Ferromagnetic shape memory alloys (MSMAs) have the ability to revert back to original shapes and properties after significant deformation. When in single crystalline form, these alloys produce up to 10% reversible magnetic field induced strain (MFIS) which can be beneficial in actuator and sensor devices in industrial and automotive applications. Advanced manufacturing techniques such as direct metal deposition additive manufacturing enables the fabrication of highly complex components. However, the obtained polycrystalline structure is expected to significantly decrease functional properties (i.e. MFIS) in ferromagnetic shape memory alloys. To address this issues, there is a critical need to establish a fundamental understanding of how non-equilibrium processing (rapid solidification) and complex thermal cycling (reheating) affect microstructural evolution and functional properties in these alloys. The overall goal of this project is to identify fundamental processing-structure-property relations over several length scales that enable laser based advanced manufacturing techniques for functional Ni-based MSMA alloys. The study focuses on the development of a cooling rate-microstructure-magnetic property map for Ni2MnGa alloy resembling conditions of laser metal deposition (LMD). The presented material is based upon work supported by the National Science Foundation under Award No. 1808145.
Multi-layer metal deposition processing creates complex microstructures due to multiple remelts and reheat cycles. In order to create fundamental processing-microstructure-property relations for Ni2MnGa alloy, high cooling rates seen in laser metal deposition (102-104 K/s) were simulated using three methods for rapid solidification in a lab setup: a) levitation-drop melting, b) tungsten electrode arc melting, and c) laser beam melting. A range of cooling rates was achieved by varying sample weight (0.02 to 5 gram). Thermocouple and infrared (IR) camera measurements were performed for levitation drop melting and electrode arc melting, but were found to be experimentally challenging. A finite element (FE) model was developed in Simufact Welding to predict cooling rates in electrode arc melting as a function of sample size (weight). The model was validated with cooling rates obtained from measurements of secondary dendrite arm spacing (SDAS). Finite element model setup and validation was done on Nickel base alloy Ni-625, since correlation between dendrite arm spacing and cooling rate are available in the literature. The model was then modified with thermal properties for Ni2MnGa alloy to estimate cooling rates for the rapid solidification experiments. Microstructural characterization of rapidly solidified Ni2MnGa alloy was performed as a function of cooling rate using light optical microscopy (LOM), scanning electron microscopy (SEM), electron dispersive spectroscopy (SEM-EDS), and X-ray diffraction. Ongoing work involves measurement of martensite and austenite start and finish temperatures and Curie temperature using differential scanning calorimetry (DSC), and measurement of magnetic properties using a vibrating-sample magnetometer (VSM).
Laser beam melting could not completely melt samples, especially at higher sample weight due to laser power limitations of the equipment used. Levitation drop melting into a copper mold was limited to sample weights higher than 2 g and produced samples with steep microstructural gradient, i.e. exhibiting large differences in grain and dendrite sizes throughout the sample. Tungsten electrode arc melting proved to be the most reliable and repeatable method to rapidly solidify metal samples with varying cooling rates up to 10,000 K/s (for 0.2 g samples weight), i.e. achieving high cooling rates that resemble what is seen in laser metal deposition (102-104 K/s). LOM of electrode arc melted samples revealed evidence of a fine cellular/dendritic structure. However, etching of the material was challenging, and extensive etching trials found Marbles reagent to be most successful. Grain size analysis as a function of cooling rate was performed using Image analysis. SEM back-scatter imaging showed evidence of martensitic lamellae and twinning throughout the microstructure. SEM-EDS analysis revealed segregation of Mn and corresponding Ga deletion at cell boundaries. Results on microstructural characterization were compared to microstructures at different locations in laser metal deposition builds, thus reflecting different thermal histories and repeated local melting and solidification. The comparison is currently extended to X-ray diffraction results on electrode arc melted samples.
The FE model was initially setup and validated with secondary dendrite arm space measurements (SDAS) on Ni-625. Efforts to obtain experimentally measured cooling rates using thermocouples and IR camera were only partially successful in providing reliable and repeatable data. The FE model was modified with thermal properties for Ni2MnGa to estimate cooling rates of electrode arc melted samples. Cooling rates as a function of sample weight were then combined with the results from microstructure characterization to create a cooling rate-microstructure map for Ni2MnGa alloy.
This talk presents efforts to develop processing-microstructure-property relations for Ni2MnGa alloy, which are needed to support process parameter development for laser metal deposition of builds with functional properties. Different rapid solidification methods were assessed. Tungsten electrode arc melting of different sample weights achieved high cooling rates that resemble what is seen in laser metal deposition processing (102-104 K/s). Experimental temperature measurements proved challenging, so estimation of achievable cooling rates was supported by a finite element model that was developed in Simufact Welding and validated with dendrite arm spacing measurements on Ni-625. Microstructural characterization of rapidly solidified Ni2MnGa alloy resulted in key information such as solute segregation to cell boundaries and evidence of martensite twinning. Features similar to what is seen in laser metal deposition build. Based on the obtained results a cooling rate-microstructure map for Ni2MnGa alloy was developed. Ongoing work focuses on measuring functional properties of rapidly solidified samples, as well as simulate the effect of cyclic reheating on microstructure and properties.