Production of Single Tracks of Ti-6Al-4V by Directed Energy Deposition to Determine the Layer Thickness for Multilayer Deposition

For the experimental studies, irregular Ti-6Al-4V powder with an average size of 50 – 150 µm and apparent density of 1.85 g/cm³ was employed as depositing material (Figure 1). The chemical analysis of the powder confirmed that the oxygen and nitrogen contents of the powder did not change before and after the deposition process, while in both cases the oxygen content was higher than the standard oxygen content of Ti-6Al-4V powder for additive manufacturing (<0.13%). However, the oxygen and nitrogen content of the bulk components increased after the deposition.

Figure 1: Starting Ti-6Al-4V powder is used as depositing material. This is an irregular powder with an average size of 100 – 150 µm and apparent density of 1.85 g/cm³.

Table 1: The chemical composition of Ti-6Al-4V powder before and after deposition (weight percentage). It shows that the oxygen and nitrogen contents of the powder do not change before and after the deposition process, while in both cases the oxygen content is higher than the standard oxygen content of Ti-6Al-4V powder for additive manufacturing.

Figure 2 shows the single tracks of the Ti-6Al-4V alloy after deposition at various laser power and laser scan speed. As can be seen by increasing the laser power and decreasing the laser scan speed, the size of single tracks increased.

Figure 2: Single tracks of the Ti-6Al-4V alloy after deposition. These single tracks were deposited at different laser power and laser scan speed and analyzed from the top. By increasing the laser power and decreasing the laser scan speed, their sizes increased.

Figure 3 shows the cross section of single tracks after the deposition, and by increasing the laser power, the height of single tracks increased considerably. Moreover, by decreasing the laser scan speed at a constant laser power, the height of deposition increased while, at low laser power and very high laser scan speed, the height of deposition was negligible. Despite the height of the melt pool, the porosity formation inside the melt pool, in particular near the interface of the melt pool/fusion zone area, was another phenomenon which was revealed after the deposition.

Figure 3: Cross section of single tracks after the deposition. By increasing the laser power and decreasing the laser scan speed, the height of melt pool decreased. Please click here to view a larger version of this figure.

The relationship between the single track height and different process parameter is shown in Figure 4. The height of single tracks at different laser scan speeds increased by increasing the laser power, which suggests that the laser power up to a certain point has a positive effect on the deposition height (Figure 4a). However, after that critical point, the laser power negatively affects the growth of deposition due to the delivery of too much energy into the melt pool. Furthermore, it was found that as the laser scanning speed increased, the energy input in the melting pool was reduced and the powder delivery rate was indirectly decreased, and consequently the deposited height decreased remarkably (Figure 4b).

Figure 4: Effect of different process parameters on single track dimension. It is clear that as the laser scanning speed increased (b), the energy input in the melting pool is reduced and the powder delivery rate (a) is indirectly decreased and, consequently, the deposited height decreased remarkably. Please click here to view a larger version of this figure.

These results clearly demonstrate the influence of different process parameters on the geometry of deposited tracks. Despite providing valuable insight into the process, the assessment of the deposited height is still challenging, due to the variety of parameters which were involved. Thus, some efforts have been undertaken to develop a new strategy to evaluate the effect of the combination of process parameters on the geometry of deposited track.

As was shown, the height of deposited layer increased by increasing the laser power, and it was understood that this was not the only parameter that affects the height of the melting pool. In fact, in the period of time needed to melt a given volume of the substrate and deposit an appropriate layer of molten material, a certain amount of energy and powder should be provided to the substrate. This energy is not only determined by laser power and laser scan speed, but also the laser spot size should be considered. For this purpose, the specific energy density per unit spot size (E) and powder feed density (F) is calculated to evaluate the effect of the combination of these parameters.

E, which is the specific energy density, shows the energy that is delivered into the melt pool by the laser, and in principle is responsible for melting the substrate and powder. This energy density is expressed as⁸

                        (1)

Where E is the specific energy density per unit spot size, P is the laser power (W), v is the laser scan speed (mm/s), and D is laser spot size (mm). To obtain an appropriate deposition level for each metallic material, there is a certain level of energy below which no fusion bonds can be achieved, and beyond that the dilution becomes too large. Another factor that shows the effect of the combination of parameters is the powder density (F), which can be calculated as follows⁸

                        (2)

Here, F is the powder feed density, and G is the powder feeding rate (g/s).

Figure 5 demonstrates the variation of deposited layer height as a function of specific energy density. As can be seen, the height of single tracks increased by increasing the specific energy density, which can be related to the higher heat input at the higher laser energy density. The empirical correlation between the energy density and the height of deposition are as follows:

h = 14.99 E – 17.85                        (3)

From this equation, the height of deposited track can be estimated through the calculation of the specific energy density and this equation. On the other hand, the variation of deposited height as a function of powder density, which is shown in Figure 6, showed that by increasing the powder density, the height of deposited track increased, and the empirical relationship between these can be expressed as follows:

h = 38477 F – 157.06                        (4)

This equation shows that the height of deposited track can be calculated by calculating the powder density and this equation. Eq. 3 and Eq. 4 show that by using the combination of process parameters and calculating the specific energy density and powder density, it is possible to forecast the deposited height, and consequently find the best domain to achieve the best deposition quality.

Figure 5: Single track height (h) versus specific energy density (E). The height of single tracks increased by increasing the specific energy density, which can be related to the higher heat input at higher laser energy density. Please click here to view a larger version of this figure.

Figure 6: Single track height (h) as a function of powder feed density (F). By increasing the powder feed density, the height of deposited track increased. Please click here to view a larger version of this figure.

In direct energy deposition of metallic materials, h (the layer height, or ΔZ) is a very important factor that affects the quality of component after the deposition. In the conventional direct energy deposition of metallic components, the height of the deposition layer is considered a constant and, apart from the geometry of component and its material, the process parameters such laser power and laser scanning speed were optimized to fabricate the final component. Indeed, slicing the layers in a constant thickness does not usually conform to the process parameters. Therefore, this thickness may be altered either manually or empirically, which sacrifices the quality of component and the fabrication rate. In general, in conventional layer slicing, over- or under-deposition may be achieved due to the approximations in the consideration of the layer thickness, which needs further corrections such as subsequent deposition or machining the extra layers (Figure 7a). Thus, in this work, effort has been undertaken to develop a new strategy to determine the layer thickness, according to the process conditions which are used in the production of components.

Figure 7: Slicing. (a) conventional slicing strategy, (b) new slicing strategy according to the optimum process parameters; in conventional layer slicing, over- or under-deposition may be achieved due to the approximations in the consideration of the layer thickness, which needs further corrections, such as subsequent deposition or machining the extra layers. In this approach, the layer thickness for the fabrication of component is determined according to a single layer height related to the specific energy density of two combined parameters. E is the specific energy density per unit spot size, F is the powder feed density, t_dep is the single layer thickness, and t_layer is the slice thickness.

In fact, in this approach, the layer thickness for the fabrication of component is determined according to a single layer height related to the specific energy density of two combined parameters. To proof this method and check the correlation between the quality of the component and different layer thickness, some simple cubes were built at various ΔZ and then their cross sections were evaluated.

Figure 8a-b show the representative cross-sections of multilayer-blocks, which were produced according to the conventional method. As can be seen in Table 2, according to the slicing strategy which considers 0.325 mm as the layer thickness, the desired height of the block shown in Figure 8a should be approximately 5.2 mm. However, in the conventional method, the final height of 10.11 mm (over-deposition) was achieved, which is a consequence of considering a high ΔZ (0.6 mm) during the process. This over-deposition process resulted in the lack of fusion between the layers, and a high level of porosity inside the specimen. On the other hand, Figure 8b illustrates that by considering a low ΔZ, the desired height cannot be achieved, and this results in a long deposition process and undesirable microstructure. These discrepancies imply that in the conventional method, slicing the layers in a fixed thickness does not usually conform to the process parameters, and thus the desired layer thickness cannot be achieved. A cross-section of the block, which was produced according to the slicing strategy, is shown in Figure 9. As can be seen through considering an appropriate ΔZ, it could be possible to achieve excellent dimensional accuracy. However, the dimensional accuracy can be decreased at a high level of laser power as a consequence of high input energy, which results in the melting of underlying layer. Table 2 shows that by using the slicing method, a more stable melting pool position can be achieved, and consequently the dimensional accuracy increases. Figure 9 shows a block which is produced according to the slicing approach, and as can be seen by using an appropriate ΔZ (~ 0.5 mm) the desired height of deposition was obtained.

Figure 8: Examples of specimens produced by the conventional method. According to the slicing strategy which considers 0.325 mm as the layer thickness, the desired height of the block which is shown in panel a should be approximately 5.2 mm. However, in the conventional method, the final height of 10.11 mm (over-deposition) was achieved, which is a consequence of considering a high ΔZ (0.6 mm) during the process. On the other hand, panel b illustrates that by considering a low ΔZ, the desired height cannot be achieved, and results in a long deposition process and undesirable microstructure. Please click here to view a larger version of this figure.

Figure 9: Example of a sample fabricated by the slicing approach. It confirms that a proper ΔZ consideration results in an excellent dimensional accuracy.

Table 2: Comparison between the deposited height and desired height in conventional and slicing methods. It shows that by using the slicing method, a more stable melting pool position can be achieved and, consequently, the dimensional accuracy increases.