Project background

With the continuous growth of demand for new energy vehicles, the application prospects of lithium batteries in the new energy vehicle industry are broad. Square batteries have become a mainstream battery packaging application due to their advantages in charge discharge rate, cycle life, safety, and other aspects.

In the production process of lithium batteries, there are mainly three stages, namely:lithium battery electrode film making process stage, battery cell production process stage, and battery module process stage, each of which has a large number of quality inspection requirements.

In the process of using machine vision technology for quality inspection, the most important thing is to combine the actual situation of the customer,Putting customers at the center, providing flexible customization solutions with the best cost and cost-effectiveness, helping enterprises reduce costs and increase efficiency。

After the hot pressing of the battery cell, the square shell battery needs to undergo laser welding of the adapter plate. After welding, the appearance quality of the weld seam of the copper aluminum electrode adapter plate needs to be inspected.

This article willtaking the detection of welding defects in the adapter laser welding machine as a case study, different solutions are designed and adopted for different defect requirements in the same process. Huahan Weiye combines 2D and 3D detection methods, not only achieving optimal cost, but also more accurately identifying and detecting welding defects, achieving better detection results.

Testing items

1) After welding, burrs, burst point detection, weld spacing, weld width, weld protrusions, pit defects, etc. of the adapter;

2) Welding deviation and breakage of adapter pieces;

Solution

• HyperShape 3D detection software

For burrs, burst points, seam protrusions, burst point defects, and pit defects after welding of the adapter, the Huahan HyperShape 3D detection software is used. The 3D camera is installed on the module and scanned sequentially to scan the copper aluminum electrode adapter weld seam. When the weld quality of the product is detected to be unqualified, an alarm signal is output.

Burr height detection

Use a height detection tool to detect the height of welding burrs using the maximum value.

Burst point detection

Use defect detection tools to detect burst points and highlight the burst point defects on the height image using a reference plane.

Weld gap

Using two straight lines to detect the distance between welding lines, the scanned grayscale images of aluminum and copper sheets have good contrast and can be used for distance detection.

Weld seam width

Use the distance between two straight lines to detect the width of the welding line.

Weld length

Use block tool to detect the length of the weld seam, extract the height mutation areas at the upper and lower ends of the weld seam through height images, and calculate the topmost position of the outer rectangle in the upper area and the bottommost position of the outer rectangle in the lower area. The difference between the two is the length of the weld seam.

• InnoVision 2D detection software

Due to the deep background texture of some products, there are a lot of noise points when using 3D images for edge detection, which affects the accuracy of edge grasping. In order to reduce the impact of noise and abnormal edge points, this project uses a 2D vision system to detect size and deviation, while 3D mainly detects defects in the height direction.

During the testing process of the plan, the 2D imaging effect is less affected by different colors and specifications of light sources, and the imaging effect is stable. Therefore, the combination of broken welding and 2D detection results is better.

Welding deviation detection

Using contour or grayscale matching tools for pre positioning, the ROI area of the line tool automatically follows the initial positioning correction, and the distance between two lines tool is used to fit the straight lines of the upper, lower, left, and right edges of the weld seam. The length, width, and spacing of the adapter weld seam are automatically measured.

Welding break detection

Use block or defect detection tools to detect defects such as weld breaks based on image feature differences.

detection result

The camera scans along the welding trajectory direction at a speed of 150 mm/s, which can meet the production capacity of 18PPM;

The Z-axis accuracy of the 3D camera is 0.005 mm, and it can stably detect defects such as 0.2 mm ² burst points and burrs;

The X-axis accuracy of the camera is 0.0125 mm, which can meet the accuracy requirement of 0.05 mm for weld length and width detection.

Detection time: ＜ 0.5 s, positioning/detection accuracy: ± 0.05 mm, missed kill rate=0%, false kill rate ＜ 0.5%, high accuracy, can effectively detect product quality.

Flexible customization solutions can help enterprises save a lot of labor and time costs, with high cost-effectiveness and the ability to ship in bulk.