Shape deviation modeling and compensation in additive manufacturing (AM) are pivotal for achieving high geometric accuracy and enabling industrial-scale production. While traditional analytical and statistical methods laid the foundation, recent advancements in machine learning (ML) have improved prediction and compensation precision. However, critical challenges persist, including generalizability across complex geometries and adaptability to position-dependent variations in batch production. Traditional methods of controlling geometric deviations often rely on complex parameterized models and repetitive metrology, which can be time-consuming yet not applicable for batch production. In this paper, we present a novel, process-agnostic approach to address the challenge of ensuring geometric precision and accuracy in position-dependent AM production. The proposed GraphCompNet presents a novel computational framework integrating graph-based neural networks with a generative adversarial network (GAN)-inspired training paradigm. The framework leverages point cloud representations and dynamic graph convolutional neural networks (DGCNNs) to model intricate geometries while incorporating position-specific thermal and mechanical variations. A two-stage adversarial training process iteratively refines compensated designs using a compensator-predictor architecture, enabling real-time feedback and optimization. Experimental validation across various shapes and positions demonstrates the framework’s ability to predict deviations in freeform geometries and adapt to position-dependent batch production conditions, significantly improving compensation accuracy (35% to 65%) across the entire printing space, addressing position-dependent variabilities within the print chamber. The proposed method advances the development of a Digital Twin for AM, offering scalable, real-time monitoring and compensation capabilities. This work bridges critical gaps in AM process control, paving the way for high-precision, automated, and industrial-scale design and manufacturing systems.

Optimizing heterogeneous elastic material distribution on a 3D part to achieve desired deformation behavior is an important task in computer-aided design and additive manufacturing. This paper presents a solution to this problem, which involves interactive design, automatic deformation generation, and optimization of spatial distribution of heterogeneous elastic materials. Our method improves previous techniques in three aspects. First, we incorporates a geometric deformation-based interactive design into FEM-based optimization, which makes the solution less dependent of initial guesses of Young’s modulus values and it more likely to produce the target design even with sparse user input of displacements and forces at a limited set of mesh vertices. Second, we formulate the problem as an L2- or L0-optimization problem. The L2 formulation outputs smoothly varying heterogeneous material distribution that accommodates multiple functions within a single part. The L0 formulation achieves the computation of sparse material distribution in one step, which is beneficial for additive manufacturing with multi-material printers. Third, we utilize the adjoint method to derive formulae for efficiently computing the gradient of the objective functions, making it possible to quickly solve the optimization problem in the full-dimensional space of materials, which was previously infeasible. The experiments demonstrate the robustness and efficiency of our approach.

Metal Sintering is a necessary step for Metal Injection Molded parts and binder jet such as HP's metal 3D printer. The metal sintering process introduces large deformation varying from 25 to 50% depending on the green part porosity. In this paper, we use a graph-based deep learning approach to predict the part deformation, which can speed up the deformation simulation substantially at the voxel level. Running a well-trained Metal Sintering inferencing engine only takes a range of seconds to obtain the final sintering deformation value. The tested accuracy on example complex geometry achieves 0.7um mean deviation for a 63mm testing part.

A finite difference-based 3D phase-field model is developed to investigate the spherulite growth at the mesoscopic scale during the isothermal crystallization of polyamide (PA) 12. The model introduces a phase-field variable to distinguish the crystalline and amorphous phases of polymers. The phase-field evolution equation is coupled with the heat conduc-tion equation that considers the latent heat of crystallization. The evolution equations in-troduce both the dimensionless diffusivity and latent heat that are dependent on the crys-tallization temperature. A high-order finite difference-based numerical framework is ap-plied to the phase-field model. Both the qualitative simulation results of the phase-field model such as the crystal morphologies and the quantitative results including the radial crystal growth rate, degree of crystallinity, and lamellar thickness are validated against ex-periments. The simulation for single-crystal growth shows that a high crystallization tem-perature results in a large crystal with a slow radial growth rate. The simulation for multi -crystal growth shows that the crystals impinge on each other and finally fill the whole domain during crystallization, which further demonstrates the capability of the model in simulating the spherulite growth during isothermal crystallization of polymer melts. (c) 2023 Elsevier Inc. All rights reserved.

Crystallization of polyamide 12 (PA12) is an essential process requiring thorough investigation for evaluating the mechanical properties after the polymer parts are manufactured. The change in crystallization temperature results in different crystallization behaviors for PA12. Hence, the crystal morphology of PA12 achieved provides important information about crystallization behavior, especially for those produced through additive manufacturing due to its heterogenous cooling rate in a single print bed. With the lack of literature investigating PA12 crystallization using phase field modeling, this paper aims to simulate the spherulite morphology of PA12 undergoing isothermal crystallization. This model is compared with the spherulite morphologies obtained from the optical microscopy test. The model shows that PA12 spherulites have thicker dendrites when the isothermal temperature is higher. The present phase-field model can determine the spherulite morphologies of bulk printed PA12 based on the crystallization condition and be used to evaluate the properties of the printed part.

Multi Jet Fusion (MJF) is a 3D-printing process capable of fabricating large-scale polymer structures. Herein, we present a framework for MJF-printed lattices with tunable stiffness and strength based on an empirical analysis of structural behavior.

This paper addresses the challenges of designing mesh convolution neural networks for 3D mesh dense prediction. While deep learning has achieved remarkable success in image dense prediction tasks, directly applying or extending these methods to irregular graph data, such as 3D surface meshes, is nontrivial due to the non-uniform element distribution and irregular connectivity in surface meshes which make it difficult to adapt downsampling, upsampling, and convolution operations. In addition, commonly used multiresolution networks require repeated high-to-low and then low-to-high processes to boost the performance of recovering rich, high-resolution representations. To address these challenges, this paper proposes a self-parameterization-based multi-resolution convolution network that extends existing image dense prediction architectures to 3D meshes. The novelty of our approach lies in two key aspects. First, we construct a multi-resolution mesh pyramid directly from the high-resolution input data and propose area-aware mesh downsampling/upsampling operations that use sequential bijective inter-surface mappings between different mesh resolutions. The inter-surface mapping redefines the mesh, rather than reshaping it, which thus avoids introducing unnecessary errors. Second, we maintain the high-resolution representation in the multi-resolution convolution network, enabling multi-scale fusions to exchange information across parallel multi-resolution subnetworks, rather than through connections of high-to-low resolution subnetworks in series. These features differentiate our approach from most existing mesh convolution networks and enable more accurate mesh dense predictions, which is confirmed in experiments.

With the advancements in 3D printing technologies, it is extremely important that the quality of 3D printed objects, and dimensional accuracies should meet the customer's specifications. Various factors during metal printing affect the printed parts' quality, including the power quality, the printing stage parameters, the print part's location inside the print bed, the curing stage parameters, and the metal sintering process. With the large data gathered from HP's MetJet printing process, AI techniques can be used to analyze, learn, and effectively infer the printed part quality metrics, as well as assist in improving the print yield. In-situ thermal sensing data captured by printer-installed thermal sensors contains the part thermal signature of fusing layers. Such part thermal signature contains a convoluted impact from various factors. In this paper, we use a multimodal thermal encoder network to fuse data of a different nature including the video data vectorized printer control data, and exact part thermal signatures with a trained encoder-decoder module. We explored the data fusing techniques and stages for data fusing, the optimized end-to-end model architecture indicates an improved part quality prediction accuracy.

The effects of build direction on the tensile properties, tension-tension low cycle fatigue behavior, and failure mechanism of polyamide 12 parts fabricated by Multi Jet Fusion have been investigated by addressing the roles of formed voids, sintering interfaces, and geometrical imperfections. Results indicate that the Young's modulus of bulk material with vertical build direction is higher than that of bulk material with horizontal build direction due to the carbon black enhanced sintering interfaces. The tensile strength of the printed bulk material with horizontal and vertical build directions is comparable. It is found that the anisotropic tension-tension low cycle fatigue behavior for the bulk material is significant, while that for the re-entrant lattice structure is weak. The anisotropic fatigue strength of bulk material is derived from the anisotropic distribution of internal voids and sintering interfaces. The fatigue behavior of the re-entrant lattice structure is determined by the geometrical imperfections. The coalescence of the large voids or clusters of voids contributes to the crack initiation and propagation processes that lead to the failure of the bulk material. While the failure of re-entrant lattice structure is mainly determined by the intersection of strut junction rather than the void defect. This result provides the fatigue data that can be used for the structural design with Multi Jet Fusion polyamide 12 as well as reveals the effects of voids, sintering interfaces, and geometrical imperfections on the fatigue behavior of the printed parts.

Multi-Jet-Fusion (MJF) is a relatively new powder-based face-sintering additive manufacturing technique that exhibits great potential in high-volume manufacturing due to its rapid printing speed. Extensive research has been conducted on the mechanical properties of MJF-printed polyamide 12 (MJF PA12) by investigating the powder characteristics and build orientations. However, no work has been conducted to study the effect of humidity and physical ageing on the mechanical performance of MJF PA12 specimens. Here, the physical and mechanical properties of MJF PA12 specimens that were printed in different build orientations and stored under ambient and dryer conditions for 474 days were investigated. It was found that specimens stored under ambient conditions exhibited a drop in glass transition temperature due to moisture absorption followed by an increase, which could indicate that ageing of the specimens occurred. The moisture absorption was also found to have an insignificant effect on the crystallinity and crystallite size of the specimens. The change in mechanical properties, such as the tensile modulus and ultimate tensile strength, was more significant for specimens stored in ambient conditions than those in dryer conditions. It was also found that the build orientation does not have a significant impact on the moisture absorption rate, glass transition temperature, and the change in the trend of mechanical properties. In addition, an artificial neural network was adopted to predict the mechanical properties of the MJF PA12 specimens under the influence of physical ageing, humidity, and build orientations. The neural network adopted in this work can predict the change in the mechanical properties of differently-orientated MJF PA12 with different moisture content over time. The neural network prediction could be useful in providing guidance on the structure and risk assessment of the parts before print.

In multi jet fusion process, the thermal history varies at different locations inside the printing chamber resulting in the dependence of crystallinities of the printed parts. As performing experimental test is time consuming and costly, it is desirable to have the crystallinity be predicted even before the parts are printed. Thus, this work presents a crystallinity prediction method based on machine learning for MJF-printed polyamide 12. In the model, the predicted thermal profiles and the experimental measurements of crystallinities were employed to train and optimize the machine learning regression model. The prediction results explain the formation of crystallinity is significantly affected by the duration of first cooling stage, temperature at the end of printing process, the duration of extremely low cooling rate, and the cooling condition of the second cooling stage. Additionally, an optimized Ridge regression model has been found to predict the crystallinity with the accuracy of 93.6 %.

Multijet fusion (MJF) is a novel additive manufacturing (AM) process for the powder bed fusion of thermoplastic polymers. In recent years, MJF has been established as a promising manufacturing technique for the rapid prototyping and production of fully functional, large‐scale components at high resolution. Like other powder bed fusion techniques, MJF permits the printing of parts without support material, and does not require chemical postprocessing steps, making it especially suitable for fabricating parts with complex and intricate internal features. However, the effects of common build volume parameters (such as part orientation, location, and interpart spacing) on mechanical properties are currently not well understood. This study aims to provide a thorough and comprehensive description of the trends in mechanical properties observed in polyamide (PA)‐11 and PA‐12 samples printed using the MJF process. Tensile, flexural, and impact specimens are printed using varying build volume packing parameters and tested according to the corresponding ASTM standards. The fracture surfaces of tested specimens are studied and characterized to further understand the method of failure propagation through MJF‐printed PA‐11 and PA‐12. As this work demonstrates, determining an optimal bed packing algorithm will be instrumental for optimizing desired mechanical properties of printed parts.

Polyamide 12 (PA12) is a semi-crystalline polymer as its crystallisation behaviour varies due to manufacturing conditions. In this work, the crystallisation behaviour of PA12 was observed through experiments conducted at a constant cooling rate from different maximum heated temperatures. From the differential scanning calorimetry (DSC) experiment, PA12 exposed at a higher temperature crystallised at a lower temperature. The crystallisation shrinkage observed from thermomechanical analysis (TMA) was lower after PA12 was exposed at a higher temperature. Judging from the crystal growth process of PA12 from in-situ optical microscope and phase field modelling, the crystals of PA12 exposed at a higher temperature were found to be smaller and more compact. The exposure to a high temperature allows PA12 to have lower melt viscosity, ensuring the material to sinter evenly, and this enhances better molecular interactions from polymer chains for the formation of crystals.

Warpage of printed parts, especially for multi-jet fusion (MJF) process, is believed to have strong correlation with crystallization shrinkage. Hence, this article investigates the shrinkage phenomenon of polyamide-12 (PA12) during crystallization process. Various cooling rates show that PA12 exhibits different crystallization behaviour, which affect the crystallinity, crystal phase and crystal shrinkage when performed in MJF. The investigation shows that crystallization shrinkage of PA12 coincides with the exothermic reaction of crystallization process. With cooling rate as slow as 0.2 degrees C/min, PA12 sample shrinks 1.64% during the crystallization process with crystallinity calculated at 24.9%. X-ray diffraction (XRD) suggests shrinkage are closely related to lamellar folding and growth of gamma crystal phase within PA12 samples during crystallization. Spherulites are found to be smaller in diameter with faster cooling rate. These investigations provide detailed understanding of the crystallization behaviour of PA12 for MJF process.

The viscoelastic-viscoplastic deformation of Multi Jet Fusion-printed polyamide 12 (MJF PA12) was investigated with experimental and numerical approaches. Multi-loading–unloading–recovery tests were conducted to distinguish between the viscoelastic and viscoplastic deformations. The influence of the void defects on deformation of PA12 was investigated through the micro-computed tomography (μCT) and field emission scanning electron microscope. A finite-strain viscoelastic-viscoplastic constitutive model based on the logarithmic stress rate for the matrix of MJF PA12 was developed within the thermodynamic framework as well as an introduction of the accumulated plastic deformation–induced damage into the proposed model. A representative volume element was modeled based on the μCT results of MJF PA12. The simulated results, such as stress–strain and strain–time curves, agreed with the experimental results. Moreover, the model revealed the mechanism that the low tensile ductility of MJF PA12 is caused by the increase in strain localization and narrowing of the shear band.

This paper presents a research vision to design a large-scale cyberphysical systems (CPS) experimental framework to enable collaborative and coordinated molecular biology studies. This framework will be based on the integration of CPS with microfluidic biochips and cloud computing. It has the potential to drastically advance personalized medicine through knowledge fusion among many research groups, and synchronization of research planning. This framework therefore leads to a better understanding of diseases such as cancer, and helps researchers in identifying effective treatments. A case study from cancer research is discussed to explain the significance of our framework in promoting coordinated genomic studies.

In the paper a digital material design framework is presented to compute multi-material distributions in three-dimensional (3D) model based on given user requirements for additive manufacturing (AM) processes. It is challenging to directly optimize digital material composition due to extremely large design space. The presented material design framework consists of three stages. In the first stage, continuous material property distribution in the geometric model is computed to achieve the desired user requirements. In the second stage, a material dithering method is developed to convert the continuous material property distribution into 3D printable digital material distribution. A tile-based material patterning method and accordingly constructed material library are presented to efficiently perform material dithering in the given 3D model. Finite element analysis (FEA) is used to evaluate the performance of the computed digital material distributions. To mimic the layer-based AM process, cubic meshes are chosen to define the geometric shape in the digital material design stage, and its resolution is set based on the capability of the selected AM process. In the third stage, slicing data is generated from the cubic mesh model and can be used in 3D printing processes. Three test cases are presented to demonstrate the capability of the digital material design framework. Both FEA-based simulation and physical experiments are performed; in addition, their results are compared to verify the tile-based material pattern library and the related material dithering method.

This book provides a comprehensive set of optimization and prediction techniques for an enterprise information system. Readers with a background in operations research, system engineering, statistics, or data analytics can use this book as a reference to derive insight from data and use this knowledge as guidance for production management. The authors identify the key challenges in enterprise information management and present results that have emerged from leading-edge research in this domain. Coverage includes topics ranging from task scheduling and resource allocation, to workflow optimization, process time and status prediction, order admission policies optimization, and enterprise service-level performance analysis and prediction. With its emphasis on the above topics, this book provides an in-depth look at enterprise information management solutions that are needed for greater automation and reconfigurability-based fault tolerance, as well as to obtain data-driven recommendations for effective decision-making.