Solution annealing at 1200°C

After solution annealing at 1200°C, the laser track boundaries are completely canceled (figures 4.17). Furthermore, a remarkable grain coarsening occurs during such heat treatment. It is worthwhile to note that coarsened grains still maintain an elongated shape oriented along the building direction.

Figure 4.17. Optical micrographs of the horizontal and vertical planes after solution annealing at 1200°C for 1 hour and 2 hours. First published in [194].

The previous profiles of the grain boundaries are still visible because of the presence of cluster of Nb-rich carbides formed along them (figure 4.18). These precipitates form because of the Nb segregation at grain boundaries. Nevertheless, at a so high solution temperature such precipitates are not able to pin the boundaries, contrary to what happens during the heat treatment at 1065°C. The diffusion of Nb

is not fast enough to follow the migration of the grain boundaries, therefore during the heat treatment carbides form and growth in the zones initially occupied by the original grain boundaries rather than along their new location. Consequently, the grain boundaries observed after solution annealing at 1200°C appear almost free from precipitates (figures 4.18 and 4.19).

Figure 4.18. SEM micrograph of the vertical plane after solution annealing at 1200°C for 1 hour showing the carbides clusters along the previous positions of the grain boundaries and a new grain

boundary unconstrained by the carbides after recrystallization (compare with figure 14). EDS point analysis on a Nb-rich carbides cluster (point 1) in comparison to the intragranular matrix

zone (point 2).

The collected SE FESEM micrographs demonstrate that the removal of the interdendritic eutectic products is not complete even after achieving at 1200°C such a high dissolution degree. The complete dissolution of the second phases at the dendrite length scale cannot be obtained both because of the high thermal stability of the carbides formed by eutectic reaction during solidification in the SLM process and because of the formation of new very small precipitates resulting from the release of Nb after dissolution of the Laves phases.

Figure 4.19. SE FESEM micrographs of the horizontal plane after solution annealing at 1200°C for 1 hour and 2 hours. Note the grain boundaries free from precipitates and the presence of residuals

of the interdendritic precipitates inside the grains (higher magnification images).

The increase in volume fraction of the finest precipitates (less than 130 nm), provided by the increase in solution soaking time from 1 hour to 2 hours of heat treatment, is evident by comparing the two size distributions reported in figure 4.20.

Conversely, the volume fraction of the coarser particles, in particular the ones larger than 370 nm, is reduced by increasing the heat treatment duration.

Figure 4.20. Size distributions relative to the samples solutioned at 1200°C. First published in [194].

The size distribution obtained for each solution condition are reported for sake of comparison in figure 4.21 where the effect of the solution temperature is considered at fixed time. As it was expected, the higher the solution temperature, the lower is the total volume fraction of the detected second phases. The major difference between the size distributions regards the peak related to precipitates finer than 100-120 nm, which in most cases is reduced as the solution temperature

increases. After 1 hour of treatment, the samples solution annealed at 1065°C contain a higher amount of precipitates finer than 80 nm with respect to the one solution annealed at 980°C, but the volume fraction of the larger precipitates is strongly reduced. The formation of new small precipitates during the heat treatment is faster at 1065°C because part of the Nb released at 980°C is used to form inter-dendritic  phases, as earlier discussed. However, after 2 hours of heat treatment, the volume fraction of precipitates on sample treated at 1065 °C is lower for almost all dimensional ranges with respect to the one recorded on samples solutioned at 980°C. On the other hand, the measured volume fractions at 1200 °C are lower with respect to the ones recorded at 1065°C for almost all dimensional ranges both after 1 hour and 2 hours of heat treatment.

Figure 4.21. Size distributions: comparison between different solution temperature after 1 hour (A) and 2 hours (B) of heat treatment. First published in [194].

The total volume fraction of precipitated second phases is tightly connected with the hardness of the material. The relationship between the mean Vickers microhardness of the as built sample and the solution annealed samples and their relative volume fraction of detected precipitates is graphically shown in figure 4.22.

A clear linear correlation stands for all the samples apart from the ones solution annealed at 1200°C, for which a Vickers microhardness lower than the value

predicted by the linear regression line was recorded. The lack of hardness in these samples, further to their lower amount of second phases, is due to the irreversible grain growth occurred during this heat treatment. Therefore, the loss of mechanical properties cannot be recovered anymore with a subsequent aging treatment. For this reason, despite the higher solution efficiency that can be obtained at 1200°C, such solution temperature is no more considered in the following study of the complete heat treatment cycle.

Figure 4.22. Correlation between the Vickers microhardness and the average volume fraction of precipitates in the as built sample (the volume fraction of the as built sample is obtained considering the eutectic products and neglecting eventual very small intradendritic precipitates as

those in figure 3.27) and the solutioned samples. The vertical error bars refer to 95% confidence intervals on HV0.01, the horizontal error bars refer to standard deviations of the volume fractions.

The regression line with relative 95% confidence band is obtained excluding the points of the samples solutioned at 1200 °C. First published in [194].

4.3 Optimization study of the complete heat treatment