Laser directed energy deposition (LDED) is a promising technology for preparing complex-shaped melt-grown Al2O3-based ceramics which are important candidates for new high-temperature structural materials. The characteristics of ceramic powder particle significantly affect the stability of LDED process and the forming quality of ceramic samples. In this study, rod-like and thin-walled Al2O3 ceramics were one-step fabricated by LDED by using the plasma spheroidized alumina powder (PSAP) and irregular alumina powder (IAP), respectively. The differences in forming quality and mechanical properties of the specimens prepared by the above two powders were analyzed and discussed. Both powders achieved high-quality forming of rod-like samples with high relative densities of more than 99% and 98%, respectively. The flowability of IAP met the fundamental forming requirements of LDED technology. The microhardness and fracture toughness of the Al2O3 ceramics obtained by using IAP were 17.77 ± 0.97 GPa and 4.58 ± 0.50 MPa·m1/2, respectively. Due to the angular shape and narrow particle size distribution of IAP, there were lack-of-fusion (LOF) pores at the grain boundaries. Intergranular oxide impurities and LOF pores reduced the flexural strength. In contrast, the flexural strength of Al2O3 ceramics prepared by PSAP reached 276.6 ± 22.9 MPa due to the columnar crystals with highly consistent growth orientation. Combining the reduction of line energy density and the supplement of additional laser energy input, crack-free thin-walled Al2O3 ceramics with a width of 30 mm were successfully manufactured using PSAP. Its relative density was close to 99%, and the forming error of width direction was only 5.7%. The study demonstrates the profound influence of powder particle shape on the forming quality of LDED, which provides an essential reference for laser additive manufacturing of high-quality Al2O3-based ceramics.
Download: 0
Show more
Improving the built quality of printed objects is commonly tackled from a material composition or manufacturing processing perspective, but mechanics also play a critical role in controlling performance of the printed parts. In this study, we investigate lattices that are manufactured with recycled PLA (Polylactic Acid) using a mechanics-guided approach, which assigned optimal printing parameters to specific struts according to a possible tensile, compressive, and flexural stress state. We identify the optimal combinations of nozzle temperature, printing speed, and layer heights using standard test specimens for each loading scenario, The strengths of specimens with optimal parameters are up to 23% higher than those with default parameters. Guided by simulations and experiments, we then compared the compressive strength and energy absorption of recycled PLA lattices fabricated by programmable printing parameter sets against those with default parameter sets without considering mechanical features. Strength and energy-absorbing ability can be improved by up to 25.52% and 140.22% respectively. Overall, we validate the role of mechanics in fabricating a 3D-printed object and envision that our approach is universal and applicable to improving the mechanical properties of any given geometries printed by other methods and with other base materials.
Download: 0
Show more
Highlights • We review conventional and modern machine learning (ML)-assisted works employing closed loop control (CLC) strategies in metal additive manufacturing (AM). • We discuss various AM defects and their causes, their observability, and controllability in terms of avoidance, mitigation, or repair. • We show that traditional CLC control solutions lack the flexibility and scalability to adequately support AM processes. • We propose an ML-assisted CLC solution framework supported by ML algorithms which solve quickly and support a broader spectrum of situations. • We focus our discussion on ML-assisted adaptive in-situ control – the topic which has received the least attention in the literature so far. In metal additive manufacturing (AM), the material microstructure and part geometry are formed incrementally. Consequently, the resulting part could be defect- and anomaly-free if sufficient care is taken to deposit each layer under optimal process conditions. Conventional closed-loop control (CLC) engineering solutions which sought to achieve this were deterministic and rule-based, thus resulting in limited success in the stochastic environment experienced in the highly dynamic AM process. On the other hand, emerging machine learning (ML) based strategies are better suited to providing the robustness, scope, flexibility, and scalability required for process control in an uncertain environment. Offline ML models that help optimise AM process parameters before a build begins and online ML models that efficiently processed in-situ sensory data to detect and diagnose flaws in real-time (or near-real-time) have been developed. However, ML models that enable a process to take evasive or corrective actions in relation to flaws via on the fly decision-making are only emerging. These models must possess prognostic capabilities to provide context-sensitive recommendations for in-situ process control based on real-time diagnostics. In this article, we pinpoint the shortcomings in traditional CLC strategies, and provide a framework for defect and anomaly control through ML-assisted CLC in AM. We discuss flaws in terms of their causes, in-situ detectability, and controllability, and examine their management under three scenarios: avoidance, mitigation, and repair. Then, we summarise the research into ML models developed for offline optimisation and in-situ diagnosis before initiating a detailed conversation on the implementation of ML-assisted in-situ process control. We found that researchers favoured reinforcement learning approaches or inverse ML models for making rapid, situation-aware control decisions. We also observed that, to-date, the defects addressed were those that may be quantified relatively easily autonomously, and that mitigation (rather than avoidance or repair) was the aim of ML-assisted in-situ control strategies. Additionally, we highlight the various technologies that must seamlessly combine to advance the field of autonomous in-situ control so that it becomes a reality in industrial settings. Finally, we raise awareness of seldom discussed, yet highly pertinent, topics relevant to adaptive control. Our work closes a significant gap in the current AM literature by broaching wide-ranging discussions on matters relevant to in-situ adaptive control in AM.
Download: 0
Show more
