Numerical framework towards the improved functional response of Ti-Ni SMA lattice structures by localized vaporization of alloying elements during LPBF.

Laser Powder Bed Fusion (LPBF) of Ti-Ni shape memory alloys (SMA) is generally accompanied by vaporization of alloying elements, which is considered a detrimental phenomenon because of the high sensitivity of functional properties to compositional variations. In the present work, this compositional sensitivity is used to improve the functional characteristics of Ti-Ni components by locally-controlled nickel vaporization during processing.

First, a numerical framework was developed to predict the influence of LPBF process parameters on the printing density and functional properties of Ti-Ni components. The cornerstone of this framework is a thermal conduction-based melt pool model. To predict printed density, this model is augmented with experimentally obtained melt pool geometry ̶ printed density relations. To predict functional properties, the same model is combined with the Hertz-Knudsen vaporization law to calculate local compositional changes and therefore, changes in transformation temperatures and stresses. This numerical framework allows to build an LPBF process map describing the range of functional properties obtainable by varying the process parameters while maintaining high printed density.

Next, a finite-element (FE) based optimization framework attributes different material laws to individual nodes and struts of the lattice structure. The optimization criterion is to maximize the volume of matter undergoing martensitic transformation during loading, thus increasing the level of reversible deformation the structure can sustain.

The developed framework was validated with diamond lattice structures produced using a TruPrint 1000 LPBF system by varying the laser power from 50 to 150 W and the scanning speed, from 500 to 1000 mm/s. These processing conditions enabled controlled compositional variations from 51.17 to 50.75 at.% Ni. Compression tests of printed structures were used to compare experimental data with the FE model predictions. Future work will focus on the direct characterization of printed structures to integrate the experimentally measured mechanical properties into the FE framework.