Laser-based powder bed fusion thermal history of IN718 parts and metallurgical considerations

Researchers have developed an efficient model that unlocks the complexities of metal additive manufacturing process simulations. Published in Advanced Manufacturing, the algorithm is presented, validated and applied to a specifically designed benchmark demonstrating the complex behavior of additive manufacturing processes and how they can affect mechanical properties. This model closes the gap between very small/short laser length and time scales and relatively large part scales enabling the prediction of defects, optimizing process parameters to avoid those defects and to design microstructures achieving the required mechanical properties.

In laser-based powder bed fusion additive manufacturing the laser is significantly smaller than 1 mm in diameter. In combination with fine metal powder particles, it is possible to manufacture complex parts with features that have not been possible to produce using conventional manufacturing processes. The laser scans the metal powder with speeds reaching several meters per second. The laser hence passes any point in the processing chamber within microseconds melting the powder, evaporating a portion of the molten metal and possibly missing some parts of the raw material leading to defects, referred to as pores. The back-and-forth motion of the laser results in fluctuating temperatures leading to high temperature differences throughout the part, which result in additive manufacturing characteristic microstructures, specific mechanical properties and high stresses.

Optimizing the printing process relies heavily on trial and error. Predictive simulations could provide better guidance to achieve the required part quality in a more reliable and quicker way. Given the large difference in laser diameter and parts to be printed (usually several cm³ in size) modelling additive manufacturing processes is computationally expensive. Researchers have limited themselves to either reducing the studied geometry size to a few mm³ or have reverted to simplified models requiring prior experimental calibration. Modelling the complete printing process resolving the laser scan path in detail has thus far not been possible.

This paper presents an efficient algorithm that resolves the laser scan path fully to predict the temperature field throughout the complete part, hence closing the gap between small and large scales. The laser scan path used to control the printing machine is the same file driving the simulation providing reliable information about part quality prior to actually starting the print job. In order to achieve this, the algorithm combines a semi-analytic model describing the laser energy distribution with high fidelity finite volume models accounting for complete part thermal history. The results are obtained within a few hours on a common desktop computer.

To verify and validate the novel modelling approach a special base plate holder was designed to separate the printed parts from the printing machine. Further, simple cones were oriented in such a way to create different thermal histories that would lead to different metallurgical behaviors and eventually different mechanical properties. Temperature measurements were performed using infrared cameras and thermocouples.

The validated thermal model results are used as input to phase diagram calculator and grain growth models to predict the metallurgical state of the printed cones. The simulations show that each point is melted twice. The remelting leads to epitaxial growth and columnar grains mostly parallel to the build direction. The cooling rates differ from one cone to the other as a result to the heat levels retained in cones of different orientations. The thermal histories are so different that the cones are expected to have different concentrations of hardening precipitates, leading to differences in cone hardness. Specimens printed during model verification and validation are analyzed confirming that columnar grains are dominant for all printed cones. The hardness differences between cones were quantified and confirm the findings of the modelling chain.

This paper ”Laser-based powder bed fusion thermal history of IN718 parts and metallurgical considerations” was published in Advanced Manufacturing.

Megahed M, Verhülsdonk M., Vervoort S., Dionne P., Laser-based powder bed fusion thermal history on IN718 parts and metallurgical considerations. Adv. Manuf. 2025(1):0003, https://doi.org/10.55092/am20250003.