Abstract :
[en] Additive manufacturing (AMing) is an expanding technique that enables producing complex shaped parts upon a substrate, under a layer by layer deposition process, each layer corresponding to the remelting of powders or wire feedstock under a heat source (laser or electron beams) prior to rapid solidification. Both singles and dissimilar alloys can be used, either for the deposit or for the substrate.
AMing techniques are increasingly used for various applications such as repairing parts, graded microstructures, coatings with improved properties, etc.
In the case of Ti-6Al-4V, the parameters used for powder based AMing processes influence the macrostructure (geometry of the melt, size and texture of primary grains, internal defects, etc.) but also the final microstructure within the deposit, which in turns will strongly influence the mechanical properties. For common applications of Ti-6Al-4V such as aeronautics or biomedical, one will often seek both strength and ductility, especially for fatigue purpose, such properties being already achieved with the conventional hot worked alloy, which exhibits bimodal and globular fine grained structure. AMing processes lead to microstructures that are more complex, such as Basketweave Widmanstätten, ’-martensite, massive, either in the form of precipitates or as retained phase, Ti3Al intermetallics, these phases being present separately or together in the deposit, according to the local thermal histories during processing.
The individualized characterization of each phase within the microstructure is a preliminary key step in the purpose of optimizing mechanical properties. Such an approach can help to know if expected properties are already achieved on the as-fabricated AMed part, and else to set the possible post-treatment to perform in order to improve the mechanical properties.
’ martensite is very often observed within Ti-6Al-4V alloy processed by Laser Metal Deposition (LMD). Both the formation mechanism and the morphology of martensite are well known, which correspond respectively to a displacive transformation under rapid cooling above the critical rate of 20°C/sec and orthogonal laths between which a substructure made of thinner laths is found. However, the hardness value of martensite ranges between 330 and 420 HV, thus overlaps with that of Widmanstätten (300 -380 HV), thus leading to misinterpretation.
The hardness depends on several parameters, including the crystal lattice, solid solution with both substitutional and interstitial elements, the dislocations density, the presence of precipitates, internal defects, etc.
In this work, the micro-mechanical behavior of ’ martensite obtained from different manufacturing conditions is studied. One reference sample is compared to two other samples obtained from LMD process. The reference sample corresponds to a fully martensite obtained after a conventional solution treated annealing (STA) above Beta transus followed by water quenching. The AMed martensitic samples are obtained while setting processing parameters so as to allow achieving large columnar grains with epitaxial growth on the one hand (under high Incident energy, IE), and fine weavy grains on the other one hand (under low IE). (Fig. 1)
The micro-mechanical behavior is obtained while carrying out hardness tests at different scales. Each scale level highlights local contributions, such as prior grain size for the macro scale, martensite lath size for the mesoscale, lath joint for the micro scale, dislocations or crystal orientation at the nanoscale.
The complete characterization of the reference serves as a basis for comparison with equivalent results obtained on martensite produced by LMD process.
The indentation size effect is established (Fig. 2).
Difference between reference and LMD samples are discussed thus highlighting the effects of various features (cooling rate, prior grain size, lath size, dislocation density, etc.) on the hardness.
This work is useful not only for better understanding the micro-mechanical behavior of martensite, but it can also serve as a reference for the establishment of quantification models that set the microstructure evolution under AMing processing.
