Time-resolved synchrotron techniques are powerful tools used to study melt pool dynamics, defect evolution, as well as phase transformations and stress evolution during the additive manufacturing (AM) processes [1,2]. These techniques provide experimental information to validate numerical models. From synchrotron diffraction experiments, heating and cooling rates involved during the AM process are determined from the changes in the lattice parameter because of the thermal expansion and contraction during temperature changes. It is often assumed that during initial cooling from the liquid state, or in the absence of solid phase transformations, the decrease in lattice parameter is solely due to thermal contraction. The temperature is then estimated from the lattice parameter using the material’s thermal expansion coefficient. However, internal stresses and changes in chemical composition change the lattice parameter, thus affecting the temperature estimation.

The study aims to test and quantify the errors associated with the aforementioned assumption on the heating and cooling rates extracted from time-resolved synchrotron diffraction experiments during AM process combining both experimental and numerical approaches. To that end, time-resolved synchrotron diffraction experiments, as well as residual strain measurements at room temperature, were performed at ID31 (ESRF, France) with a miniaturized laser melting deposition machine that we have developed. A 316L austenitic stainless steel has been investigated due to its absence of solid phase transformation during AM and strong partitioning of elements. A numerical approach to the AM process based on thermal (semi-analytical calculations) and mechanical (finite element method) analyses [3,4] has been used to understand the lattice strain evolution measured by synchrotron experiments. This approach allowed us to deconvolute the contribution of the thermal and elastic strains from the total strain measured at the synchrotron and thus more accurately estimate heating and cooling rates. The study shows that up to one order of magnitude difference can be observed between the “real” and estimated heating and cooling rates under the aforementioned assumption.

[1] S. Hocine, et al., Materials Today, 2020, 34, 30-40

[2] A.C.F Silveira, et al., Additive Manufacturing, 2023, 63, 103408

[3] D. Weisz-Patrault, et al., Additive Manufacturing, 2020, 31, 100990

[4] D. Weisz-Patrault, et al., Additive Manufacturing, 2022, 56, 102903