Additive manufacturing

Percutaneous tracheotomy represents a critical airway procedure, yet training opportunities in pediatric patients remain limited due to the low incidence and high risk associated with the procedure. Simulation-based training plays a key role in enabling physicians to develop procedural skills in a safe environment. This study presents the design, fabrication, and validation of a high-fidelity physical simulator for pediatric percutaneous tracheotomy using the Frova screw technique. The simulator reproduces the anatomy of an eight-year-old patient in a hyperextended neck position, allowing rigid bronchoscope insertion, and includes relevant anatomical landmarks. The presence of a blood pocket enhances procedural realism. The modular design, with a reusable base and disposable cartridge, enables rapid reset between sessions while minimizing maintenance costs. Soft tissues were reproduced through silicone molding, while rigid components were fabricated using fused deposition modeling 3D printing. Validation was performed by 39 physicians using a structured five-point Likert scale questionnaire. The simulator achieved a mean score of 4.2/5, with item scores ranging from 3.6 to 4.7, indicating a high level of perceived realism, procedural fidelity, and educational value, as well as highlighting potential areas of improvement. These findings suggest that the proposed simulator is a useful tool for simulation-based education, with the potential to improve operator confidence and patient safety in complex airway emergencies.

Keywords: 3D printing , CAD modeling , Frova screw technique , pediatric simulator , percutaneous tracheotomy

Additive manufacturing (AM) represents a quite interesting technology for manufacturing components of nuclear reactors. This work investigated the microstructural stability of 316 L steel fabricated via Laser Powder Bed Fusion (L-PBF) from room temperature to 650 °C. Despite the reduced susceptibility of the material to sensitization owing to its low carbon content, temperature variations may induce deleterious effects in nuclear safety-critical components. In as-printed condition, the microstructure is not stable and undergoes significant changes induced by thermal cycling up to 650 °C in Mechanical Spectroscopy (MS) tests: the typical melt-pool pattern disappears, a population of equiaxed grains substitutes the original ones elongated in the build direction, the average size of the cells forming a finer sub-structure inside the grains increases, texture changes, and the excess of vacancies induced by the rapid cooling is recovered. Although the current literature reports that the microstructure is stable up to 500 °C, MS results indicate that the aforesaid irreversible phenomena start at a lower temperature (~230 °C). The present results suggest that the microstructure of the printed material must be stabilized through suitable heat treatments before its application in structural components for nuclear reactors.

Keywords: 316 L , austenitic stainless steels , laser powder bed fusion , microstructure

This study investigates the origin of distortion during sintering of 316 L stainless steel components produced by binder jetting, focusing on friction between the sample and the support surface and on density inhomogeneity in the green state. A design of experiments (DoE) approach evaluates the influence of key printing parameters on the sintering behavior of two geometries with different through-hole sizes. Dimensional measurements, and density profiling, are performed in both green and sintered states. Sintering simulations use the Skorokhod-Olevsky viscous sintering (SOVS) model and include experimentally measured density gradients and frictional effects. Results show that green density varies significantly (52% to 58%) depending on printing parameters, especially binder saturation, and exhibits directional dependence. These variations lead to measurable distortions during sintering. Simulations that include both friction and density gradients match experimental deformations with deviations below 4%. A compensation strategy that places parts on co-sintered 316 L support plates with interposed refractory particles reduces distortion to <1.5%. This work demonstrates the combined role of friction and density gradients in sintering distortion and presents a practical method to improve dimensional accuracy in binder jetting.

Keywords: Additive manufacturing , Dimensional accuracy and precision , Distortion , friction , Metal binder jetting , Shrinkage on sintering

Purpose – Additive manufacturing enables complex geometries, but support structures still compromise surface quality, increase postprocessing effort and cause material waste. This study aims to evaluate the feasibility and process optimization of a dual-material support strategy using polylactic acid (PLA) and polyethylene terephthalate glycol (PETG) on a single-head fused filament fabrication (FFF) system, with the objective of improving surface quality and support removability while preserving mechanical performance. Design/methodology/approach – An experimental campaign was conducted using PETG parts fabricated with PLA as an interfacial support material. Surface quality was assessed through roughness measurements, while mechanical performance was evaluated via tensile tests and digital image correlation. Manufacturing time, material consumption and ease of support removal were also quantified and compared with conventional single-material PETG printing. Findings – The introduction of a PLA interfacial layer significantly improved surface quality and support removal. Surface roughness anisotropy was reduced by up to 99% in vertically printed specimens and 86% in horizontally printed ones compared to single-material PETG. Supports were easily removed without surface damage. Ultimate tensile strength decreased by approximately 20% in vertical specimens and 49% in horizontal specimens, whereas the secant modulus showed limited variation (<10%). Manufacturing time and material usage increased by 57% and 35%, respectively, due to material purging during switching. Finally, the force required to remove the supports completely dropped by 63% on average from the single-material to the dissimilar-material configuration. Practical implications – The proposed approach is suitable for prototypes and functional parts where surface quality and dimensional consistency are critical, especially when only single-nozzle machines are available. Originality/value – This study demonstrates a practical and effective multimaterial support strategy using dissimilar polymers on a single-head FFF machine, offering an alternative to soluble supports and multiextruder systems.

Keywords: Dissimilar materials , Fused filament fabrication (FFF) , Multimaterial , Support optimization , Surface quality

The increasing pressure to decarbonize manufacturing systems is pushing industry beyond conventional lightweighting strategies toward material and process paradigms, capable of delivering functional performance with radically lower environmental impact. In this context, polymer-based composite Additive Manufacturing (AM) offers an underexplored yet highly promising pathway for sustainable production of load-bearing components. This study presents a preliminary comparative cradle-to-gate Life Cycle Assessment (LCA) of a Formula SAE brake pedal, assessing the environmental transition from conventional sheet metal fabrication and finishing operations of Aluminum 7075-T6 to additive manufacturing solutions, with specific focus on Carbon-Fiber-Reinforced Polymer (CFRP) composites. Two topology-optimized designs, respectively for Powder Bed Fusion (PBF) in AlSi10Mg and Material Extrusion (MEX) in Polyethylene Terephthalate Glycol with Carbon Fiber (PETG-CF) are compared to conventional fabrication aluminum benchmark. The analysis is integrated in the product and process design following ISO 14040/14044 standards and is implemented using the Environmental Footprint 3.0 methodology within the 3DEXPERIENCE platform. Results outline that Material Extrusion (MEX) composite manufacturing achieves the lowest environmental impact across all evaluated categories. Compared to conventional manufacturing, the PETG-CF solution enables an approximate 50% reduction in Global Warming Potential and an almost complete elimination of mineral depletion. Unlike metal additive manufacturing, which remains constrained by high process energy demand, MEX benefits from low processing temperatures, minimal auxiliary systems, and highly efficient material deposition. Crucially, these sustainability gains are achieved while maintaining functional performance through design-driven topology optimization. AM composite solutions, by merging advanced material science with additive flexibility, may lead to design approaches which cease to be \'potential\' enablers of sustainable manufacturing for the Industry 5.0 transition.

Keywords: Additive Manufacturing (AM) , Automotive Lightweight Design , Carbon Footprint Reduction , Design for Sustainability , Fused Deposition Modeling (FDM) , Life Cycle Assessment (LCA) , Sustainable Manufacturing , Topology Optimization

Fused filament fabrication (FFF) is one of the additive manufacturing methods used to transform digital models cost-effectively into prototypes, mockups, and functional parts for industrial customized applications, mainly aerospace, automotive, and biomedicine. In an industrial standard design-to-manufacturing workflow, the slicing software is responsible for translating the digital model of the object into a set of instructions for the FFF machine. However, setting printing profiles for FFF machines is a painstaking process in the operative environment due to the long time needed to carry out the required tests and tuning phases. Moreover, the scientific literature needs to include the influence of digital model topologies on the more influencing manufacturing parameters. Thus, this paper proposes a reproducible methodology to understand how the choice of the manufacturing parameters affects the time estimation and mass of the production process. Through a half-factorial Design of Experiment approach, the manufacturing parameters that most significantly affect the time required are identified; furthermore, the methodology aims to suggest adjustments to enhance the accuracy of build time predictions in commercial slicing software. Several case studies in the paper provide empirical support for the findings, highlighting that proper configuration of commercial slicing software can substantially enhance manufacturing process accuracy. In particular, the results show that the best configuration cannot be chosen a priori since the topology of the component affects the optimal choice of parameters. Moreover, a rigorous statistical approach allows for producing functional components with excellent printing times and optimal material consumption, compared to a more random approach that may lead to non-functional components. The methodology suits the industrial environment where processes must be set up quickly with satisfying results.

Keywords: Cost , Design of experiment (DoE) , Fused filament fabrication (FFF) , Manufacturing time , Printing parameters , Slicing

Nowadays environmental sustainability is a hot topic, especially with regards to the transportation sector. In fact, political strategies are oriented to the transition towards the cleaner technologies to reduce polluting and climate-altering emissions. However, even fully electric vehicles are not characterized by zero global emissions, due to the production of electricity from sources that are not always renewable. Moreover, this type of vehicle is afflicted by the limited range provided by the battery pack and its long recharging time. So, it is useful to reduce the energy consumption through appropriate strategies, for example by means of the vehicle lightweighting. In this paper we focus on the battery pack lightweighting by considering different passive battery cooling systems as a replacement for the standard active one. The passive systems considered are air and PCM-based (Phase Change Material) cooling systems. In addition to the primary lightweighting given by the replacement of the cooling system, the secondary lightweighting obtained by reducing the capacity of the battery pack to return to the range of the starting reference vehicle has been also considered. Three tools were used for the study: VI-CarRealTime and another consolidate vehicle model to obtain the power demand on a standard driving cycle; and an ad-hoc battery system model, configurable according to the cooling system. The simulations showed that the air-cooled system leads to greater lightweighting, but it makes the battery cells work far from 20 °C (optimal operating temperature) and therefore it could lead to greater battery cell degradation and its field of application must be limited to vehicles operating in fleets, with predictable mission; this can be overcome by using an appropriate PCM-based cooling system, stearyl alcohol. Furthermore, using a PCM, glycerol, with a melting point close to the optimal operating temperature of the batteries, allows to reduce the cell degradation.

Keywords: Electric vehicle , Energy consumption , PCM (phase change material) , Secondary lightweighting , Traction battery thermal management system , Vehicle lightweighting

Design optimization through the integration of multiple functions into a single part is a highly effective strategy to reduce costs, simplify assembly, improve performance, and reduce weight. Additive manufacturing facilitates the production of complex structures by allowing parts consolidation, resulting in optimized designs, where multiple functions are integrated into a single component. This study presents a design for additive manufacturing method for integrating multiple lattice structures to achieve thermal management and shock absorption functions. The method follows modeling and simulation phases for dimensioning and optimizing solutions to deliver the design functions at different macro- and mesoscale levels. Hierarchical complexity was leveraged to design the two-levels structure in a single part, each delivering a specific function. Specifically, the external layer addresses energy absorption and thermal insulation, while the internal layer acts as a thermal battery by incorporating a phase change material. The design of a container carried by an unmanned aerial vehicle for the transport of healthcare and biological materials is presented. The container is shock-resistant and can maintain the content at 4 ± 2 °C for at least 1 h. As it operates passively without the need for additional energy-consuming devices, it is easy to operate and contributes to increased flight autonomy. A flight mission experiment for urgent transport of blood bags confirmed the capability of the container to preserve blood integrity. This case study demonstrates that the two-layer lattice structure design represents a highly efficient approach to multifunctional design optimization.

Keywords: 3D printing , cellular structure , functional integration , fused deposition modeling , gyroid , impact absorption , multi-scale design , thermal insulation , triply periodic minimal surface , unmanned aerial system

Clinical simulation is fundamental for the healthcare staff to learn and enhance their procedural skills without causing harm to the patients. Despite its importance, in literature appears a deficiency of pediatric pathological mannequins, especially those simulating difficult airway management due to the obstruction of the passage of tubes, fiberscopes, or catheters. Given the importance of simulating complex scenarios in the medical staff's training, the authors decided to realize a modular high-fidelity pathological mannequin with nasal access using reverse engineering and additive manufacturing techniques within T3Ddy, a joint laboratory between Meyer Children's Hospital of Florence and the Department of Industrial Engineering of the University of Florence. The mannequin is developed from diagnostic images of a significant 30-month-old polymalformative patient also affected by Pierre-Robin syndrome modifying the tracheobronchial tree to reproduce an abnormal status. Rigid parts and silicone cast molds are manufactured using 3D-printed ABS/ASA while platinum-cure-silicones are used for the soft ones. Meyer's anesthesiologists collaborated to the realization providing feedback during design and production. The device is evaluated with a 5-point Likert scale questionnaire and results in a useful tool for the training of procedural skills related to difficult intubation as its realism, anatomical geometry, and tactic feedback are positively evaluated.

Keywords: 3D printing , mannequin , pathological airway , simulation

The paper focuses on composites based on inner Al-alloy sheet-based Triply Periodic Minimal Surfaces (TPMSs) structures, which can be manufactured with high porosity (ε), leading to two non-interconnected domains. These conditions favor designing Composite Phase Change Materials (C-PCMs) in which the void domains are filled by the same or different PCMs, which enable thermal energy storage in the form of latent heat. In this paper, we demonstrate that C-PCMs can be designed based on models of their effective thermophysical properties. To this aim, the effective thermal conductivity (λeff) of various Al-based TPMS structures, among which Primitive-Schwarz (PS), Gyroid (G), Diamond (D), and I-graph and wrapped package graph (I-WP), was calculated and analytically modeled. Different filling phases, such as tin, paraffins, or water, were considered to evaluate the influence of the thermal conductivity ratio of the two C-PCM phases at different ε. Furthermore, the transient thermal behavior of Al-based PS C-PCMs was numerically simulated in the extreme cases of temperature ramp or constant heat flux inputs. Low-conductive paraffin and high-conductive tin were selected as filling materials. The results of the analyses revealed that in the first case (temperature ramp), the hybrid Al/paraffin C-PCM exhibited fast phase change, corresponding to peak-type thermal power storage and higher differences between Al and paraffin phase temperatures. Constant heat flux led to a more gradual paraffin melting and heat storage. Instead, the choice of the boundary conditions is less influential on the fully metallic C-PCM response. Combining two different PCMs with the PS lattice further modulates the thermal response of C-PCMs, making them appealing for Temperature Management (TEM) purposes. Finally, an analytical model of the permeability of PS structures was developed based on numerical simulations and compared to highly scattered literature data. Permeability estimation allows the calculation of the Rayleigh-Darcy parameter, setting thresholds for the onset of the convection of the molten PCM phase within the TPMS skeleton, which modifies the thermal response of the C-PCM.

Keywords: Additive manufacturing (AM) , Composite Phase Change Materials (C-PCM) , Energy storage , Triply periodic minimal surfaces (TPMS)

To tackle the challenge of optimizing structural performance and minimizing weight, this research explores the potential of multi-scale bio-inspired design principles in combination with 3D printing. Besides, multi-objective genetic algorithm-based optimization strategies are applied to enhance structure bending and compression resistance. Hollow cylinders, commonly used in piping, structural supports, and biomedical implants, are ideal for this approach as reducing their weight often compromises stiffness and stability, highlighting the need for new design strategies. Herein, the successful combination of bio-inspiration, including plant- and animal-based solutions, and multi-objective optimization plays a core role in the architected structure design to minimize volume and maximize reaction force. Among the various bio-inspired models evaluated, the Horsetail structure has emerged as the most effective, demonstrating 100% increase in the reaction force under bending loads, a 40% improvement in maximum displacement before instability, and a 20% increase in compressive resistance if compared to the reference hollow cylinder. Validation through a combined experimental and numerical approach confirms the accuracy and reliability of the developed models. The successful application of the Material Extrusion process to fabricate these bio-inspired structures proves their practical feasibility. Here, integrating these optimized designs into real-world applications can replace conventional cylindrical pylons with lighter, load-optimized 3D-printed alternatives.

Keywords: Additive manufacturing , Bio-inspired , Lightweight , Multi-axial loadbearing structures , Multi-objective optimization

The solid oxide fuel cell (SOFC) technology has significant potential to become a large-scale, environmentally friendly solution for electricity generation, but realizing this goal hinges on advancements in performance and manufacturing. To succeed in this intent, additive manufacturing (AM) is considered a promising option. This study provides a comprehensive review of patent literature about the AM of SOFC, identifying and comparing goals, implemented AM strategies and clusters of intersections among them. Compared to previous reviews, limited to the scientific literature, some new solutions characterized by unknown goals and AM strategies have been identified. Most common clusters are: lattice structure optimization in SOFC electrodes to optimize porosity solving a compromise between gas diffusion, ionic conductivity and mechanical resistance; the printing of thin walls and self-supporting structures to build complex fuel channels; graded or 3D interface between electrolyte and electrode to improve the ionic flow. Patent bibliometrics and discussion of the results show how industrial and scientific research are patenting ideas rather than solutions ready to be implemented about the AM of the SOFC. The results of this review may be useful to industries interested in patent intelligence on the topic, as well as to researchers to understand which lines of research arouse greater economic interest.

Keywords: Additive manufacturing , Patents , SOFC , Solid oxide fuel cell

Several industries make an extensive use of sandwich structures in structural design, especially in the aerospace and automotive Engineering fields. Sandwich structures are characterised by two thin layers (skins) with a thick, but light, core in the middle, for increased stiffness to bending. The use of additive manufacturing allows us to adopt very light and geometrically complex cores based on the lattice structures concepts. Lattice structures performances are strongly affected by geometrical factors such as the array configuration, the unit cell shape or topology and the relative density. The research presented in this paper is aimed at adopting hybrid sandwich structures with lattice cores to develop structural engineering solutions characterised by high energy absorption properties and low weight for occupant protection in transportation. Actually, these kind of hybrid shock absorbers have been already demonstrated to be a good compromise between crashworthiness performances and weight efficiency. To improve the shock absorbers designs presented in [1], the additive manufacturing technology has been used to produce three different sandwich panels based on a Body Centred Cubic (BCC) unit cell type, such as a Waved Body Centred Cubic (WBCC) unit cell. The Direct Metal Laser Sintering (DMLS) technology has been adopted by using the EOS M290 3D Printer for metals. The shock absorbers' configurations have been numerically designed by modifying the core unit cell topology starting from a linear shape to a curved shape with two different orientation.

Keywords: Additive Manufacturing , FEM , Impact , Lattice Structures

Most of the research on Binder Jetting 3D Printing of AISI316L stainless steel has focused on optimizing the sintering cycle, with less emphasis on the effect of printing parameters on the properties in the sintered state. This study aims to investigate the influence of printing parameters on density, mechanical properties in both the green and sintered states. A Taguchi model was employed to experimentally study four parameters (printhead speed, dark body, powder applicator speed, and shell thickness) at three levels using an L9 orthogonal array. The results showed that green density ranged from 51.3% ± 0.9%–57.0% ± 0.9%, while a sintered density of 98.6% ± 0.08% was achieved. Mechanical testing revealed green strengths between 1.8 ± 0.1 MPa and 6.1 ± 0.8 MPa, while tensile tests provided yield strength of 175.1 ± 3.5 MPa, ultimate tensile strength (UTS) of 532 ± 9.7 MPa, and elongation at break of 87.7% ± 7.5%. ANOVA analysis indicated a statistically significant effect of the dark body (binder saturation) on density and mechanical properties. In the green state, increasing binder saturation enhances green density and bending strength. However, in the sintered state, higher binder saturation results in lower densification and UTS. Powder applicator speed also demonstrated statistical relevance for green density, where lower speeds produced higher density. Conversely, shell thickness and printhead speed had negligible effects on the properties investigated. These findings highlight the most significant printing parameters and their impact of target properties. Additionally, linear models were developed as a tool for optimizing printing parameters based on the desire density and mechanical outcomes.

Keywords: BinderJet 3D printing , Printing parameters , Stainless steel , Taguchi analysis

Lattice structures are architectures of interest for lightweight applications in different fields such as aerospace, biomedical, automotive and structural engineering. Their properties, such as strength to weight ratio, specific strength and energy absorption can be potentially tailored or generally improved by functionally grading. In this study differently graded lattice structures based on Body Centered Cubic (BCC) unit cell, which has a simple basic geometry that can be advantageous for easier 3D printing and a wider use in industry were considered. Linearly graded and symmetrically linearly graded geometries were modelled with a similar value of average relative density in order to compare their structural behaviour with a uniform BCC lattice structure. The different geometries of the graded lattices were 3D printed with stereolithography (SLA), which allowed to obtain a good geometrical accuracy and superficial finish. The printed models were subjected to compression tests and main parameters, specifically, stiffness and energy absorption capability were evaluated.

Keywords: Additive Manufacturing , Functionally grading , Lattices

Additive Manufacturing (AM) of Polyether ether ketone (PEEK) via Fused Filament Fabrication (FFF) presents a versatile and effective solution for producing complex, lightweight, and customized parts with intriguing mechanical properties. PEEK, known for its high-temperature resistance, chemical stability, and biocompatibility, finds applications in diverse fields, from aerospace component manufacturing, benefiting from its lightweight nature to biomedical engineering, requiring patient-specific implants or prostheses. However, AM of PEEK via FFF poses significant challenges, and conventional low-cost 3D printers may not always be well-equipped for this demanding task. In the present study, a low-cost solution was proposed and evaluated through simulations and experimental measurements to enhance the performance of a 3D printer valued at less than $ 6.000. The solution entails designing an accessory that utilizes the printer's second nozzle, incorporating a basic aluminum plate. Through simulations and experimental measurements, it is demonstrated that this accessory provides localized heating in the printing area, thereby increasing the average temperature and improving temperature uniformity.

Keywords: Additive Manufacturing , local heating system , low-cost equipment , PEEK

This research is about designing multi-degree-of-freedom (multi-DOF) compliant mechanisms with decoupled inputs that can be independently locked/unlocked using bistable switches to achieve different combinations of DOFs. A case study mechanism achieving two decoupled rotational DOFs (tip and tilt) is designed, fabricated, and characterized. It can be triggered using two pairs of bistable switches, achieving drastically different states of torsional stiffness for each DOF in four sets of DOF combinations - no DOFs, a tip DOF, a tilt DOF, and both tip and tilt DOFs. Bistability and stiffness cancelation principles are exploited to achieve the desired changes in stiffness. Two flexure elements can be identified - the switch providing a negative stiffness and the cross-axis-flexural-pivot (CAFP) producing a positive stiffness. The mechanism is tuned to achieve static balancing, reaching a near-zero stiffness over much of its range. The pseudo-rigid body model and two-dimensional (2D) finite element model (FEM) are combined defining a fast method to dimension the system. The 3D FEM is simulated to validate the obtained results. For each DOF, the system is tested in two configurations (stiff and compliant) for three cycles over a ±10 deg rotation, achieving a stiffness reduction of around 99%. Comparable stiffness values were measured after triggering the switches more than once, repetitively reaching the required two states of stiffness, confirming the system's usability in practical applications.

Keywords: 3D-printing , binary stiffness , bistable mechanisms , compliant mechanisms , compliant mechanisms and robots , decoupled actuation , flexure systems , lockable degrees-of-freedom , mechanism synthesis and analysis , multi-degrees-of-freedom , negative stiffness , static balancing , stiffness cancelation

Additive manufacturing represents one of the most significant improvements in Industry 4.0. Design for additive manufacturing is the discipline that studies integrated CAD/CAE tools with guidelines for optimizing 3D printing in terms of cost, process time, quality, and precision. In this context, machine learning is used to support control and decision-making activities in additive manufacturing. However, the use of machine learning methods is generally limited to one single process phase. No studies are proposing a machine learning approach focused on different phases of the product lifecycle, from the early design phase to manufactured parts. In the literature, machine learning applications for additive manufacturing regard only one specific phase of the production process. This paper describes current improvements in the integration of additive manufacturing and machine learning, highlighting limitations, and proposes to include different phases of the product lifecycle while designing with machine learning tools. The research provides a guide to develop a new design platform where machine learning supports the engineers in the definition of the product design and process parameters.

Keywords: Design for additive manufacturing , Laser-powder bed fusion , Machine learning , Product lifecycle

Additive manufacturing technology has made remarkable advancements, enabling the development of innovative structures by leveraging existing constraints. A notable area of exploration lies in lattice structures, which offer a promising alternative to solid components. Among the various lattice designs, the body-centered cubic (BCC) structure has gained popularity. Typically, BCC structures comprise diagonal struts, with equal diameters. In this study we propose a control anisotropic strategy by combining BCC and Crossing Cylinder (CC) structure in order to improve characteristics regarding shear and normal Young's modulus. This study aims to investigate the effect of varying the diameters of the center struts of CC and diagonal struts of BCC on two key parameters: the effective Young's modulus and the Zener anisotropic index A. To achieve this, a homogenization technique is employed, allowing the derivation of a stiffness matrix that characterizes the lattice. Subsequently, the Zener anisotropic index A and effective Young's modulus are computed based on the stiffness matrix. To establish a controlled comparison, a specific control strategy is implemented by systematically altering the ratio between the diameters of the center and diagonal struts. This ratio is maintained at values greater than 1 and less than 1 to ensure diverse structural configurations. The results of this research demonstrate that a ratio of 2.5 between the center and diagonal strut diameters yields the most favorable outcome, as evidenced by an obtained Zener anisotropic index A value of 1.

Keywords: Body-centered cubic , Isotropic strategy , Lattice structure

The development of additive manufacturing has made these technologies suitable for fabricating end products. This encourages companies to identify quickly parts in large databases for which switching from traditional manufacturing technologies to additive manufacturing is convenient. Typically, the manufacturing process selection is made by experts who weigh various parameters, but evidence suggests that intelligent systems could beneficially replace or aid this manual selection. One challenge in using manufacturing data for advanced analysis and machine learning is that it is usually unlabeled, and manual data labelling is expensive and time-consuming. This paper deals with the application of an enhanced unsupervised learning algorithm that automatically identifies parts suitable for additive manufacturing based on parts geometry as a preliminary step of process selection. One hundred randomly selected parts were evaluated by manufacturing experts through a survey and then clustered by the proposed algorithm. The comparison of the manual and algorithmic classifications, using unsupervised learning, regarding suitability for additive or traditional manufacturing is the main original contribution of this study. Overall, 78% convergence between most experts' designations and the unsupervised learning algorithm is achieved. For those parts where expert opinions are substantially aligned, the algorithm showed a 90% convergence rate with human choices.

Keywords: Additive manufacturing , Computer-aided design , Hierarchical clustering , Image processing , Machine learning , Unsupervised learning