Colloquia

Samenvatting

During the process known as Laser Metal Deposition (LMD), a stream of powder particles is molten by a laser and deposited on a substrate material forming a bead. However, the geometry and dimensions of this bead changes due to undesirable environmental disturbances and varying process variables. Therefore, it is desirable to implement a sensor system able to measure some of the changes in the geometry in real-time. One of these geometrical changes is the height of the bead, which can be measured by measuring the distance between the laser processing head, and the deposited bead, known as the axial displacement.

In this thesis, several possibilities for sensors where identified that can be integrated onto the currently existing LMD set-up at the Chair of Laser Processing of the University of Twente. It was chosen to use the method known as Coaxial Laser Triangulation (CLT).

 In coaxial laser triangulation a probe beam is “injected” into the optical path of the high-power laser beam, which reflects of the geometry of the bead, and generates an image on the coaxial camera if the axial displacement increases.

 Over the course of the research, the CLT sensor method was implemented on the existing laser processing setup. In support of this, multiple models were created to examine the behaviour of the sensor. This included: a geometrical model, an optical transmittance model, and a ray tracing model. Using these models, an implementation of this sensor was designed. Next the sensor was implemented, and experiments were performed using this implementation.

 From the results gathered, it was concluded that the created models where relatively accurate however, the behaviour of a diaphragm remained problematic. Furthermore, using the current implementation, the stated accuracy requirement of 0.1 mm could not be met, and a maximum accuracy, in ideal conditions, of 0.2-0.3 mm was achieved. While in non-ideal conditions, an accuracy of 0.6-0.7 mm was reached.

 Overall, the final recommendations where to redesign the system to improve the overall rigidity, axial resolution, and accuracy. This would be done by replacing the probe laser, camera, and optical tubing. Based on these aspects, it will be possible to meet the requirements.