Laser powder bed fusion (LPBF) is a bottom-up, layer-wise additive manufacturing technique commonly used for metals and alloys, offering tailorable physical and mechanical properties according to the end use of the parts. Desired properties are achieved by careful control over numerous variables inherent in the process, such as printing parameters, build geometries, and physical and chemical properties of the feedstock materials. Nowadays, the microstructure of many built parts can be designed and altered locally on the run, which requires reliability and a deep understanding of any evolution occurring under transient conditions. One way to gain insight into the bulk during the LPBF process is by operando or in situ measurements with neutrons that can penetrate metallic materials up to a few cm. In order to facilitate the intended investigations, a downsized LPBF machine (n-SLM) was developed at the Paul Scherrer Institut, Switzerland.

The n-SLM device can accommodate different neutron measurement techniques to characterize, for example, the evolution of residual strains/stresses, dislocation densities, phase transformations, defects, and temperatures during printing. We used neutron diffraction to observe the dislocation density evolution as a result of local laser rescanning to promote recrystallization on 316L. Neutron diffraction is also used to follow the progression of residual stresses throughout the printing of duplex steels. In addition, we exploited the high sensitivity of polarized neutron imaging to spatially capture phase transformations involving ferromagnetic phases in multi-material systems. Evolving defects in the printed specimen, like porosity and cracks, can be identified through conventional neutron imaging, profiting from the difference in local transmission. By focusing on the cold wavelength range beyond the last Bragg edge, we can furthermore produce spatially resolved temperature maps revealing the evolution of temperature gradients over time.

This contribution will showcase the technical aspects and potential of the n-SLM machine, as well as some examples of the findings, for the wider user communities.