Use of Sensors in Laser Welding Applications
- Apr 1
- 3 min read
Modern production lines aim to achieve zero defects to avoid wasting high value goods. Therefore, sensors are required to ensure correct alignment and positioning of welding components. Misalignment and rotation can be compensated by adjusting the laser beam path. Sensors are not only needed prior to the welding process but also during the process itself to monitor and respond to irregularities and potential defects. Depending on the workpiece, sensor type, and defect, the process can be adapted in real time to ensure high-quality welds, or the workpiece can be marked for further inspection and rework.
The Right Sensor for the Right Application
Workpiece Alignment
High-precision positioning of workpieces can be time-consuming and expensive. Modern camera systems and computer vision algorithms can help reduce this cost in serial production lines. Workpieces can be positioned within a defined tolerance and fixed using flexible clamps. The camera system detects characteristic geometries of the workpiece. Points derived from these geometries are compared to a reference, and the weld trajectories are adjusted accordingly. Depending on the system, the camera can be placed coaxially or off-axis to the laser beam path, providing flexibility in implementation.
Seam Tracking
Sensors can be used not only to align complete workpieces but also to track individual weld seams or edges and compensate for manufacturing tolerances and deviations. For this purpose, line sensors and tactile sensors can be used. These sensors can be easily integrated when using a robot to move along the planned trajectory. In both cases, the sensor observes the area in front of the laser beam, detects the correct weld path, and provides feedback to an offset-compensating linear axis or laser scanner. In the case of a wider gap in a butt weld or a lift-off in a lap joint configuration, the laser process can be adjusted accordingly, for example by increasing wobble amplitude or laser power.
Weld Monitoring
One option to monitor the welding process is photo diode-based devices. These measure light emissions from the welding zone across different spectral ranges. The number of spectral ranges varies depending on the manufacturer, but in principle, welding defects can be detected by comparing the photodiode signal of the current weld to defect-free reference welds.
Figure 1 shows an example of a recorded welding process. Each color in the graph represents a different spectral range. In the first section, the weld performs as expected. The rising signal amplitude indicates increasing heat accumulation in the workpiece. In the second section, melt is expelled from the welding zone and the signal rises until saturation. This may be caused by evaporation of contamination, entrapped gases, or full penetration of the laser beam with material evaporation on the backside. Afterwards, the signal becomes irregular with intermittent spikes, indicating an unstable collapsing keyhole. In the final section, the process stabilizes again.
By analyzing this information, the process can be adjusted to prevent such defects, or workpieces can be selected for further inspection, reducing the number of required test samples. Additionally, if multiple weld signals deviate from the reference, this may indicate a systematic issue, such as contamination of the protective window in the laser optics. A trained operator can respond accordingly, reducing costly failures and machine downtime.
Implementation of Sensors into a Civan Dynamic Beam Laser System
The Civan Dynamic Beam Shaping Laser is very flexible in sensor integration and system integrators can easily mix and match with industry proven sensor solutions.
Cameras for workpiece alignment can be installed stationery and communicate with the beam steering device, like galvo scanners or robots. In applications where the optical head is mounted on a robot, the camera can also be mounted on the robot to increase flexibility. In scanner-based welding, a coaxial camera can capture the region of interest prior to welding.
Seam tracking devices, such as line sensors, can be installed on robots or linear axes to measure ahead of the welding zone.
Photodiode sensors are best placed coaxially to the laser beam so that the welding zone is continuously monitored. A schematic illustration of a possible integration into the optical head of a Civan Dynamic Beam Laser is shown in Figure 2.
The standard port of the coaxial camera can be adapted with a suitable beam splitter which reflects and transmits light the relevant wavelengths. Examples of an actual setup using a 4D.TWO and Laser Weld Master sensor are shown in Figure 3 and Figure 4, respectively. Figure 5 shows the integration of a Plasmo to a Civan Dynamic Beam Laser. Here, off-axial fibers which are connected to photo diodes are pointing at the workpiece together with a co-axial fiber, which is also connected to the photo diodes.
Overall, sensors are essential tools in industrial laser welding applications, enabling easier workpiece positioning and consistent weld quality. Civan Dynamic Beam Lasers support the integration of the most relevant sensor types, allowing for adaptive and controlled welding processes.
