Formnext 2024: flexible beam-shaping platform optimizes LPBF processes

13 Nov 2024

Fraunhofer ILT says custom beam profiles improve component quality, cut material losses.

A new approach to beam shaping will soon make additive manufacturing more flexible and efficient: the Fraunhofer ILT (Fraunhofer Institute for Laser Technology) has developed a new platform to optimize laser powder bed fusion (LPBF) processes.

The ILT states, “Customized beam profiles improve component quality, reduce material losses and enable scaling of the build-up rate of the single beam process. The research institute is presenting the test system at the Formnext exhibition in Frankfurt, Germany, from November 19 to 22.

The ILT and the Chair of Technology of Optical Systems at RWTH Aachen University are working together to create a test system enabling them to investigate complex laser beam profiles in power classes up to 2 kW, in order to customize solutions for industrial partners.

Currently, laser powers of around 300 to 400 watts are common in many LPBF processes. However, the standard Gaussian laser beam they use has significant disadvantages, says the ILT: the high concentration of power in the beam center leads to local overheating and undesirable material evaporation as well as process instability, both of which can impair component quality due to spatter and pores, meaning that the laser power available in LPBF systems – often up to 1 kW – cannot be utilized for most materials.

“One way to speed up the process is to use several lasers and optical systems in parallel,” said Marvin Kippels, PhD student in the Laser Powder Bed Fusion Department at Fraunhofer ILT. “However, the costs scale at least proportionally to the number of systems installed.”

Boosting productivity

These systems cannot always be utilized homogeneously in real applications, which means that productivity cannot be increased proportional to the increase in power. A promising approach is, therefore, to increase the productivity of the single beam process, which can also be transferred to multi-beam systems.

Previous studies have shown that even simple beam shapes with rectangular, ring-shaped or a combination of two Gaussian distributions produce promising results for both component quality and process speed. The potential of more complex beam shapes has so far been largely unexplored, as the necessary system technology was lacking.

“The interaction of laser beam and material in the process is so complex due to its dynamics that simulations can only provide indications of the actual melt pool behavior,” said Kippels, who is currently setting up a new type of system that uses LCoS-SLMs (Liquid Crystal on Silicon - Spatial Light Modulator), which will enable researchers to investigate almost any beam profile in the LPBF process.

As it has a laser power of up to 2 kW, the innovative system is a platform for testing new beam shapes at very high power levels in the LPBF process, which allows the suitable system technology to be identified for an individual LPBF task. “We can optimize the LPBF process in a targeted manner,” said Kippels.

Flexible profiles

Currently, system technology is often promoted as able to produce specific beam shapes such as ring or top hat profiles. However, the choice of these beam shapes is not based on an in-depth understanding of the underlying process mechanisms, which is reflected in the sometimes contradictory literature on the subject.

Only by fundamental understanding the processes can research specifically define which adjustments achieve a defined target, such as a certain melt track geometry. This means that a beam shape must be developed and optimized for the application, which can then ideally be implemented in the company without needing LCoS-SLM technology.

“We are still at the very beginning, but we can already see the enormous potential that beam shaping can offer for the LPBF process,” said Kippels. “There is no one perfect beam shape; every application has its own requirements. Thanks to our flexible beam shaping, we can find the ideal distribution for each process, the best process parameters for the task in question.”