Page 113 - MSAM-4-3

Materials Science in Additive Manufacturing Sustainable manufacturing composite material optimization

Gray Relational Analysis (GRA), has yielded superior suffer from inefficiencies, such as excessive material
multi-objective optimization outcomes. These methods usage, high-energy consumption, and limited control
have proven effective in identifying process parameter sets over process parameters. AI-driven approaches address
that offer an optimal balance among mechanical strength, these challenges through predictive modeling, real-time
energy efficiency, and surface finish – key factors in high- process optimization, and adaptive control mechanisms.
performance impeller fabrication. By leveraging AI-based systems, research institutions and
industrial entities can significantly reduce material wastage,
Recent advancements have also seen the application
of MCDM techniques in the production of composite enhance energy efficiency, and maximize overall process
performance, aligning 3D printing practices with circular
impellers, particularly those fabricated using high- economy principles. A major source of inefficiency in AM
performance thermoplastics such as carbon fiber- is failed prints, which often result from over-supporting
reinforced polyether-ether-ketone (CF-PEEK). These structures, suboptimal parameter settings, warping, or
materials are highly valued for their exceptional strength- poor interlayer adhesion. Studies have demonstrated
to-weight ratios and thermal resistance, making them that defective prints account for approximately 10 –
suitable for demanding applications. Optimization studies 30% of material waste in FDM processes. To mitigate
44
employing a hybrid Fuzzy AHP-TOPSIS framework have these losses, AI-based real-time defect detection systems
identified an optimal combination of process parameters utilizing computer vision and CNNs have been developed.
for CF-PEEK impeller production, including a 70% infill These systems continuously monitor the printing process
density, 0.15 mm layer height, 60 mm/s print speed, and and detect potential defects as they emerge, enabling
an extrusion temperature of 445°C. This parameter set has real-time corrective actions such as adjusting extrusion
been ported to deliver superior mechanical strength while temperature, print speed, or feed rate to prevent print
maintaining energy efficiency. These findings underscore failure. Furthermore, ML-based predictive models trained
the growing relevance of MCDM-based optimization on historical print data have achieved defect detection
models in the development of next-generation impellers accuracies of up to 90%, offering a proactive means of
for aerospace, marine, and advanced industrial fluid minimizing material loss and enhancing process reliability.
systems. 43
Another major sustainability challenge in 3D printing
Despite significant progress, a key challenge that is the overuse of support structures, which not only leads
persists is the limited generalizability of current MCDM- to significant material wastage but also increases post-
based models to diverse impeller geometries and material processing time and costs. To address this, AI-enabled
compositions. Future research must focus on further topology optimization and generative design techniques
advancing AI-enhanced MCDM frameworks through the have been employed to minimize support requirements
integration of real-time sensor data, enabling adaptive without compromising part strength or manufacturability.
control of printing parameters based on in-situ mechanical Combined with lattice structure optimization algorithms,
and thermal feedback. This would allow for dynamic these approaches allow for the design of mechanically
process optimization tailored to varying design and robust and lightweight parts, resulting in substantial raw
material requirements. Furthermore, the incorporation material savings. Empirical studies have demonstrated
of sustainability metrics – such as carbon footprint material reductions of 20 – 40% through AI-driven
assessments and life cycle analysis (LCA) – into MCDM topology optimization, while maintaining mechanical
models will be essential in steering FDM-based impeller performance. Beyond waste minimization, AI also plays
45
manufacturing toward circular economy principles. With a pivotal role in optimizing energy consumption in AM
continuous advancements in AI, the IoT, and decision- processes. Energy usage in FDM systems is highly sensitive
making algorithms, the future of impeller production to parameters such as print speed, nozzle temperature,
is poised to become more sustainable, intelligent, and layer thickness, and infill density. AI-powered energy
autonomous, supporting the creation of high-performance, simulation systems – leveraging deep learning architectures
energy-efficient components for industrial fluid systems. such as long short-term memory (LSTM) networks
and RL – enable real-time parameter prediction and
1.3. AI-based waste reduction and energy adjustment to lower energy usage while preserving print
optimization in 3D printing
quality. Coupled with real-time power monitoring through
The integration of AI and ML into AM has become a smart meters and intelligent sensors, these systems enable
driving force in minimizing material and energy waste, adaptive energy management strategies, achieving energy
thereby advancing the sustainability of 3D printing. savings of 15 – 30% during production. Recent advances
46
Conventional AM techniques, particularly FDM, often in AI-based production scheduling further enhance energy

Volume 4 Issue 3 (2025) 5 doi: 10.36922/MSAM025200033