Documents published in Scipedia

• SAMPLE3D: A Versatile Numerical Tool for Investigating Texture and Grain Structure of Materials Processed by PBF Processes

  Z. Yang, Y. Kuesters, R. Logvinov, M. Markl, C. Körner

  SIM-AM2023.

  
  

  Abstract

  Powder Bed Fusion (PBF) not only enables the fabrication of metal parts with complex geometries in near-net-shape, but also offers the potential to tailor the microstructure and, consequently, the mechanical properties of the final product. In this contribution, we present our in-house developed simulation software SAMPLE3D (Simulation of Additive Manufacturing on the Powder scale using a Laser or Electron beam in 3D), which is designed specifically for simulating grain structure evolution during PBF processes. The core of SAMPLE3D is composed of a finite difference model and a cellular automaton model. The finite difference model is used to obtain the temperature field caused by an electron or laser beam. This temperature field is further used in the cellular automaton model to simulate grain structure development where grain selection as well as nucleation is considered. A range of information can be extracted from the simulation results, such as texture, grain morphology, and grain boundary arrangement. SAMPLE3D provides a way to get insight into the relationship between PBF process strategies and microstructures. SAMPLE3D has been employed to investigate the texture and grain structure evolution of various materials in different research projects.

  Abstract
  Powder Bed Fusion (PBF) not only enables the fabrication of metal parts with complex geometries in near-net-shape, but also offers the potential to tailor the microstructure [...]

  • 190
  •  read

• Multi-Scale Topology Optimization of Bodies with TPMS-based Lattice Structures and Mortar Contact Interfaces

  N. Strömberg

  SIM-AM2023.

  
  

  Abstract

  The combination of topology optimization, lattice structures and 3D printing has quickly emerged as a potential alternative for the design and manufacturing of lightweight components. However, the size of the building chamber restricts the size of this kind of lightweight designs. A possibility to overcome this limitation is to design assemblies of 3D printed lightweight components put together with contact interfaces. To design such an optimal lightweight assembly, the components should not be optimized separately, but the wholeassembly should be optimized simultaneously with all components including their unilateralcontact interfaces. This is the topic of the following work. In this paper, a framework formulti-scale topology optimization of assemblies of bodies with triply periodic minimal surfaces(TPMS)-based lattice structures and unilateral contact interfaces is developed and implementedin 3D. The contact interfaces are formulated for finite element bodies with non-matching meshesusing the mortar approach which in turn is solved by the augmented Lagrangian formulationand Newton’s method. The multi-scale topology optimization formulation, suggested in [1],is set up by defining two density variables for each finite element: one macro density variablegoverned by RAMP (Rational Approximation of Material Properties), and a micro densityvariable governed by representative orthotropic elastic properties obtained by numerical finiteelement homogenization of representative volume elements of the TPMS-based lattice structure. Thus, the macro density variable defines if an element should be treated as a void or be filled with lattice structure, and the micro density variable sets the local grading of the lattice. The potential energy of the system is maximized with respect to the design variables, in such manner no extra adjoint equation is needed for the sensitivity analysis. Both density variables are treated with a density filter, and the macro density variable is also passed a Heaviside filter. The final optimal assembly design is realized by transforming the optimal density fields to implicit surface-based geometries using a support vector machine and Shepard’s interpolation method, which then can be 3D printed as the corresponding stl-file obtained by applying the marching cube algorithm. The implemented framework is demonstrated for three-dimensional benchmark problems.

  Abstract
  The combination of topology optimization, lattice structures and 3D printing has quickly emerged as a potential alternative for the design and manufacturing of lightweight [...]

  • 19
  •  read

• Fluid Flow-Based Topology Optimization of Internal Channels of a LPBF-Manufactured Calibrator Side Lath

  F. Gallego-Bordallo, H. Erdelyi, W. Six, I. Marco, B. Van Hooreweder

  SIM-AM2023.

  
  

  Abstract

  In this study, a method is presented to design embedded cooling channels in an additively manufactured metal part. A fluid flow-based Topology Optimization (TO) methodology was applied on a specific industrial case study with thermal objectives and constraints. The resulting design was 3D-printed and assessed numerically. In addition, the cooling efficiency is compared against that of the original design, which is machined. This work was performed using commercial software tools Simcenter STAR-CCM+ to perform the thermal and fluid flow optimization and simulations; NX to generate a final geometry from optimization results and 3DXpert to assess part printability.

  Abstract
  In this study, a method is presented to design embedded cooling channels in an additively manufactured metal part. A fluid flow-based Topology Optimization (TO) methodology [...]

  • 105
  •  read

• Thermal and Mechanical calibration of multi-material DED AM process simulations on part-scale

  J. Vroon, A. van der Eems, W. van der Brink

  SIM-AM2023.

  
  

  Abstract

  Additive Manufacturing (AM) processes, such as Directed Energy Deposition (DED), offer great potential for producing complex and customized components. To optimize these processes, accurate simulations and numerical modeling techniques are essential. This paper presents a study on the thermal and mechanical calibration of DED AM process simulations on a part-scale. The research aims to develop a comprehensive finite element model that incorporates the multi-physics nature of the DED process, accurately predicting thermal behavior, internal stresses, and distortion of manufactured components. The calibration process involves experimental measurements and simulations using Abaqus software. The thermal calibration involves calibrating parameters such as emissivity, absorptivity, and convection coefficients, while the mechanical calibration focuses on plastic strain properties. Additionally, the study explores the simulation of multi-material prints and functionally graded materials. The results demonstrate that the models can accurately represent thermal and mechanical phenomena, with calibration of material properties playing a crucial role. The paper concludes with recommendations for further validation, including demonstrator prints and investigations into simulation parameters. This research contributes to advancing the understanding and application of DED AM simulations, enabling more accurate and reliable predictions for industrial applications.

  Abstract
  Additive Manufacturing (AM) processes, such as Directed Energy Deposition (DED), offer great potential for producing complex and customized components. To optimize these processes, [...]

  • 71
  •  read

• Simulation-Based Process Optimization Towards Homogeneous Ti6Al4V L-PBF Components

  T. Koenis, M. Montero-Sistiaga, M. De Smit, E. Amsterdam

  SIM-AM2023.

  
  

  Abstract

  In this study, macro-scale thermal simulation of the laser powder bed fusion (LPBF) process is employed to predict and limit geometry-induced overheating of complex Ti6Al4V components. First, the overheating effect is reproduced in tensile specimens. Overheating is found to increase the local oxygen content by almost 80% and lower the elongation at break by over 70% in overheated regions. By employing macro-scale thermal simulations, an automated routine is developed to efficiently optimize the L-PBF process to prevent local overheating. Variable interlayer wait times are numerically optimized to allow cooling of the material without adding manufacturing time where this is not required. In this way, local overheating can successfully be prevented resulting in a more homogeneous temperature distribution during the L-PBF process. This method was found to fully restore the mechanical properties in geometries prone to overheating, resulting in more homogeneous and predictable Ti6Al4V components.

  Abstract
  In this study, macro-scale thermal simulation of the laser powder bed fusion (LPBF) process is employed to predict and limit geometry-induced overheating of complex Ti6Al4V [...]

  • 51
  •  read

• Numerical model for the laser metal deposition additive manufacturing process: Multiphysics modeling and experimental validation

  G. Dali

  SIM-AM2023.

  
  

  Abstract

  Metallic Additive Manufacturing (AdM) technologies (3D printing) is rapidly spreading to a variety of industrial applications. In recent years, advances in AdM have gradually transformed the way in which manufactured products are designed and produced. It enables easy manufacturing of complex shaped parts with high performance, less material waste and short development cycle. Laser Metal Deposition (LMD) is one of the processes in this growing field. This process can produce high performance parts by the injection of powders into a melt-pool created by a laser heat source. However, the LMD is complex and several defects may appear during the printing process. In this context, numerical simulation could be a helpful tool to describe the involved physical phenomena and then to predict the impact of process parameters on the material state. Such numerical tool can predict the heat exchanges and the fluid flow within the molten pool enabling defect prediction and process optimization. In this work, a multi-physics numerical model of the LMD process, at a mesoscopic scale, (i.e. at the layer thickness scale) is developed to predict thermal cycles during fabrication, as well as the complex relationships between part construction and operating parameters. For this purpose, the finite element code COMSOL Multiphysics is used. The developed model takes into account fluid flow and heat transfer in the different phases (gas, substrate and melt pool). As a key feature, the developed model simulates the growth of the track using the generation of droplets when the powder flow is intercepted by the laser beam. Material addition, interface tracking, and strong topological changes are handled using the level set technique. The numerical results are compared to the experimental results for validation purposes. This validation includes the comparison between the predicted molten pool cross-section and measurements from macrographs and high-speed videos.

  Abstract
  Metallic Additive Manufacturing (AdM) technologies (3D printing) is rapidly spreading to a variety of industrial applications. In recent years, advances in AdM have gradually [...]

  • 48
  •  read

• Multiphysical modeling of soft-tissue stent interaction (Plenary Lecture)

  S. Reese, K. Manjunatha, J. Shi, M. Sesa, M. Behr, F. Vogt

  coupled2023.

  
  

  Abstract

  The efficacy of cardiological interventions, including the implantation of prostheses, is highly dependent on patient-specific immunohistology and can be enhanced with computational predictive tools. Therefore, an in silico replication of neointimal hyperplasia shall provide the necessary insights about the biochemical and cellular interactions within the vessel wall, and eventually address the risks of in-stent restenosis in a patient-specific manner. In this context, we set up a multiphysics framework considering key mediators of restenosis and couple them to a continuum mechanical vessel wall model. The governing set of coupled partial PDEs for the underlying mechanobiological system is solved via the finite element method and the results are compared to those obtained using a deep learning framework employing physics-informed neural networks (PINNs). Another interesting cardiological intervention-related problem is the maturation of tissue-engineered cardiovascular implants wherein the evolution of the collagen density affects the tissue’s mechanical behavior. The model we present allows us to predict the evolution of collagen density within textile-reinforced heart valves.

  Abstract
  The efficacy of cardiological interventions, including the implantation of prostheses, is highly dependent on patient-specific immunohistology and can be enhanced with computational [...]

  • 6
  •  read

• A coupled model for the formation of Atherosclerosis due to inflammation processes

  P. Wriggers, M. Gierig, A. Haverich

  coupled2023.

  
  

  Abstract

  Atherosclerosis is a disease in blood vessels that often results in plaque formation and lumen narrowing. It is an inflammatory response of the tissue caused by disruptions in the vessel wall nourishment. Blood vessels are nourished by nutrients originating from the blood of the lumen. In medium-sized and larger vessels, nutrients are additionally provided from outside through a network of capillaries called vasa vasorum. It has recently been hypothesized [1] that the root of atherosclerotic diseases is the malfunction of the vasa vasorum. This, so called outside-in-theory, is supported by a recently developed numerical model [2] accounting for the inflammation initiation in the adventitial layer of the blood vessel. Building on the previous findings, this presentation proposes an extended material model for atherosclerosis formation that is based on the outside-in-theory. Beside the description of growth kinematics and nutrient diffusion, the roles of monocytes, macrophages, foam cells, smooth muscle cells and collagen are accounted for in a nonlinear continuum mechanics setting. Cells are activated due to a lack of vessel wall nourishment and proliferate, migrate, differentiate and synthesize collagen, leading to the formation of a plaque. Numerical studies show that the onset of atherosclerosis can qualitatively be reproduced. Thus, the in silico model backs the new theory.

  Abstract
  Atherosclerosis is a disease in blood vessels that often results in plaque formation and lumen narrowing. It is an inflammatory response of the tissue caused by disruptions [...]

  • 2
  •  read

• Reduced-Order Models in Bayesian solvers for inverse problems

  P. Diez, S. Zlotnik, A. Muixí, A. García-González

  coupled2023.

  
  

  Abstract

  Challenging inverse problems aim at identifying large sets of parameters using data from different sources and diverse accuracy. This is the case of data assimilation for geophysical crust dynamics, were the number of parameters to identify amounts to thousands. In this context, Bayesian inverse solvers combined with Markov-Chain Monte Carlo (MCMC) strategies are an affordable strategy, accounting for the uncertainty of the input data and quantifying also uncertainty of the output. Despite the efficiency of the MCMC approach, the direct problem has to be evaluated an extremely large number of times, many (after the burn-in phase) with the input parameters lying in a narrow range. This is the ideal situation for Reduced-Order Models (ROM): many repeated queries to the model corresponding to parameters lying in a limited manifold. Thus, we aim at applying ROM to large-dimensional parametric forward problems. In this case, it is important optimising the dimensionality reduction technique inherent to the ROM strategy. For instance, Proper Orthogonal Decomposition (POD) is associated with a linear Principal Component Analysis (PCA). PCA is linear in the sense that assumes the reduceddimension manifold to be Euclidean. We explore using kernel PCA (kPCA) to further reduce the dimension, thus devising a kPOD approach. Different options to select physically inspired kernels, based on the knowledge of the problem under consideration, are discussed. Moreover, the computational strategy to explore the feature space (the reduced-dimensional space) is also discussed. 

  Abstract
  Challenging inverse problems aim at identifying large sets of parameters using data from different sources and diverse accuracy. This is the case of data assimilation for [...]

  • 6
  •  read

• Model Order Reduction in the Parametrized Multi-Scale and Coupled Setting

  K. Veroy-Grepl

  coupled2023.

  
  

  • 0
  •  read