In the first part of the article we’ve seen how AlTiN thin film deposition process was carried out via reactive Physical Vapor Deposition High-Power Impulse Magnetron Sputtering (PVD HiPIMS) to coat Ti6Al4V substrates, realized via Selective Laser Melting (SLM).
Two different SLM process conditions were employed for modifying the obtained part surface morphology and, later, the samples were heat-treated under high vacuum.

Now it’s time to proceed with the analysis of the data acquired during the experimental phase. Do not miss this in-depth investigation and the conclusions at the end, we’ll provide you all the information to get the most out of these three technologies.

Analysis of the Selective Laser Melting results

In the experimental phase we used two different scanning design to realize the cylindrical samples. The upper surfaces of two Ti6Al4V substrates, manufactured using the meander and contour scanning strategies, are shown in the following secondary electron SEM images, respectively.

Whilst their topological features by three dimensional maps are listed in the following figures, together with linear profiles concerning peculiar features of the analyzed surfaces.

1. Meander strategy: it is usually selected for manufacturing massive parts.
2. Contour strategy: it is employed for thin parts, like lattice or trabecular structures, requiring more delicate energy deposition because of their limited dimensions.

For these reasons, the two resulting morphologies are significantly different, because the laser scanning paths and the corresponding values of density of power, namely fluence, are largely modified in the two samples.
For both the samples, the direction of the laser scanning is well visible, which can be directly associated to a preferential direction of melting, similar to a welding process. In addition to this, adjacent laser tracks can be also seen, indicating how the central part of the sample is processed at a certain powder layer.

The impact of the laser fluence

For pulsed wave lasers the fluence F can be calculated as follows:

\[F = {P \times t_{exp} \over d_p \times d_h \times s}\]

where P, t_exp, d_p, d_h and s indicate laser power, exposition time, point distance, hatch distance and layer thickness, respectively.
From the previous formula it can be calculated that the fluence increased from 35.5 J/mm³ up to 60 J/mm³ by changing the scanning strategy from the contour to the meander one. This almost double increase of fluence can promote the formation of larger liquid pools, inducing smoother surface of the liquid material during its solidification. Moreover, it is also visible from the 3D maps that the use of the contour strategy is not able to melt all the powders, due to the lower density of energy irradiated to the single layer. As a result, some residual powders can be found not completely melted but partially joined to the upper surface and the irregular surface possessed high roughness (Sa=13.3 µm).
On the contrary, the meander scanning strategy led to a more regular topology with lower roughness (Sa=4.2 µm) and without prominent asperities and seemed to be more suitable to be coated by a PVD process.

Therefore, the contour strategy resulted inapt to be efficiently coated by a protective layer without any deep modification, rather due to the presence of pronounced valley and hills. Even in case of a polished surface, the residual porosity could remain a critical issue.

For this reason, in this work only meander selective laser melted samples were taken into account to the coating process.

Analysis of the heat-treating results

In the following figure you can evaluate how a 1-hour vacuum thermal treatment at 950°C, into a TAV all metal vacuum furnace, gave rise to a dramatic changing of selective laser melted samples.

After that titanium-aluminum-nitride coatings were deposited onto such kind of treated substrates in order to evaluate their properties.
In order to characterize the coated samples by nanoindentation, the substrate surfaces were polished to almost a mirror like finishing. This process, moreover, led to exclude the influence of the several peculiar surface morphologies.

Analysis of the coating results

During reactive sputtering process, N₂ partial pressure affects the deposition rate in addition to the nitrogen content present in the growing film.
In particular, a high rate leads to a coating with high internal stresses, which can be detrimental for its mechanical stability.
Indeed, a nitrogen partial pressure of 14% gave rise to a film (deposited onto Si) which spalled completely, whilst 20, 25, and 30% permitted to obtain mechanically stable coatings with sputtering rate of 2.3, 2.2 and 1.5 µm h^-1, respectively.
Energy dispersive X-Ray spectroscopy quantitative analysis confirmed the stoichiometry of Al_0.5Ti_0.5N_1.0, highlighting that the three partial pressures did not affect the material composition.

The following figure shows a secondary electron micrograph of the film cross-section where it is possible to observe the typical PVD columnar dense morphology of the obtained films.

The following chart reports the X-Rays diffraction profile of the AlTiN film, which can be attributed to a cubic phase (note that the peak of the substrate present between 65° and 74° was removed).

The coating deposited onto polished commercial Ti6Al4V samples resulted having a quite different behavior from those deposited onto Si.

The adhesion of the coating

The following photo shows how only the films deposited with pN₂ of 25% demonstrated to have a satisfactory adhesion to the substrate of polished commercial Ti64Al4V.

Whilst the films obtained with pN₂ of 20 and 30% partially spalled likely due to high internal stresses relaxation phenomena.

Hardness and elastic module of the coating

The following two graphs show the hardness and the elastic module of the deposited AlTiN thin film, respectively.

Nanoindentation data indicate that the thin films deposited with pN₂ of 25% and 30% exhibited very similar values of hardness H and elastic module E, leading to conclude that N₂ partial pressure did not affect significantly the mechanical properties of the coatings.
Even though the film hardness values resulted slightly lower than the typical ones of this kind of coating, the improvement of substrate surface features were remarkable, if compared to the typical titanium alloy hardness (around 1 GPa).
The two rightmost bars refer to the selective laser melted substrate, prior and after vacuum heat treatment respectively: in this cases values of H and E of the thin film increased considerably. Note how the vacuum heat treatment seemed to decrease those physical characteristics, likely due to a more relaxed substrate microstructure, which reduce the entity of the film internal stresses.

Correlations between mechanical and tribological properties of the film are very important for an assessment of its quality. Although hardness is considered generally as the main property affecting wear resistance and load-carrying capability of materials, properties such as toughness and elastic modulus are also pivotal in describing the mechanical behavior of a coating.

Let’s now examine nanoindentation data in terms of elastic strain to failure, associated to H/E ratio, and resistance to plastic deformation H³/E², associated to the resistance to local contact load.

Elastic strain to failure and plastic deformation resistance of the coating

In the following two graphs, the ratios H/E (elastic strain to failure) and H³/E² (plastic deformation resistance) suggested as the film deposited onto Ti6Al4V selective laser melted substrate possessed better tribological behavior. As described above, the thermal treatment modulated these properties slightly decreasing them, even if the values remained within the errors.

Now we have to find an answer to one last question.
What happens if we try to coat a pristine surface?

Coating a pristine 3D printed surface

A pristine selective laser melted substrate was coated with AlTiN thin films avoiding the polishing process, in order to evaluate the influence of the peculiar morphology on the mechanical compatibility between the two materials.

The following two figures show a backscattered electron and secondary electron SEM (Scanning Electron Microscopy) image respectively, showing the AlTiN coating deposited onto pristine SLMed surface.

Unfortunately, how it is possible to observe, the coating spalled where the morphology was particularly arcuate. In particular, the presence of a very thin film onto the spalled zones (less bright gray areas in the first figure), confirmed that the spalling process occurred during the deposition, and not later.

Conclusions

In this work, AlTiN hard coatings were deposited by reactive PVD HiPIMS with the aim of improving the surface features of Ti6Al4V substrates, realized via Selective Laser Melting.

Preliminary results demonstrated that it is possible to coat the titanium alloy substrates with dense films with proper stoichiometry and crystallographic phase, obtaining a good improvement of the mechanical surface features.

The laser scanning paths can modify significantly the surface morphology, affecting the resulting adhesion of AlTiN coatings. In particular, between the meander and contour scanning strategies, the first ones seemed to be more suitable to be coated with the protective layer, when the surface is adequately treated by means of polishing process.

The tribological properties of the resulting films showed a good behavior of the couple film/substrate, even when the latter is vacuum thermal treated. Unfortunately, the mechanical matching resulted not so good in case of coating onto a pristine surface. Indeed, the surface morphology gave rise to coating spalling where the roughness is particularly marked. In future, further investigations will be mainly addressed to work on the substrate morphology as well as the deposition coating conditions, in order to prevent those kinds of detrimental phenomena.