Akademik Veri Yönetim Sistemi
2nd MecaNano Workshop on Machine Learning for Micro- and Nano-Mechanics, Budapest, Macaristan, 4 - 05 Eylül 2025, (Yayınlanmadı)
Additive manufacturing (AM) offers significant
advantages in producing metal components with complex geometries and tailored
mechanical properties that are difficult to achieve using conventional methods.
However, the complex interactions of multiple process parameters often lead to
manufacturing defects such as porosity, microstructural irregularities, and
dimensional distortions. Therefore, identifying optimal processing conditions
is crucial to ensure part quality and structural reliability. Considering the
high cost and time consumption of experimental optimization, machine learning
(ML) algorithms have emerged as a promising alternative for modeling the
intricate relationships among process parameters, material behavior, and
performance.
Molybdenum (Mo), despite its high melting point, has excellent
thermal and electrical conductivity, and outstanding corrosion resistance yet
it exhibits inherent brittleness due to crack-prone grain boundaries.
AM-fabricated pure Mo typically suffers from high porosity and intergranular
fracture behavior. To overcome these limitations, rhenium (Re) is commonly
added as an alloying element to enhance mechanical performance. Re contributes
significantly to high-temperature strength, oxidation resistance, and ductility,
making it a strategic element for components subjected to extreme thermal and
mechanical loads, such as those found in aerospace and energy systems.
This study investigates the use of ML algorithms to
predict the mechanical performance of Mo-Re alloys produced via AM. A
structured experimental dataset was compiled, incorporating Re content,
processing parameters, and mechanical properties. Various regression-based ML
models were trained and validated to identify the non-linear correlations among
alloy composition, processing conditions, and material behavior. The results
demonstrate the strong prediction capability of ML models and emphasize their
potential to support process optimization in AM.
In summary, this study highlights the benefits of integrating ML techniques into the additive manufacturing of rhenium-containing alloys, offering a data-driven approach for the efficient design and development of high-performance components.