Resumo:
Tool steels produced by powder metallurgy, such as Vanadis® 8, stand out due to their excellent
mechanical properties, making them suitable for applications in cutting tools, punches, and dies.
However, the chips obtained from machining this material are sold as scrap to companies that use
remelting, an inefficient method due to non-uniform cooling, which promotes the coarse formation of
carbides and compromises the material’s microstructural properties. Therefore, the objective of this
work is to utilize a new optimization algorithm called the multi-objective algorithm of Lichtenberg
to determine the best parameters for the high-energy milling (MAE) and sintering processes, with
the aim of making the reuse of this material economically viable. The response surface methodology,
a statistical technique used to model and optimize processes, was employed to generate metamodels
with the equations. The milling parameters used were 300–400 rpm, a ball-to-weight ratio of
1:15, and a duration of 12–50 hours. To validate the model, microstructural characterization analyses,
such as SEM, XRD, and laser diffraction, were performed. The responses analyzed included particle
size, particle distribution, and mill energy consumption. The metamodel exhibited an overall error of
6.82% when compared to the confirmation experiment. The milling results showed that it is possible
to save up to 26 process hours with the optimization and obtain powders of excellent quality for subsequent
processes, such as compaction and sintering. For the sintering process, the parameters used
were 1200-1300 ◦C and 1-2 hours. The evaluated responses were apparent density, furnace energy
consumption, and microhardness. The equations derived from the metamodels showed satisfactory
fits, with adjusted R² values of 81.17% for apparent density, 91.58% for energy consumption, and
85.94% for microhardness, enabling the identification of optimized sintering temperature and time
variables. The sintering metamodel demonstrated a global error of 7.2% compared to the validation
experiment. Following heat treatment involving quenching and double tempering, the recycled samples
achieved an average hardness of 96.9% compared to the commercial material, demonstrating
good performance despite microstructural differences caused by the recycling process. In the compression
test, the recycled samples attained 66.84% of the commercial material’s elastic modulus.
The novelty of this study lies in the integration of Lichtenberg’s new multiobjective algorithm with
an innovative approach to recycling tool steels manufactured through powder metallurgy, offering
significant added value.