. Corrective action: repair the cutting tool and handling guide. It may disturb the degassing operation and raise the leakage risk.

7. Slot mark position error

Slot marks are used as reference features for alignment in later automatic operations. Causes: marking tool drift or cell positioning error. Detection: CCD position check against the nominal point. Corrective action: recalibrate the marking station and check the fixture.

8. Barcode dimension error

Causes: wrong printer resolution, scaling error in the print file, or surface deformation. Detection: CCD size check and a barcode grade or scan test. Corrective action: correct the print settings and the recipe, and flatten or clean the print surface. Unreadable codes break traceability.

Quality checks and measurements

Start-of-shift check with a master (golden) sample that has known dimensions, and a known-defect sample to prove the system can reject.
Regular calibration of the cameras with a calibration plate, repeated after any camera or lens adjustment.
Gauge repeatability checks: measure one cell many times to confirm that the system itself is stable.
Periodic comparison of the CCD result with manual measurement on selected cells.
Trend charts of key dimensions so drift is noticed before cells go out of tolerance.
Review of rejected images each day to separate real defects from false calls.

Safety and handling notes

Work in the controlled-humidity or dry-room conditions the plant requires. The electrode stack inside the pouch absorbs moisture if exposed.
Handle cells by the edges, avoid bending the tabs and never lay cells where tabs can touch metal trays or each other.
Stop the conveyor and use lockout before clearing jams, adjusting fixtures or cleaning around moving parts and pneumatic cylinders.
Bright lighting, and any laser used for marking or sensing, must be shielded. Follow the machine's eye-protection rules.
Any cell with a damaged pouch, leakage or heat must be separated at once into a marked container, and the area cleaned per the electrolyte spill procedure.
Keep the camera glass clean using the approved wipes and never use solvents that can damage the lens coating.

Interview questions

Q1. Why is calibration needed for a CCD system?
It converts pixel distances into real millimetres, so wrong calibration gives wrong measurements.

Q2. What is a golden sample?
A cell of known good dimensions used to confirm the system reads correctly.

Q3. What can cause a false reject?
Dirty optics, changed lighting, poor cell positioning or incorrect threshold settings.

Q4. Why does tab distance matter?
It decides how the cell connects in the module and may show a shifted jelly roll.

Q5. Why check seal geometry with a camera?
Seal width and height changes can indicate sealing process drift, which links to leakage risk.

Q6. What should you do if many cells fail the same dimension?
Stop and check the upstream station, such as cutting or sealing, rather than only sorting the rejects.

FAQs

Can CCD inspection replace seal strength testing?

No. It measures appearance and geometry. Seal strength needs a separate mechanical test as defined in the plant's quality plan.

Is CCD used only at one place in the line?

Plants may use vision checks at several steps, for example after the cutting and again at final inspection, as needed by the process plan.

How often should the system be calibrated?

The interval comes from the equipment supplier and the quality procedure. It is also repeated after any maintenance, lens change or impact on the camera.