More Lasers, More Layers

Powder bed-based melting of metals by means of a laser beam, PBF-LB/M for short, is the most widespread additive manufacturing process for metallic materials. This manufacturing process is used in particular for complex geometries, small series production, or customized products that cannot be made using conventional manufacturing methods. The process makes it possible to integrate functions such as internal cooling channels and lightweight structures directly into components and manufacture them in a single process step.

Although the advantages of this technology compared to conventional processes are well known, there are various challenges with regard to quality assurance and productivity when integrating it into series production. Both quality and productivity can be enhanced by increasing the number of lasers and associated laser optics. Part quality in particular can be improved with the help of innovative scanning strategies such as using pre- and post-heating lasers. However, implementation is associated with numerous challenges: During parallel exposure in adjacent areas, the laser beams with their welding by-products such as welding smoke and spatter can influence each other, depending on the direction in which the shielding gas flow is moving. This causes fluctuations in the energy input, which destabilize the molten pool and reduce process and component quality. If several lasers are used in parallel, it is therefore important to make sure that the lasers do not get in contact with each other or their welding by-products.

A research team at Fraunhofer IPK has developed an innovative algorithm for this purpose, which can be used to control multiple lasers in parallel: Based on existing scan vectors, the algorithm adapts the scan strategy to avoid unwanted shadowing effects. To this end, it takes into account the time-dependent positions of all lasers involved in the process. Information about the geometry and direction of the welding fume trail is of particular importance in order to prevent the laser's energy from being compromised by the welding fume. Therefore, the first step was to characterize the smoke trail using visual image processing. This allows the algorithm to define »forbidden zones«, i.e. areas where no laser is allowed to enter, and to adjust the scan vectors accordingly.

In order to simplify the application of the algorithm and optimize the process, the researchers are also developing a novel welding fume monitoring system for measuring the smoke trail. The aim is to comprehensively record the smoke trail on the basis of a multi-sensor system. The scientists are currently using an optical approach: The smoke trail is monitored based on images. In addition, photodiodes are used to detect the welding smoke. Thanks to the precise knowledge of its position and size, the algorithm can automatically adapt different process parameters. To achieve high efficiency and compatibility with various software applications, the algorithm is programmed in Python. In the further development of the monitoring system, the scan vectors are to be set automatically by the algorithm so that coverage by welding fumes can be completely avoided, thus significantly improving the component quality.