Kapitel

• FrontmatterSommitsch, Christof; Enzinger, Norbert; Mayr, Peter; 10.3217/978-3-99161-089-2-000
• EVAPORATION AND MASS TRANSFER IN WELDINGMENDEZ, P. F.; 10.3217/978-3-99161-089-2-001This paper reviews the foundations of evaporation and mass transfer in welding operations, presenting for the first time a treatment with simultaneous barriers to mass transfer in the molten metal, at the evaporating surface, and in the surrounding gases. Evaporation is a crucial aspect of welding operations, as it controls fume emissions, arc properties, droplet temperature, porosity, and onset of explosive metal transfer in GMAW of Al-Mg alloys. When evaporation is large a recoil pressure is significant, leading to keyhole penetration in focused beams. The treatment accounts for multicomponent mass transfer, and explicit expressions for the resistance to mass transfer for each component and each stage are provided. A novel treatment for diffusion in the melt is proposed by using an equivalent mass transfer resistance in the melt which allows for the simultaneous consideration of all mass transfer resistances. The treatment of mass transfer also includes a treatment of recoil pressure. The theory is applied to mass transfer and evaporation in the electrode tip in GMAW, with detailed analysis of three experiments with results supporting the theory. Based on the theoretical foundations a criterion for explosive transfer is proposed, reproducing the type and intensity of explosive transfer observed. The treatment presented is also a foundation for a comprehensive formulation of mass and energy balances in deposition rate.
• NUMERICAL SENSITIVITY ANALYSIS OF MATERIAL PARAMETERS IN TIG WELDING ARC MODELLINGLE GAL LA SALLE, M.; CADIOU, S.; COURTOIS, M.; CARIN, M.; BROSSE, A.; 10.3217/978-3-99161-089-2-002The simulation of the TIG/GTAW process with arc-melt pool coupling involves several material properties that are not well known. The main goal of this study is to quantify the numerical sensitivities of these arc material properties on the melt pool behaviour and provide recommendations on influential parameters. The most sensitive properties are thermal conductivity, density, heat capacity and viscosity of argon. For instance, a 10% increase in 𝐶𝑝 can result in an increase of more than 23% in the melt pool volume. This sensitivity study includes a detailed presentation of the 2D axisymmetric magneto-thermo-hydrodynamic finite element simulation developed with a dynamic coupling between the arc and melt pool and free surface deformation of the plasma-anode interface.
• ENHANCING GMAW SIMULATIONS THROUGH A HYBRID EULERIAN AND LAGRANGIAN METHOD CONSIDERING AN INCLINED WELDING TORCHMOKROV, O.; WESTHOFEN, L.; BENDER, J.; SHARMA, R.; REISGEN, U.; 10.3217/978-3-99161-089-2-003The comprehensive simulation of gas metal arc welding (GMAW) remains challenging due to the continuously evolving topologies of both the welding arc region and the weld seam formation zone. GMAW involves phenomena, such as melting, hydrodynamic flows within the weld pool, and crystallization, occurring on timescales significantly distinct from the rapid plasma physics processes in the arc. This plasma exhibits pronounced heterogeneity in voltage gradients, necessitating highly precise modeling of anode and cathode layers. In contrast, traditional simulation methods couple liquid metal and arc behaviour within a single domain, typically employing Volume of Fluid (VOF) or Level Set techniques. These methods often rely on substantial simplifications in the plasma boundary layers, thereby compromising predictive accuracy. To overcome these limitations, this paper presents a hybrid modeling approach coupling a Lagrangian Smoothed Particle Hydrodynamics (SPH) model for weld pool and droplet dynamics with an Eulerian framework tailored for the arc zone. The SPH method effectively captures large-scale phenomena in the weld pool, including heat and mass transfer, droplet detachment, and solidification. On the other hand, the Eulerian model accurately represents the complex electromagnetic and thermal behaviour within the anode and cathode layers of the arc plasma This study also addresses torch inclination effects on weld pool dynamics, heat flux distribution, and current density on the cathode, providing insights into their influence on weld geometry. Through the synthesis of both, Lagrangian and Eulerian approaches, the proposed method provides a robust tool for addressing the multiscale, multiphysics nature of the GMAW process. Results from 3D calculations include quasi-stationary Eulerian solutions for the plasma domain and transient SPH simulation of the heat and mass transfer for a drag/push weld. Although still under active development, this hybrid model shows significant potential for enhancing the predictive capabilities of GMAW simulations, thus offering additional possibilities in process optimization and quality improvement in industrial welding.
• TOWARDS ACCURATE THERMAL SIMULATIONS IN PBF: A NOVEL CALIBRATION STRATEGY FOR GAUSS-GOLDAK HEAT-SOURCE MODELSHOFMANN, M.; MAYER, T.; 10.3217/978-3-99161-089-2-004Laser Powder Bed Fusion (PBF-LB) is an advanced additive manufacturing process that fabricates near-net-shape metal components by selectively melting and solidifying powder layers with a laser beam. The intense, localized energy input generates complex thermal cycles that strongly influence microstructural evolution, residual stresses, and mechanical performance. Accurate prediction and control of this thermal history are therefore critical for optimizing part quality and process reliability. Finite Element (FE) models employing the Gauss-Goldak heat source are widely used to simulate melt pool temperature fields and support residual stress analyses. However, the accuracy of these requires precise calibration of the heat source parameters, particularly the laser absorptivity and volumetric heat flux distribution, which vary with process conditions. This study introduces a novel calibration approach combining melt pool geometry data and in-situ thermographic measurements from the bottom side of a substrate with thermal simulations to inversely identify these parameters. The method enables tailored tuning of the thermal heat input model for specific processing conditions, substantially improving the accuracy of transient temperature predictions. Bayesian optimization was used to effectively identify absorptivity from the far-field temperature, and the shape parameters of the Gauss-Goldak heat source from the melt pool geometry. The calibrated model demonstrates excellent agreement with experimental data, achieving deviations in simulated melt pool dimensions below 2.8%. Thereby, absorptivity emerges as the key parameter governing the far-field thermal response, underscoring the importance of its precise calibration for accurate part-scale thermal simulations. Furthermore, it is shown that, within the investigated parameter space, the shape parameters of the Goldak model exert negligible influence on the far-field temperature distribution, justifying a two-step approach. Although the calibration was performed using the Gauss-Goldak model, the approach is generalizable and can be extended to alternative heat input formulations. This enhanced modeling fidelity enabled by an accurately calibrated heat-source can facilitate more reliable residual stress and distortion forecasts in PBF manufacturing.
• MULTI-PHYSICS SIMULATION AND TRAJECTORY OPTIMIZATION IN LASER WELDING OF COPPER HAIRPIN WINDINGS10.3217/978-3-99161-089-2-005The development of a digital process chain for laser welding copper hair pin windings is essential for advancing the manufacturing of components used in electrical vehicles, electronic devices, and energy storage systems. Copper's high thermal conductivity and low infrared light absorption can lead to process instabilities during laser welding, such as melt pool spatter and porosity, resulting in inconsistent weld quality. Understanding the physics involved in laser welding - from keyhole and melt pool formation to the solidification of the melt pool - is crucial. This study employs Cradle CFD for multi-physics analysis to simulate and optimize the laser welding process. This method captures the laser-material interaction, thermal and fluid dynamics, and material solidification. By adjusting welding trajectories, the behaviour of the keyhole and melt pool formation is analyzed. To limit the number of experiments one of the simulation results are validated using experiments. The cut sections are used to validate the melt pool profile, and satisfactory correlation with the experiment results is observed. Additionally, this study investigates thermal and mechanical responses of hairpin windings in laser welded assemblies using FEM-thermo-mechanical and coupled electric-thermal-structural simulations. Six welding locations were analyzed, with cooling simulations revealing heat transfer, thermal stresses, and distortions near constraints. When a 300 A, 20 kHz alternating current was applied, different air cooling strategies showed that forced convection achieved lower and more stable temperatures and deformation than natural convection. The results highlight the need for thicker internal connectors and enhanced support to withstand high temperatures and deformation during operation.
• A STATISTICAL INVESTIGATION OF THE LASER ENERGY ABSORPTION AND KEYHOLE STABILITY IN HIGH-POWER LASER BEAM WELDINGMENG, X.; PUTRA, S.; BACHMANN, M.; RETHMEIER, M.; 10.3217/978-3-99161-089-2-006The behaviour of the weld pool in high-power laser beam welding is significantly influenced by laser absorption and keyhole stability, which determines the final weld qualities. However, their dynamic features and multi-coupled interactions make in-depth analyses difficult. This study addresses the challenges by conducting a thorough statistical evaluation of the effects of welding parameters on laser absorption and keyhole fluctuations, using experimental investigations and a well-validated numerical model. From a statistical aspect, the laser energy distribution and the keyhole collapse, commonly considered highly time-varying, show clear regularities. Three distinct regions of the time-averaged energy distribution are identified. The possibility of the keyhole collapse positions obeys a universal normal distribution. The statistical data show greater potential in revealing some well-known, industry-related but unclearly explained findings, such as the saturation of the weld penetration with increasing heat input and the physical basis of the contributions of different welding parameters in the porosity reduction.
• UNDERSTANDING THE ORIGIN AND PREDICTION OF LASER WELDING DEFECTS IN HIGHLY REFLECTIVE METALS: A NUMERICAL STUDY APPLIED TO A2219 ALLOYCHETAN, B. F.; COURTOIS, M.; CADIOU, S.; CARIN, M.; NAIN, V.; MACHI, F.; ALVES FERREIRA, A.; 10.3217/978-3-99161-089-2-007A new approach to modeling and simulating multiple reflections of a laser beam has been developed under COMSOLMultiphysics®. It is based on a special coupling between a modified ray tracing method and Level-set. Initial application results for laser welding on aluminum in static pulse and fusion line cases are compared with theory and experiments.
• RECONSTRUCTION OF THE TIME-AVERAGED KEYHOLE GEOMETRY IN LASER BEAM WELDING ASSISTED WITH ELECTROMAGNETIC EFFECTYANG, F.; MENG, X.; PUTRA, S. N.; BACHMANN, M.; RETHMEIER, M.; 10.3217/978-3-99161-089-2-008In laser beam welding (LBW), the time-averaged keyhole shape provides statistical insights into the process compared to its transient geometry, offering a deeper understanding of the overall keyhole behaviour. However, capturing the time-averaged keyhole shape through experimental methods remains challenging. In this paper, a reconstruction algorithm for the time-averaged keyhole is developed and integrated into a three-dimensional transient multi-physical coupled numerical model. The algorithm can accurately capture the key characteristics of the keyhole, including its diameter and centroid. In addition, it can also successfully reproduce the experimentally observed phenomena of keyhole tailing. The overall shape of the keyhole appears smooth, without exhibiting obvious instability features. Furthermore, the time-averaged keyhole shape is compared under different magnetic flux densities when an external oscillating magnetic field is applied. The results indicate that the application of external magnetic fields does not fundamentally alter the overall keyhole shape. With increasing magnetic flux density, the trailing tail becomes progressively less pronounced and a noticeable increase in the curvature of the rear wall is observed. The spatial-average of standard deviation of the keyhole diameter can serve as an effective index for evaluating the keyhole instability. Keyhole stability in LBW of aluminium alloys is improved under the assistance of electromagnetic fields, and this stabilization is positively correlated with increasing magnetic flux density.
• IMPACT OF MODELING ASSUMPTIONS ON SIMULATED MELT POOL DYNAMICS COMPARED TO HIGH-SPEED X-RAY OBSERVATIONSPLEWINSKI, J.; HENNIG, F.; FORSTER, C.; BELOUSOV, S.; ZAKIROV, A.; KORNEEV, B.; SPURK, C.; HUMMEL, M.; OLOWINSKY, A.; BECKMANN, F.; MOOSMANN, J.; SCHMIDT, M.; KÖSTLER, H.; MARKL, M.; 10.3217/978-3-99161-089-2-009Laser beam welding is a widely used advanced manufacturing process for joining metallic components across various industries. The process is prone to various defects, with solidification cracks being among the most common. Understanding the evolution of the solidification front, including temperature gradients, cooling rates and solidification velocity, is crucial to assess the conditions leading to crack formation. Direct experimental measurement of transient solidification conditions is extremely challenging, while numerical simulations of melt pool dynamics provide complementary insights by predicting local solidification conditions. In this work, we combine high-speed synchrotron X-ray imaging, which captures the evolution and velocity of the solidification front, with state-of-the-art numerical melt pool simulations that are able to provide local solidification conditions. Comparison of experiments and simulations reveals a mismatch in the quasi-stationary welding regime, primarily arising from the neglect of the gas phase in the simulations. Using a reduced model and simulation setup, we demonstrate how changes in modeling assumptions influence melt pool dynamics and shape. Our results highlight that, for deep-penetration laser welding, proper physical modeling of the vapor-filled keyhole is essential for accurate predictions of melt pool and solidification conditions.
• NUMERICAL SIMULATION OF LASER BEAM WELDING WITH DYNAMIC BEAM SHAPE SEQUENCINGDURÁN, C.; ZENZ, C.; FLORIAN, T.; BUTTAZZONI, M.; OTTO, A.; 10.3217/978-3-99161-089-2-010Dynamic beam shaping enables diverse possibilities to influence the performance of laser-based manufacturing processes, for example by tailoring the spatial intensity distribution of light in real time. One such possibility is shape sequencing, which involves switching between two or more beam shapes during laser-matter interaction, with each shape being active for a specified time frame (shape duration) within the sequence. When properly understood and utilized, shape sequencing could serve as an effective optimization approach in high-power laser materials processing applications such as Laser Beam Welding (LBW). In this work, a multi-physical simulation model is used as a comprehensive framework to evaluate the impact of shape duration on the dynamics of the LBW process, with the aim of mitigating instabilities that can lead to defects such as pores, spatter, lack of penetration, cracks, etc. The long-term goal is not only to identify optimal welding parameters, but also to understand why certain shaping strategies lead to better or worse process outcomes. As a demonstration, an aluminum-magnesium alloy is chosen as the workpiece material for a deep penetration welding (DPW) scenario. The numerical model is initially validated by comparing simulation results with macrographs of experimentally produced welds and is subsequently used to analyze a welding setup incorporating dynamic beam shape sequencing. The study reveals that shape sequencing can effectively induce vibrations at predefined frequencies on the keyhole surface, manifesting as periodic pulses whose width is determined by the shape durations. This pulsed excitation is subsequently transmitted to the melt pool, influencing its behavior. Certain shape durations favor a more or less effective excitation of the melt pool.
• MULTIPHYSICS SIMULATION OF SPATIAL-TEMPORAL BEAM SHAPING IN LASER POWDER BED FUSION USING OPENFOAMZHOU, B.; BARODE, J.; NADIMPALLI, V. K.; BAUCH, A.; HERZOG, D.; HATTEL, J. H.; BAYAT, M.; 10.3217/978-3-99161-089-2-011Metal additive manufacturing (MAM) has been applied in various advanced fields due to its flexibility and superiority in producing complex geometries, compared to subtractive manufacturing methods. The powder bed fusion using a laser beam (PBF-LB) process, as a major subbranch of MAM, involves multiple complex physical phenomena, e.g. phase changes, buoyancy, surface tension, evaporation, Marangoni effects, melting, solidification, and microstructural evolution. The interplay between process settings and melt-pool conditions has a significant influence on part quality. Multiphysics simulations that incorporate these phenomena are considered as a robust and reliable tool to study this complex interplay. This study simulates the PBF-LB process for copper and stainless steel 316L (SS 316L) under infrared laser irradiation at the deposition scale, using the open-source code OpenFOAM. The simulation results demonstrate good agreement with benchmark experimental data in terms of the transverse cross-section profiles of the melt regions when applying spatial and temporal beam shaping in PBF-LB. These beam shaping techniques are efficient ways to adjust the melt-pool morphology and eventually to improve the print quality. In this study, continuous Gaussian and ring-shaped laser beam profiles, and pulsed beam shaping techniques are modelled, and their effects on the melt pool are investigated. Moreover, for copper, there has been long controversy in the literature in terms of the correct absorption rate of laser radiation. We hence explore the influence of absorptivity on the melt-region cross-sectional dimension. This work focuses on simulating the melt pool on metal substrates and offers insights into the fundamental impacts of the process settings on melting behavior and absorptivity while using beam shaping.
• ADVANCED PROCESS MODELING AND SIMULATION OF MULTI-MATERIAL DIRECTED ENERGY DEPOSITIONOSMANOGLU, E.; CANAZ, C.; HEMPEL, N.; MAYR, P.; 10.3217/978-3-99161-089-2-012As technology advances, materials face increasing demands, especially in the aerospace industry. Metal additive manufacturing (AM), specifically directed energy deposition (DED), addresses these needs by enabling multi-material deposition and in-situ alloying, producing functionally graded materials (FGMs) with tailored properties by combining different materials. Numerical simulations can be used to predict temperature fields and distortions in the DED process, reducing costly trial runs. However, incorporating FGMs in such thermomechanical simulations of DED process requires development of new process models and creation of material models for specific alloy combinations. Using Simufact Welding and extending the capabilities of the software, a DED process is modelled, and material compositions are assigned to layers to construct FGMs virtually in this work. Initial process simulations of CuCrZr-IN718 FGM deposition on a CuCrZr substrate demonstrate successful modelling of the process. These models correlate well with experimental results, predicting temperature fields and distortions effectively. Future work includes further validation and scaling simulations for industrial applications.
• ADVANCED THERMAL MANAGEMENT STRATEGY FOR ARC-BASED DIRECTED ENERGY DEPOSITION ASSISTED BY NUMERICAL MODELINGREINDL, T.; HEMPEL, N.; MAYR, P.; 10.3217/978-3-99161-089-2-013Arc-based Directed Energy Deposition (DED-Arc) offers major industrial potential due to its versatile application possibilities and attractive advantages, such as the flexible, time-efficient near-net-shape manufacturing of large-volume components. However, there are obstacles to overcome, such as the complex thermal history, which limits process stability, component quality, and process efficiency. Therefore, establishing this technology in industry requires solutions capable of monitoring and managing temperature fields. In this context, the focus is increasingly shifting toward real-time-capable, physically based digital solution frameworks. Current efforts primarily emphasize monitoring, simulation, or trial-and-error methods. There is a lack of direct coupling between energy-temperature metrics, standardized analyses and classifications, and effective process optimizations, including adaptive control strategies and design foundations for advanced thermal management. This work aims to strategically analyze and classify the coupled data from an industrial-grade digital shadow (DS) and numerical modeling. Based on this, targeted optimization measures and process control indicators are identified to enable advanced thermal management. To achieve this, a novel thermal optimization strategy (TOS) was developed to optimize the thermal process of an exemplary tube geometry. Optimizations in path planning, interlayer temperature (ILT) profiles, and temperature change rates were detected and addressed with specific countermeasures. An analysis of the heat balance revealed specific temperature and heat flow gradients providing input for targeted cooling actions. The results highlight the benefits of a coupled monitoring-simulation-control architecture, which eliminates process errors early, enhances thermal stability and control, and optimizes process efficiency.
• SOLUTION APPROACHES FOR THE EFFICIENT MODELLING OF THE LAYER BUILD-UP IN THE WAAM MANUFACTURING PROCESS USING SMOOTHED PARTICLE HYDRODYNAMICSMOKROV, O.; WARKENTIN, S.; WESTHOFEN, L.; ANTONISSEN, J.; BENDER, J.; SHARMA, R.; REISGEN, U.; 10.3217/978-3-99161-089-2-014Within Additive Manufacturing (AM) technologies, Wire Arc Additive Manufacturing (WAAM) stands out as a particularly promising process variant. WAAM enables the efficient production of large metal structures and offers significant advantages over traditional processes, especially in the field of repair welding. A key component for optimizing the WAAM process is simulation, particularly in relation to the concept of a virtual factory. A digital twin could significantly accelerate the wire-and-arc-based AM manufacturing process, especially regarding development, optimization, and certification. Developing a precise simulation model is crucial to correctly map a weld's geometric shape. In the presented work, Smoothed Particle Hydrodynamics (SPH) is used to develop a model that simulates multiple weld beads building on each other – an area that is not yet fully established due to complex physical interactions and the enormous computing resources required. The focus of the work is on heat and mass transfer during the WAAM process, aiming to create a model that calculates interactions within a geometry of numerous layer structures while significantly reducing calculation time. The proposed approach fully accounts for mass and heat transfer in the cathode area, while droplet formation and detachment are captured by an equivalent model for more efficient use of computing resources. The concept for accelerated calculation incorporates a coarse mass discretization and the neglect of a direct resolution of the anode area. Accordingly, the model focuses on the processes in the cathode area concerning the weld to be formed. A side effect of this concept are local deviations in the weld geometry arising from surface tension models in the SPH formalism, affecting global component geometry. The potential of the model for predicting the temperature field could be demonstrated in the simulative reproduction of an eccentric pin as an industrial reference product. However, there is still no satisfactory agreement between the calculated geometry and the real component. While geometric agreement requires further calibration, the temperature field prediction shows strong consistency with the industrial process. Calculation time still holds potential for improvement. This work makes a significant contribution to the further development of simulation methods in the field of WAAM and provides valuable information for improving process control and increasing efficiency.
• NUMERICAL ANALYSIS OF MECHANICAL BEHAVIOR DURING FRICTION STIR WELDINGMAEDA, S.; IKUSHIMA, K.; SHIBAHARA, M.; 10.3217/978-3-99161-089-2-015A novel numerical analysis method based on a Coupled Eulerian–Lagrangian (CEL) approach with an implicit formulation was developed to simulate the transient thermal and mechanical behavior during Friction Stir Welding (FSW). The proposed method combines an operator-split-based CEL scheme with a dynamic implicit solver optimized for GPU-based parallel computation, enabling stable and accurate simulations with over 1.2 million nodes. This approach permits larger time steps than conventional explicit methods without sacrificing accuracy, achieving a computational cost of about 5.5 hours for each simulated second of welding. The simulation results successfully reproduced key features observed in experiments, including periodic temperature and load fluctuations associated with tool rotation, which are characteristic of onion ring formation. Furthermore, by varying the tool tilt angle, the influence of tool orientation on thermal fields, material flow, and defect generation was quantitatively investigated. A forward tool tilt was found to move the high-temperature region to the front of the tool, leading to the formation of tunnel defects. This provided insight into the defect formation mechanism by revealing that a forward tilt alters material flow and thermal concentration in a way that promotes tunnel defects. These findings demonstrate the effectiveness of the proposed method in capturing complex thermo-mechanical phenomena during FSW and its potential for optimizing welding parameters and improving joint quality.
• NUMERICAL ANALYSIS OF ULTRASONIC VIBRATION ENHANCED FRICTION STIR WELDING OF TI/AL DISSIMILAR ALLOYSZHANG, X.; SHI, L.; WU, C.; 10.3217/978-3-99161-089-2-016Ti/Al hybrid structures offer advantages in lightweight design and cost reduction. However, joining them remains challenging due to the distinct thermophysical properties of the two materials. Conventional friction stir welding (FSW) often results in leads to insufficient material flow capacity when applied to Al/Ti joints. To address this issue, a novel ultrasonic vibration enhanced friction stir welding (UVeFSW) technique was employed to join 2024-T4 aluminum alloy and TC4 titanium alloy. Furthermore, a multi-phase numerical model was developed using computational fluid dynamics (CFD) coupled with the volume of fluid (VOF) method to quantitatively analyze the heat and mass transfer behavior during the Ti/Al dissimilar UVeFSW process. The results reveal that ultrasonic vibration effectively reduces material flow stress and enlarges the plastic flow zone. Overall, the findings demonstrate that UVeFSW is a promising and efficient technique for producing high-quality Ti/Al dissimilar joints, offering a novel approach for advanced lightweight structural applications.
• MULTIPHYSICS SIMULATION OF RESISTANCE SPOT WELDINGBEN BAHAFFA, I.; COURTOIS, M.; CADIOU, S.; GESLAIN, E.; COURTOIS, E.; DUPUY, T.; CHEN, X.; 10.3217/978-3-99161-089-2-017Resistance spot welding is widely used in the automotive industry for assembling steel car bodies. However, increasing demands for safety, durability, and weight reduction have led to the development of more advanced steels, which are also more sensitive to welding defects. To better understand issues like material expulsion, this study relies on an innovative multiphysics finite element numerical simulation. The developed model integrates all the main physical phenomena involved in the process: heat transfer, fluid flow, electromagnetism, and solid mechanics. To manage this complexity, a multi-mesh approach is used, allowing each physics to be solved on a dedicated mesh while maintaining efficient coupling at each time step. The results of this new method will be compared with a more classical Electro-Thermo-Mechanical model, which is the industry standard to highlight the advantages of our approach for studying interfacial expulsion.
• THERMOHYDRODYNAMIC MODELING OF LASER CUTTING FOR RESIDUAL LASER ENERGY QUANTIFICATION USING A HYBRID LEVEL SET – MOVING MESH METHODMEILLOUR, R.; COURTOIS, M.; NAHED, C.; DOYEN, I.; CARIN, M.; 10.3217/978-3-99161-089-2-018Although widely used and investigated in manufacturing, the laser cutting process remains an emerging technology for nuclear dismantling, where the challenges differ significantly. Indeed, while cutting quality is essential in manufacturing processes, it is not a relevant criterion in nuclear dismantling. Nevertheless, the residual laser energy, which is the focus of this study, is of particular relevance to nuclear safety issues for dismantling. This paper presents a thermohydrodynamic model for simulating the laser cutting process of 5 𝑚𝑚 thick 304L stainless steel, accounting for heat transfer and fluid flow in all phases (gas, liquid and solid), and employing the Level Set method to dynamically capture the gas-metal interface. The model is first validated through a comparison of the cutting front angles with experimental observations, demonstrating good agreement with a difference of less than 6%. The simulation results show good agreement with experimental observations regarding the cutting front angle, with a difference of less than 6%. Then, a hybrid Level Set – moving mesh method is proposed to discretize the interface in order to quantify the residual laser energy using the ray tracing approach.
• MODELING AND SIMULATION OF THE ANISOTROPIC THERMAL EXPANSION OF TI-6AL-4V PROCESSED BY PBF-LB/MMAYER, T.; CAPOZZI, P.; HOFMANN, M.; FRISO, F.; RADIS, R.; 10.3217/978-3-99161-089-2-019Additive manufacturing (AM), particularly Laser Powder Bed Fusion of Metals (PBF-LB/M), enables the production of geometrically complex Ti-6Al-4V components with high precision and design flexibility for a wide range of applications in the aerospace, chemical and process engineering, energy, and biomedical sectors. However, rapid solidification in PBF-LB/M produces a non-equilibrium martensitic α′ microstructure with pronounced texture, leading to anisotropic mechanical and thermal behaviour. In particular, the apparent thermal expansion of PBF-LB/M Ti-6Al-4V has been found to be anisotropic due to strains induced during the 𝛼′→𝛽 phase transformation (850–1000 °C). These transformation strains lead to pronounced apparent expansion along the build-direction and reduced in-plane apparent expansion. In PBF-LB/M manufactured components, this anisotropy can lead to unexpected residual stresses during the AM process or mechanical integrity issues during post-manufacturing heat-treatments or operation. This study focusses on the modelling of the anisotropic apparent expansion of Ti-6Al-4V and its implementation for Finite Element Analysis (FEA) to enable simulative investigations of the residual stress and distortion evolution in AM components. A new model is introduced that separates isotropic thermal and transversally isotropic transformation strains. The model accounts for the path and temperature dependent transformation characteristics with varying transformation strains developing during heating (𝛼′→𝛽) and cooling (𝛽→𝛼+𝛽). To make this model available to FEA, it was implemented in the FE software Abaqus via an UEXPAN subroutine. The effectiveness of the apparent expansion model and its implementation is demonstrated by comparing simulated and experimentally measured dilatometer curves. In addition, the impact of this anisotropic expansion on the residual stress development is investigated using thermo-mechanical process simulations of 1) a single laser welding track, and 2) the distortion of a cantilever following support structure removal. The model successfully predicted the apparent expansion in a cylinder printed at a 45° angle. Furthermore, significant differences in vertical stress components and induced plastic strain in a single weld track were found. These are attributed to the additional constraints imposed by the transversally isotropic expansion underpinning the relevance of accounting for the transformation strain. Predictions of distortions in a cantilever following support removal were, however, significantly better when considering isotropic thermal strain only. This could potentially be explained by differences of the PBF-LB/M microstructure in the presence of a support structure. Understanding these mechanisms is crucial for optimizing the mechanical integrity and dimensional stability of PBF-LB/M-fabricated Ti-6Al-4V components in engineering applications subjected to post-processing heat-treatments or high-temperature applications.
• DEVELOPING A MULTI-PHASE WELDING SIMULATION WITH DAMAGE PREDICTION SOURCING ON STANDARDIZED TENSILE TESTS AND EXPERIMENTAL VALIDATION BY DAMAGING A WELDMENTSCHRÖDER, C.; LOOSE, T.; BRODT, M.; 10.3217/978-3-99161-089-2-020To accurately simulate the strength and potential damage of welded joints in structural steel, a multi-phase welding simulation is required. This is because phase transformations in the heat-affected zone (HAZ) change the local mechanical properties significantly. The main microstructural phases of steel (ferrite/pearlite, bainite, and martensite) exhibit distinct mechanical properties such as tensile strength and fracture elongation. For the prediction of damage in welding seams, these properties play a critical role and must be considered. Welding seams can be simulated with computational welding mechanics (CWM) by calculating stresses, strains, distortions, and microstructural transformations in the HAZ. For the microstructure transformations, the effects on structural properties are considered using multi-phase material models [1], [2]. Such a multi-phase material is developed and implemented in the commercial software LS-DYNA by [3], which is available as keyword *MAT_GENERALIZED_PHASE_CHANGE (*MAT_254). The model was specifically developed for process chain simulations using a consistent material description [4] [5]. The presented work focuses on the numerical simulation of damage evolution in welded joints, under consideration of local phase changes and their underlying properties. The aim was to keep the same material model from welding to loading while considering the effects of local phase transformations for the prediction of damage. To do so, the damage model *MAT_ADD_GENERALIZED_DAMAGE, which is a generalized form of *MAT_ADD_DAMAGE_GISSMO developed by [6], was adopted. Three optimized sets of parameters of GISSMO, one for each of the three main phases, were identified based on numerical simulations and experimental results using LS-OPT. The approach proposed in [7] reassigns elements to parts characterized by the same mechanical properties after welding. This technique of creating local material clusters was adopted to perform the damage simulation based on the respective present phase of the steel. The three unique sets of parameters were then combined into a multi-phase damage model, which uses a mixing rule to calculate the material properties for arbitrary phase combinations. For the final damage calculation to verify the correct damage prediction of a welded segment in a three-point bending test, it was realized that the damage model does not work properly in combination with *MAT_254. The predicted cracks occurred in an unrealistic manner. Further research must be carried out to determine the exact cause and to fix this problem.
• FRACTURE MECHANICS BEHAVIOR OF HSS STEEL WELDED JOINT HETEROGENEOUS STRUCTURE - EXPERIMENTAL AND NUMERICAL EVALUATIONTOMERLIN, D.; GUBELJAK, N.; KOZAK, D.; LI, W.; TRIŠOVIĆ, N.; 10.3217/978-3-99161-089-2-021Due to the physics and chemistry behind the fusion welding process, the welded joints are consequently significantly heterogeneous structures. They can be fundamentally divided into characteristic regions of Base Metal (BM), Weld Metal (WM) and Heat Affected Zone (HAZ). Such heterogeneity of the welded joint material microstructures in relation affects their general mechanical properties, which can vary significantly in relation to individual weld zones. The fracture behavior of welded joints consequently depends on the mechanical properties of the material zone in which the crack is located. The paper considers double-V butt joint (X joint) made of High Strength Steel (HSS), welded with filler metal having slightly overmatching mechanical properties as compared to base metal. The experimental work relies on the initial tensile testing in order to obtain stress-strain material properties, covering all the welded joint material regions of interest. In the continuation, the fracture mechanical testing is done to determine the fracture toughness characteristic parameters, as well as all the characteristic fracture behavior curves. The experimental investigation is carried out on Single Edge Notch Bend (SENB) specimens, located in welded joint zones of interest, namely BM, HAZ, and WM. The prediction of fracture behavior of metallic materials, through the initiation and evolution of damage, can be numerically simulated using the Ductile Damage material model. Such a model in general relies on stress-strain curves obtained from tensile tests of individual material zones within the observed heterogeneous structure. In this paper, using the example of SENB test specimens, the general methodology for modeling and simulating material heterogeneity of welded joints is presented. Finally, experimentally determined fracture mechanics curves and behavior of characteristic welded joint regions are simulated numerically for HSS steel grade. In this way, the possibility of numerically predicting the fracture behavior of heterogeneous structures is demonstrated.
• DEVELOPMENT OF RESIDUAL STRESS MODEL FOR WELDED BOX SECTIONS USING PROBABILISTIC APPROACHHORVÁTH, A.; KOLLÁR, D.; 10.3217/978-3-99161-089-2-022Manufacturing processes significantly contribute to the formation of imperfections such as residual stresses and distortions in components. This study focuses on residual stresses induced by welding in box sections made from a combination of normal strength steel (NSS) and high strength steel (HSS) plates. Advanced manufacturing simulations are carried out using a deterministic, uncoupled transient thermomechanical analysis incorporating a double-ellipsoidal heat source model. These simulations evaluate the residual stress distribution in weldments across various steel grades, geometries, and welding parameters (e.g., heat input, welding speed). Furthermore, the study presents a methodology for probabilistic welding simulation. Longitudinal residual stresses are analysed using Monte Carlo simulations with Latin hypercube sampling. The findings support the development of a comprehensive residual stress model that integrates key welding and fabrication characteristics, reflecting the specificities of manufacturing technologies.
• HIGH-TEMPERATURE STRAIN GAUGE MEASUREMENTS IN WA-DED: VALIDATION OF NUMERICAL MODELSDREXLER, H.; MOSCHINGER, M.; ENZINGER, N.; 10.3217/978-3-99161-089-2-023In this study, high-temperature strain gauges are used to monitor the in-situ stress during plasma Wire Arc Directed Energy Deposition (WA-DED). A computational temperature compensation is applied to correct the strain signals for thermal effects, enabling accurate stress measurements at high temperatures. The experimental results are compared against a fully coupled thermo-mechanical finite element simulation. The comparison shows good agreement between measured and simulated stresses, demonstrating that high-temperature strain gauges can reliably capture the stress evolution during WA-DED. This approach goes beyond conventional post-process residual stress measurements by providing time-resolved validation data. The findings underline the potential of in-situ high-temperature strain gauge measurements as a robust tool for validating WA-DED simulations.
• HEAT TRANSFER AND FLUID FLOW MODEL ASSISTED MECHANICAL ANALYSIS FOR WIRE ARC DIRECTED ENERGY DEPOSITION TO IMPROVE RESIDUAL STRESS AND DISTORTION PREDICTIONDILJITH, P. K.; ARYAN, A.; ZHANG, W.; DE, A.; 10.3217/978-3-99161-089-2-024A substantive design of process variables and scanning strategy in wire arc directed energy deposition (DED-Arc) requires an estimation of the thermal distortion and residual stress. Routine sequential heat transfer and thermomechanical stress analyses neglect the convective heat transport in the melt pool that commonly affects the computational accuracy. We propose here to compute the deposit profile and temperature field using a finite volume-based heat transfer and fluid flow model and use the computed temperature field for a subsequent mechanical analysis. The computed melt pool profile, deposition profile and residual stress showed a maximum discrepancy of around 10% against experimentally measured results for single- and multi-track DED-Arc depositions. An assessment of the increased computational demand vis-à-vis enhanced prediction accuracy due to the integration of a heat transfer and fluid flow model over a heat conduction only model is presented. Further efforts to simulate multi-track multi-layer DED-Arc using the proposed approach are underway.
• PREDICTION OF RESIDUAL STRESSES OF REPAIRED AUSTENITIC STAINLESS STEEL WELDED JOINTSHILAL, S.; HENDILI, S.; DELMAS, J.; PEREIRA ALVAREZ, P.; ROBIN, V.; DERNIAUX, E.; BOUTIN, T.; 10.3217/978-3-99161-089-2-025The ability to predict residual stresses in welds is important in assessing the integrity of a component subject to degradation mechanisms such as Stress Corrosion Cracking (SCC). In the as-welded condition, multi-pass pipe circumferential butt welds made of austenitic stainless-steel exhibit a typical residual stress profile through the wall of the welded joint. In the case of a non-repaired weld, the shell bending profile places the inner half of the weld joint wall and its vicinity in compression in the axial direction. In contrast, the distribution of residual stresses may be different in the event of a repair, particularly in the case of localised repairs. The impact of repairs on SCC crack propagation was suspected following an inspection of a weld with deep SCC cracks. The particularity of this weld is that it was subjected to two successive repair stages: a complete circumferential repair followed by a localised repair. The specific character of the weld required 3D modelling, since it provides a description of each circumferential zone of the weld and enables verification of the way in which this weld has a particular behaviour (residual stresses, hardness, etc.). This would make it possible to identify potential aggravating factors regarding SCC for this type of weld configuration. This work presents the simulation of this weld, which was carried out in three stages: the initial weld, the first repair, and the localised repair. The expertise carried out following the cutting of the weld made it possible to characterise the strain hardening in the repaired areas and is compared with the simulation results by considering the measurements obtained on a sample taken in the middle of the repair. This comparison is made on hardness profiles in the axial direction of the assembly, at different heights in the thickness. A good correlation is observed between measured and simulated hardness.
• RESIDUAL STRESS SIMULATION OF ALUMINUM CASTINGS USING A TEMPERATURE-DEPENDENT HEAT TRANSFER COEFFICIENT DURING IMMERSION QUENCHINGHELML, L.; MAIER, S.; WLANIS, T.; ARTNER, W.; KAHLENBERG, R.; FUCHS, G.; KOZESCHNIK, E.; 10.3217/978-3-99161-089-2-026A simulation model is developed to predict residual stresses in aluminum castings during immersion quenching. To describe the cooling conditions, a simplified approach is used, in which the heat transfer coefficient is defined as a function of the surface temperature of the component. Two geometries are used in this study: a simple plate to verify the thermal boundary conditions and a stress lattice to evaluate the residual stress distribution. The heat transfer coefficient, which is experimentally validated using the plate geometry, is implemented in a finite element simulation and then applied to the stress lattice. The mechanical behaviour during quenching is characterized using dilatometer experiments. The simulated stress profile in the stress lattice is verified by X-ray diffraction measurements.
• SIMULATION OF STANDARDIZED WELDABILITY TESTS AND LINKAGE WITH FRACTURE AND FATIGUE BEHAVIORFU, H.; NIED, H. F.; 10.3217/978-3-99161-089-2-027This study examines the thermo-mechanical strains and stresses experienced by test specimens during weldability qualification tests, with a specific focus on the popular Varestraint and Houldcroft tests. These weldability tests are commonly used to identify critical thresholds for the initiation of hot cracking during the welding process, as well as subsequent crack propagation behavior in the presence of weld-induced residual stresses. The computational analysis in this study is primarily focused on the implementation of finite element models developed using ANSYS and SYSWELD software to simulate the coupled thermo-mechanical behavior during weldability tests. A notable aspect of the Varestraint weldability test is the application of a sudden external load during the welding process, introducing an additional external mechanical strain component in the test specimen. During this loading, the metal in the cooling weld pool is highly susceptible to fracture and is most likely to exhibit "hot cracking" behavior. To accurately determine the localized strains associated with this phenomenon, the finite element models used in this study employ element “birth” and “death” techniques, enabling precise tracking of strain and stress variations over time. This study attempts to accurately quantify the critical threshold strains for hot cracking initiation based on modified Brittleness Temperature Range (BTR) curves. Simulations of room-temperature fatigue crack propagation under the influence of combined welding residual stresses and external mechanical loads are also included. This is enabled by incorporating R-ratio crack growth rate effects in the simulations. Ultimately these simulations provide a more detailed and comprehensive understanding of the failure mechanisms in welded joints. An interesting fatigue crack growth example presented in this study is the numerical simulation of fatigue crack propagation in welded structures subjected to purely compressive external loading (compressive fatigue).
• EXPERIMENTAL AND NUMERICAL STUDY ON DUCTILITY CAPACITY FOR SOLIDIFICATION CRACKING IN LASER BEAM WELDINGHABIBI, N.; GUMENYUK, A.; RETHMEIER, M.; 10.3217/978-3-99161-089-2-028Solidification cracking remains a critical challenge in laser beam welding, influenced by thermal, metallurgical, and mechanical factors. This study investigates the local conditions that contribute to this phenomenon in stainless steel 304 by measuring the temperature and strain evolution in the critical region. A custom measurement system was developed to estimate local temperature, strains, and strain rates, during controlled tensile weldability (CTW) experimental tests. This method enables a direct assessment of crack onset by applying a defined strain and strain rate during welding. To complement experimental findings, a parallel numerical study was conducted using a three-dimensional finite element (FE) model in ANSYS to trace the strain and stress development. The model incorporated sub-modeling of the critical area using predefined boundary conditions. This approach allows for the optimization of mesh size and time step to minimize numerical artifacts. Numerical results showed good agreement with experimental, in terms of level and distribution of temperature and strain, confirming the model’s reliability. An iterative simulation approach was used to determine the critical strain conditions leading to crack initiation. These results were constantly compared with the experimentally observed data. This integrated method sheds light on the evolution of ductility and local straining conditions, enabling a more precise assessment of solidification cracking susceptibility. Corresponding author: andrey.gumenyuk@bam.de.
• MACHINE LEARNING-ACCELERATED CALIBRATION OF COMPLEX HEAT SOURCE MODELS FOR WELDING AND ADDITIVE MANUFACTURING PROCESS SIMULATIONLUO, Y.; KUDVA, A.; JANG, E.; MATAN, S.; VEGA MICHALAK, N. M.; PAULSON, J. A.; PERRAULT, A.; ALEXANDROV, B.; 10.3217/978-3-99161-089-2-029The heat source (HS) models dictate the spatial distribution of applied heat during welding and additive man-ufacturing processes, making it a critical factor for accurate simulation. However, the problem of calibrating the HS model, i.e., selecting parameters for it that match experimental conditions, has often been overlooked, with parameter selection relying on empirical heuristics and human trial-and-error. This lack of systematic calibration can lead to poor simulation performance and thus limits the expressiveness of HS models that can be used. We present a machine learning framework for efficiently calibrating HS parameters, thereby facili-tating use of HS models with greater expressiveness and practical applicability than the conventional double-ellipsoidal Gaussian model. The effectiveness of the proposed methodology is validated through case studies involving spray transfer gas metal arc welding (GMAW) and pulsed GMAW additive manufacturing. Exper-imental validation, including comparisons of thermal histories and heat-affected zone properties, confirms strong alignment between simulation and real-world observations. This work provides a systematic frame-work for improving welding simulation accuracy by addressing the long-standing limitations of empirical HS models and facilitating the adoption of more expressive and precise modeling approaches. Ultimately, we conclude that (1) machine-learning optimization-based calibration of HS parameters significantly enhances simulation accuracy compared to using HS models that are calibrated using standard methods; (2) using our framework, expressive HS models that can represent complex energy distribution better align with experi-mental data compared to the standard double-ellipsoid Gaussian model.
• WA-DED SPECIFIC TOPOLOGY OPTIMIZATION OF A KINGPIN PLATE FOR SEMI-TRAILERSHAUNREITER, F.; SILMBROTH, M.; BHARADWAJ, K.; DREXLER, H.; SCHWENDINGER, M.; BRUHNS, H.; BAUER, A.; 10.3217/978-3-99161-089-2-030Additive manufacturing (AM) offers significant design freedom compared to conventional manufacturing methods. This enables the production of complex, topology-optimized metal components. While powder bed processes are well-established for small to medium-sized parts, large-scale applications benefit from directed energy deposition (DED) techniques such as wire-arc directed energy deposition (wa-DED). Despite its potential, limited research addresses topology optimization specifically tailored to the constraints of wa-DED. This study presents a workflow for wa-DED-compatible topology optimization, using a semi-trailer kingpin plate as a demonstrator geometry. All simulations were conducted using Altair OptiStruct. Multiple optimization scenarios were evaluated to compare the impact of different optimization approaches and manufacturing constraints. The most suitable design achieved a 59 % weight reduction compared to the initial design using minimized mass fraction with stress, deformation and draw type constraints. A nonlinear finite element (FE) analysis confirmed the structural integrity of the optimized geometry under relevant load cases. To validate manufacturability, a quarter section of the design was successfully fabricated using wa-DED.
• DIFFERENT APPROACHES TO VISUALIZING SIMULATIONS OF WELDING PROCESSESKESSELBURG, L.; WARKENTIN, S.; MOKROV, O.; SHARMA, R.; REISGEN, U.; 10.3217/978-3-99161-089-2-031In modern industry and research, vast amounts of data are being generated continuously, both through real-world measurements and numerical simulations. In the case of numerical simulations, particularly transient simulations, this creates significant challenges for human comprehension. Consequently, there is a critical need to create expressive visualizations while keeping computational resource requirements manageable. These visualizations can range from simple images to videos, and ultimately to interactive renderings, depending on the specific use case. Furthermore, a more established simulation system can offer a highly integrated interface, which allows the user to parameterize and schedule new simulations in addition to the interactive visualization of running and archived simulations. The requirements for effective visualization change depending on the end-user’s needs. On one end of the spectrum, developers and researchers involved in the data generation process require full access to all raw data, often coupled with sizable computational resources. This access facilitates early validation, error tracing, exploratory analysis, and further enhancement of the visualizations themselves. However, this configuration often necessitates a more intricate setup process. On the opposite end, reviewers, researchers, or executives generally require straightforward access to a refined version of the data, accompanied by an optimized and stable visualization pipeline. This enables tasks such as reviewing scientific publications or making data-driven decisions in a business context. In this paper, we explore various visualization strategies tailored to meet the diverse needs of these two user groups, with a particular focus on simulations of welding processes. By demonstrating these strategies, we illustrate the trade-offs between computational resource demands and usability. We show two variations of the Server-Client-Architecture. In the first case, the client is a desktop program, presenting the user with a wide variety of options for analyzing and visualizing data, while utilizing computational resources from the client or the server. In the second case, the client is a web-browser, with the main benefit of removing any setup required. Additionally, we want to show an approach that utilizes WebAssembly to execute the full visualization pipeline in a web-browser. This includes loading the data, filtering, mapping and rendering. While this approach is more constrained in terms of data volume and processing power, it aims to enhance exchange between institutions through its ease of use and ease of deployment. To further emphasize the interchangeability of simulation results, the open vtk file format is used, as it is widely utilized in academia. This allows us to leverage several KitWare software programs and libraries, thereby streamlining the development process by providing advanced visualization tools, reducing the necessity to work directly with raw visualization primitives, thus enabling the reuse of code across different strategies.
• A SIMPLIFIED HYBRID METHODOLOGY FOR DETERMINING WELDING HEAT SOURCE PARAMETERS IN HIGH-STRENGTH STEELSDADKHAH, M.; NITSCHKE-PAGEL, T.; DILGER, K.; 10.3217/978-3-99161-089-2-032Welding simulations are critical for predicting the thermal and mechanical behavior of high-strength steels, as accurate thermal analysis, particularly under varying heat inputs, directly influences microstructure evolution and mechanical performance. Traditional heat source models rely on complex parameterization, often requiring variables that are computationally intensive or absent from standard material databases. This study addresses these limitations by introducing a simplified, hybrid methodology to determine heat source parameters for the Goldak’s double ellipsoidal heat source model. The approach integrates various elements, including the molten pool's length components captured through high-speed videography (HSV), size of phase transformation zones, local temperature measurements combining with Design of Experiments (DOE), and finite element simulations. This integration aims to generate optimized parameters with enhanced accuracy and efficiency. Three acceptance criteria were used to validate the optimized parameters derived from DOE: identical temperature gradients at thermocouple locations, equivalent macro-diagrams, and consistent cooling rates. These criteria ensured alignment between thermal analysis phase of the welding simulations and experimental data, facilitating robust calibration of heat source geometries and heat inputs. This framework significantly reduced computational costs compared to traditional trial-and-error calibration methods. The results illustrate the capability of this method in predicting temperature distributions and related residual stresses afterwards. Furthermore, the adaptability of this methodology is a key strength, allowing it to be extended to other heat source models or different welding conditions. This versatility enhances its utility across a wide range of applications in welding research and industry, while its flexibility ensures that the methodology remains relevant as new materials and welding technologies emerge.
• A REVIEW OF WELDING MODELLING APPROACHES FOR NUCLEAR INDUSTRYBROSSE, A.; KHELIF, N.; LEVEILLE, T.; GALLÉE, S.; 10.3217/978-3-99161-089-2-033In the nuclear industry, welding process is widely used to assemble components. As a manufacturer, Framatome has developed over the years specific tools and approaches to answer the needs of our welders. Indeed, for an industrial use, welding modelling must be efficient, robust and reliable but unfortunately no approach can achieve all these goals, so choices must be made. Many approaches exist in the literature to model residual stresses and distortions. This paper will focus only on thermomechanical approaches of welding by finite element analysis such as: moving heat source, thermal cycle, inherent strain, .... Each of these approaches has advantages and disadvantages such as computation time or accuracy of results. In this paper we propose to make a comparison between these approaches and to define the assumptions and limits for each of them in the context of welding and Wire Arc Additive Manufacturing modelling for nuclear industry. First, the paper will present a review of each approach capabilities and then their applications on a specific mockup designed as an Additive Manufacturing wall are described. The distortions and residual stresses results are compared for each approach with experimental data with the goal to give limits and guidelines for each application.
• EVALUATION OF FEM-BASED MACRO-SCALE AND MICRO-SCALE THERMAL SIMULATION FOR EFFICIENT PROCESS ADAPTATION IN DIRECTED ENERGY DEPOSITION ADDITIVE MANUFACTURINGPAKDEL SEFIDI, A. M.; MASHETTY, V.; GHORBEL, M. BEN; OSSENBRINK, R.; SCHRICKER, K.; 10.3217/978-3-99161-089-2-034Directed Energy Deposition (DED) involves complex thermal interactions that significantly influence microstructural properties, distortion and other final part properties. Accurate thermal modeling is essential to predict these effects and to further optimize processes. This study is motivated by the need for reliable simulation models to enable process optimization, in particular to adjust process parameters based on evolving thermal conditions. A computational framework requires efficient thermal predictions to enable adaptive control strategies, that optimize deposition parameters. This study presents a comparative analysis between two thermal modeling approaches in DED simulations: the macro-scale element clustering approach, as implemented in ANSYS DED Toolbox, and a micro-scale modeling approach based on high-resolution transient thermal simulations using the finite element method. The macro-scale simulation uses a computationally efficient approach where material is deposited in clusters and heat is applied using either a power-based or temperature-based method. While this method allows for reduced computational costs, it may lack fine-scale resolution of localized thermal gradients. On the other hand, the micro-scale FEM approach uses a mathematically defined heat source model with detailed mesh refinement and element activation, allowing for higher accuracy in capturing transient thermal effects but at the cost of increased computational requirements. While micro-scale FEM models provide accurate transient thermal analysis, their computational cost could limit their applicability for control and optimization tasks such as predictive path planning or optimized microstructures. The macro-scale simulation model offers a computationally efficient alternative, but its ability to capture relevant effects should be systematically evaluated. In this respect, the results provide a detailed insight into the relationship between computational efficiency, predictive accuracy and model applicability.
• PIPEWELD, A SOFTWARE SUITE FOR COMPUTATIONAL WELDING MECHANICS FOR NUCLEAR APPLICATIONSDELMAS, J.; HENDILI, S.; HILAL, S.; PEREIRA ALVAREZ, P.; ROBIN, V.; 10.3217/978-3-99161-089-2-035In the nuclear industry, numerical modelling of welding processes has become a key tool for understanding and decision-making. It is used to evaluate the impact of welding on degradation mechanisms and to accelerate the development and qualification of welding and repair techniques. Computational welding mechanics, which involves modelling the solid-state effects of welding on the base material and weld metal, such as temperature fields, microstructural evolution, and stress-strain distributions, serves this purpose. Since 2021, EDF has observed stress corrosion cracking (SCC) in certain austenitic stainless-steel pipe-to-elbow welds within safety injection systems. Accurate prediction of welding residual stresses (WRS) is essential for assessing the structural integrity of components exposed to SCC and similar degradation mechanisms. During welding, coupled thermo-mechanical and metallurgical phenomena occur, including microstructural transformations such as dynamic recovery and recrystallization. These transformations influence the material’s hardening behaviour and must be accounted for in residual stress simulations. In the as-welded condition, multi-pass circumferential butt welds in austenitic stainless steel typically exhibit a residual stress profile characterized by a shell bending effect. This results in compressive axial stresses in the inner half of the weld wall, although tensile zones may still develop near the weld axis on the inner surface. This behaviour is well-documented in the literature and has been validated through simulations using code-aster (a general-purpose finite element software for solid mechanics) and confirmed by experimental measurements on mock-ups. In this context, the Pipeweld software suite has consolidated over two decades of research on welding and its numerical modelling. Built on code-aster and Salome_Meca, numerical software developed by Electricité de France, Pipeweld enables the prediction of residual stress states for component lifetime assessment and process qualification optimisation.
• AN AUTOMATIC ANISOTROPIC MESHING ALGORITHM ADAPTED TO THE MULTIPHYSICS SIMULATION OF A TUNGSTEN INERT GAS WELDING ARCGOUNAND, S.; 10.3217/978-3-99161-089-2-036In the context of the multiphysics numerical simulation of arc welding using finite elements, meshing difficulties are often encountered. Indeed, the geometry of welding tools and of the parts to be joined can be quite intricate. Also, the multiphysics model can introduce more constraints on the mesh such as maximal layer thickness for cathodic and anodic layers. Fluid mechanics may also trigger boundary layers in the physical solution. Moreover, as multiphysics frequently involve many physical unknowns per mesh node, it is important to keep the number of mesh elements under control to ensure feasible computations. This contribution presents a meshing algorithm based on the construction of a tensor field of the desired mesh size by metric interpolation, which is then fed into an automatic anisotropic mesher and a multiphysics model of the Tungsten Inert Gas (TIG) welding arc.
• LARGE-SCALE NUMERICAL ANALYSIS OF MULTI-PASS WELDING USING AUTOMESH AND IDEALIZED EXPLICIT FEMWANG, W.; SASAKI, S.; MAEDA, S.; IKUSHIMA, K.; SHIBAHARA, M.; 10.3217/978-3-99161-089-2-037This study introduces a numerical framework that combines the highly efficient automatic mesh generation system, Automesh, with the Idealised Explicit Finite-Element Method (IEFEM) to enable large-scale thermo-elastic–plastic analyses of multi-pass welding. Automesh converts engineering parameters into graded, analysis-ready meshes while enforcing equal unit heat-input density per pass and consistent element quality and pass placement across cases, thereby ensuring fair comparisons. IEFEM employs a dynamic-explicit formulation with local (node/element-level) updates, substantially reducing memory usage and runtime and admitting efficient GPU parallelisation. Two parametric campaigns on X-groove and V-groove joints deliver pass-resolved, time-accurate fields of temperature, stress, and strain. Within the investigated ranges, groove depths near half the plate thickness minimised deformation; decreasing groove angle reduced accumulated lineal heat input and thus deformation; and root gap showed little effect for 𝑔=1 ~ 10 mm. Discrete jumps in angular distortion were traced to changes in the number of passes within a dominant height band, which placed sub-passes near the peak of an empirical 𝑄/ℎmax2 response. Automesh reduced meshing from hours to under 10 seconds per case, and IEFEM cut solution time by roughly an order of magnitude versus representative implicit baselines, yielding an efficient and reliable workflow for industrial studies and database construction.