Figure 3。3。 Illustration of a typical laser scanner sensor [3] mounted ahead of the welding torch

If the gap varies, there are a few things to control: (i) the metal deposition should be done in a way to obtain a constant weld shape, (ii) a large gap is sensitive for burn-through and a control action could be to lower the travel speed together   with

lower wire feed speed (weld current), (iii) a too low current may however result  in a lack of fusion and cracking in the weld and to keep a higher current a weaving motion of the welding torch can be applied to avoid the problem with burn- through。 This shows that the control scheme needs to consider many issues, of which some are boundary conditions and some are counteracting each other, leading to new ways to perform the weld operation。

In robotic welding of thicker plates, the welding is usually performed in several passes, e。g。 one root pass and additional passes to fill up the weld joint。 In such cases, the use of a tracking sensor can be applied in several ways。 Normally, tracking is applied for the root pass。 During this operation the robot records the weld path and subsequent passes can be overlaid with respect to the first path based on the actual weld joint geometry。

Figure 3。4。 Typical standard joint types。 Left column: fillet and corner joint。 Right column: lap, butt and V-groove joint [3]

A laser-based seam tracker is typically mounted on the weld torch and has the weld joint in its field of view some distance ahead in the weld path direction, see  Figure

3。3。 This means that the robot must use one degree of freedom to keep the sensor in alignment with the weld joint during welding or, alternatively, use a separate motion so that the sensor can rotate around the weld torch to maintain the alignment relationship between the sensor and the weld torch。 It should also be noted that the seam tracking sensor must measure and deliver target positions of the weld torch continuously and that these must be time stamped and stored in a buffer for later use by the robot controller。

In order to use the data from the laser scanner, weld joint features must  be extracted from the image and a target position must be determined which is stored in the input buffer to the robot controller。 The feature extraction algorithm is dependent on the weld joint to detect and is defined beforehand。 Examples of different weld joint types are shown in Figure 3。4。

Figure 3。5。 Example of the steps of feature extraction of the segmentation process: (1) outlier elimination from the scan, (2) line segmentation generation based on the specific joint template, (3) join the line segments, and (4) validate against templates and tolerances [3]

The feature recognition process includes the following basic tasks: (i) identification and elimination of outliers, (ii) contour generation of the weld profile and generation of line segments based on predefined templates, (iii) merging line segments, and (iv) validation of joint parameters so that they are within predefined tolerances and match the joint template, see Figure 3。5。

From a control point of view, seam tracking is usually performed with full compensation of the position error measured。 Seam tracking is typically only performed using a nominal path。 The nominal path is the assumed trajectory of the weld joint and during tracking, the robot controller receives new target positions from the sensor and the robot controller overrides the nominal path by changing the position of the TCP while keeping a constant orientation。 This has some benefits and drawbacks。 The benefits are that, given a nominal path, it is rather straightforward to verify the ability of the robot to follow the path with some minor changes while keeping the orientation constant。 This means that issues related to joint limits, singularities and possible collisions are minimized。 Drawbacks are that the user must define and program a nominal path。