Sigurd Villumsen

Flexible Laser Metal Cutting: An Introduction to the ROBOCUT Laser Cutting Technique

This paper describes a new flexible and fast approach to laser cutting called ROBOCUT. Combined with CAD/CAM technology, laser cutting of metal provides the flexibility to perform one-of-a-kind cutting and hereby realises mass production of customised products. Today’s laser cutting techniques possess, despite their wide use in industry, limitations regarding speed and geometry. Research trends point towards remote laser cutting techniques which can improve speed and geometrical freedom and hereby the competitiveness of laser cutting compared to fixed-tool-based cutting technology such as punching. This paper presents the concepts and preliminary test results of the ROBOCUT laser cutting technology, a technology which potentially can revolutionise laser cutting.

Process time optimization of robotic remote laser cutting by utilizing customized beam patterns and redundancy space task sequencing

This dissertation is written as a part of the ROBOCUT project which concerns the development of a new laser cutting technology. The technology seeks to increase the performance of traditional and remote laser cutting by using beam shaping technologies. The resulting customized beam patterns are obtained by using a diffractive optical element (DOE). The main work of this dissertation falls within three areas of focus. The first area of focus covers the development of a flexible laser processing laboratory for cutting with customized beam patterns. A laboratory was created with a single mode fiber laser used mainly for DOE cutting and an industrial robot and XY table for positioning the laser beam. Furthermore, software interfaces to all pieces of auxiliary equipment were created. An iterative learning control (ILC) algorithm was implemented on the control system in an effort to increase the accuracy of the positioning system. The implemented algorithm was capable of reducing the tracking error by a factor of almost 2. The second area of focus covers the establishment of a benchmark with which the performance of the ROBOCUT technology could be evaluated. It was chosen to evaluate the performance of the ROBOCUT technology based on the remote fusion cutting technology. It was shown through several experimental series that ROBOCUT cutting performed significantly better when concerning stability than RFC. This increase in stability did however come at the cost of speed, as the ROBOCUT cutting was conducted at approximately 20% of the cutting speed of RFC. This reduced speed is however due to the low duty cycle of the cutting process, which was also in the range of 20%. The third and final area of focus concerns the development of a framework for task sequencing and path planning that can be used with the developed DOE based cutting technology. The developed framework was based on sampling the redundant axes of the laser cutting system and transforming the sequencing problem into a generalized traveling salesman problem (GTSP) which

Optimizing tracking performance of XY repositioning system with ILC

Controlling complex mechanical systems is often a difficult task, requiring a skilled developer with experience in control engineering. In practice however, the theoretical difficulties of designing a good controller is only a first step as the implementation itself on the various pieces of equipment is also often challenging. This paper investigates if iterative learning control (ILC) can be used as an alternative to tuning existing controllers for improving system performance. This is evaluated by a case study on a high speed XY-positioning system used for laser cutting. An ILC algorithm is implemented by using a server client structure from Matlab. After tuning the parameters an implementation is found which is able to increase the tracking accuracy significantly for cutting speeds up to \(0.5\;{\text{m}}/{\text{s}}\). This is done only by implementing code on the master control unit and thus without changing subsystem controllers.