Kapitel

• High-fidelity numerical modelling of cold spatter formation during laser powder bed fusion of 316-L stainless steelBayat, M.; Hattel, J. H.; 10.3217/978-3-85125-968-1-01Spatter and denudation are two very well-known phenomena occurring mainly during the laser powder bed fusion process and are defined as ejection and displacement of powder particles, respectively. The main driver of this phenomenon is the formation of a vapor plume jet that is caused by the vaporization of the melt pool which is subjected to the laser beam. In this work, a 3-dimensional transient turbulent computational fluid dynamics model coupled with a discrete element model is developed in the finite volume-based commercial software package Flow-3D AM to simulate the spatter phenomenon. The numerical results show that a localized low-pressure zone forms at the bottom side of the plume jet and this leads to a pseudo-Bernoulli effect that drags nearby powder particles into the area of influence of the vapor plume jet. As a result, the vapor plume acts like a momentum sink and therefore all nearby particles point are dragged towards this region. Furthermore, it is noted that due to the jet’s attenuation, powder particles start diverging from the central core region of the vapor plume as they move vertically upwards. It is moreover observed that only particles which are in the very central core region of the plume jet get sufficiently accelerated to depart the computational domain, while the rest of the dragged particles, especially those which undergo an early divergence from the jet axis, get stalled pretty fast as they come in contact with the resting fluid. In the last part of the work, two simulations with two different scanning speeds are carried out, where it is clearly observed that the angle between the departing powder particles and the vertical axis of the plume jet increases with increasing scanning speed.
• Comparison between green and infrared laser in laser powder bed fusion of pure copper through high fidelity numerical modelling at meso-scaleAlphonso, W.E.; Bayat, M.; Hattel, J. H.; 10.3217/978-3-85125-968-1-02Laser Powder Bed Fusion (L-PBF) is a Metal Additive Manufacturing (MAM) technology which offers several advantages to industries such as part design freedom, consolidation of assemblies, part customization and low tooling cost over conventional manufacturing processes. Electric coils and thermal management devices are generally manufactured from pure copper due to its high electrical and thermal conductivity properties. Therefore, if L-PBF of pure copper is feasible, geometrically optimized heat sinks and free-form electromagnetic coils can be manufactured. However, producing dense pure copper parts by L-PBF is difficult due to low optical absorptivity to infrared radiation and high thermal conductivity. To produce dense copper parts in a conventional L-PBF system either the power of the infrared laser must be increased above 500W, or a green laser should be used for which copper has a high optical absorptivity. Increasing the infrared laser power can damage the optical components of the laser systems due to back reflections and create instabilities in the process due to thermal-optical phenomenon of the lenses. In this work, a multi-physics meso-scale numerical model based on Finite Volume Method (FVM) is developed in Flow-3D to investigate the physical phenomena interaction which governs the melt pool dynamics and ultimately the part quality. A green laser heat source and an infrared laser heat source are used individually to create single track deposition on pure copper powder bed above a substrate. The effect of the dissimilar optical absorptivity property of laser heat sources on the melt pool dynamics is explored. To validate the numerical model, experiments were conducted wherein single tracks are deposited on a copper powder bed and the simulated melt pool shape and size are compared. As the green laser has a high optical absorptivity, a conduction and keyhole mode melting is possible while for the infrared laser only keyhole mode melting is possible due to low absorptivity. The variation in melting modes with respect to the laser wavelength has an outcome on thermal gradient and cooling rates which ultimately affect the mechanical, electrical, and thermal properties.
• Simulation of microstructure evolution during WAM processKronsteiner, J.; Drexler, H.; Hovden, S.; Haunreiter, F.; O‘Toole, P.; Molotnikov, A.; Easton, M.; 10.3217/978-3-85125-968-1-03The development of microstructure during Wire-Based Additive Manufacturing (WAM) is of major interest for the Additive Manufacturing (AM) industry. The resulting geometry, mechanical properties, and quality of WAM parts are directly affected by the process conditions. Numerical simulations of WAM processes can predict and optimize the process settings (heat input, thermal conditions, e.g., by pre-heating, …) and therefore speed-up the trial-and-error phase of the manufacturing process (first time right). The WAM process can be considered as a welding process accompanied by continuous heat-treatment processes (during the build-up of walls the heat source is reheating and even partially remelting existing layers multiple times), the formation of the microstructure mainly depends on the temperature evolution in the part. Two modes of microstructure development are considered: grain nucleation and growth during solidification and growth/recrystallization in the solid phase. To compute grain size considering the interplay of nucleation and growth during solidification, the Interdependence (ID) model is used. Based on the initial grain size distribution, chemical composition, inoculant particles, among other influence factors, the evolution of the microstructure and hot cracking susceptibility during the cooling and reheating cycles of an AM process can be calculated. Both models have been implemented into the Finite Element (FE) solver LS-DYNA®. Depending on the current element temperature either of the two grain morphology models are activated. Here, the evolution of the microstructure including the hot cracking susceptibility during the manufacturing of wall structures is presented. The results obtained from the calculations are compared and validated against our experimental trials. A good agreement between the measured and calculated grain size could be obtained.
• Numerical prediction of bead formation and build-up toward WAAM process optimizationBen Hamouda, H. ; Felice, I. O.; Oliveira, J. P.; Antonissen, J.; 10.3217/978-3-85125-968-1-04Wire and Arc Additive Manufacturing (WAAM) is a promising solution to produce complex shapes with low buy-to-fly ratio outperforming conventional subtractive manufacturing methods. Generally, this approach relies on robot programs that define the printing path based on a set of pre-defined process parameters. To obtain a net shape WAAM part, the printing path requires controlled bead shape estimation in the preprocessing step. Unlike trial-and-error approach, numerical simulation tries to determine the bead shape by solving coupled physical problems such as fluid dynamics (CFD) and heat transfer. In this work, a numerical tool was deployed to predict single and multiple bead formation given three key process parameters: the wire feed speed, torch travel speed, and voltage. Calculations were performed using ER-90S steel alloy as a printing wire and compared to experimental measurements printed using a MIG/MAG process. As a result, prediction of dimensional features such as the bead height (BH) and width (BW) showed a good agreement with the experiment. Furthermore, the effect of the process parameters was investigated and a parametric study was performed to establish a process guideline that feeds the robot printing strategy.
• Physical mechanisms governing deposition rate in arc welding with a consumable electrodeMendez, P. F.; Yan, Z.; Sánchez, V. Núñez; Chen, S.; 10.3217/978-3-85125-968-1-05This paper presents a model of deposition rate in gas metal arc welding (GMAW). Some aspects of this model are also helpful to understand related processes such as MCAW, FCAW, SAW, SMAW, EGW, ESW and how-wire additions. Deposition rate is one of the essential factors in calculation of welding costs and times to completion in practical applications. The relationship of deposition rate with current is also determinant of dilution of deposited material (essential concept in overlays) and is a tool of diagnostic of the welding process. For some common materials such as aluminum alloys, the variations in deposition rate with current are not well understood, and common explanations might be misleading. The foundations of the model are mass and energy balances together with the mass and heat transfer mechanisms involved. Heat transfer mechanisms considered include the energy deposited by the fall voltage of the arc against the consumable, Joule heating of the electrode extension, evaporation losses, and heat exchange with the contact tip. The calculation of Joule heating considers the amount and shape of electrical resistivity variation with temperature. The approach presented is in contrast with the common use of company literature for the relationship between current and wire feed speed, or with the use of a second order expression with tabulated constants specific for particular materials and process configurations. The model developed allows to predict the effect of electrode extension and droplet temperature, which are not explicit in current approaches.
• Numerical study on the formation of a bulging region in partial penetration laser beam weldingArtinov, A.; Meng, X.; Bachmann, M.; Rethmeier, M.; 10.3217/978-3-85125-968-1-06A transient three-dimensional thermo-fluid dynamics numerical model was developed to study the formation of a bulging region in partial penetration laser beam welding. The model accounts for the coupling between the fluid flow, the heat transfer, and the keyhole dynamics by considering the effects of multiple reflections and Fresnel absorption of the laser beam in the keyhole, the phase transitions during melting and evaporating, the thermo-capillary convection, the natural convection, and the phase-specific and temperature-dependent material properties up to the evaporation temperature. The validity of the model was backed up by experimentally obtained data, including the drilling time, the weld pool length, the local temperature history outside the weld pool, the process efficiency, and a range of metallographic cross-sections. The model was applied for the cases of partial penetration laser beam welding of 8 mm and 12 mm thick unalloyed steel sheets. The obtained experimental and numerical results reveal that the bulging region forms transiently depending on the penetration depth of the weld, showing a tendency to transition from a slight bulging to a fully developed bulging region between penetration depths of 6 mm and 9 mm, respectively.
• Modelling of the melt pool behaviour during a pulsed TIG welding operation in a narrow grooveCadiou, S.; Baumard, A.; Brosse, A.; Bruyere, V.; 10.3217/978-3-85125-968-1-07Arc welding is one of the main processes for assembling metal components in the nuclear industry. To guarantee the quality of the welded assemblies and to predict the characteristics of the weld, it is necessary to master the welding process and have a thorough understanding of the interactions within the melt pool. To this end, the objective of this work is to develop a transient numerical model allowing for the prediction of the behaviour of the melted zone during current pulsation in reasonable computational times. The relevant industrial application in this study is the welding of a narrow groove gap of a stainless steel pipe. The welding process used is pulsed TIG and different synergies are studied. In this work, numerical simulation is used as a predictive analysis tool providing data that complete the experimental ones. Knowing that the predictive aspect of the simulations depends on the modelling choices, it is necessary to consider the main physical phenomena governing the melt pool (thermal transfers, fluid flow, electromagnetism) and to model the mass feeding process using the Arbitrary Lagrangian Eulerian (ALE) method. The development of the magneto-thermohydraulic model with material supply is carried out using the Comsol Multiphysics® software.
• A numerical study on the suppression of a detrimental weld pool profile in wire feed laser beam welding by magnetohydrodynamic techniqueMeng, X.; Artinov, A.; Bachmann, M.; Üstündağ, Ö.; Gumenyuk, A.; Rethmeier, M.; 10.3217/978-3-85125-968-1-08The weld quality and the possible defect formation are directly determined by the weld pool shape and the thermo-fluid dynamics therein. In this paper, an untypical weld pool profile, i.e., elongated at its top and bottom but narrowed at the middle, is found experimentally and numerically in the wire feed laser beam welding. The detrimental influence of the weld pool narrowing on the element transport is analyzed and discussed. A magnetohydrodynamic technique is utilized to suppress the narrowing, aiming at a more homogenous element distribution. It is found that a low-temperature region is formed in the middle of the weld pool due to the interaction of the two dominant circulations from the top and bottom regions. The weld pool is significantly narrowed due to the untypical growth of the mushy zone in the low-temperature region, which results in a direct blocking effect on the downward flow and the premature solidification in the middle region. The Lorentz force produced by a transverse oscillating magnetic field shows the potential to change the flow pattern into a single-circulation type and the low-temperature-gradient region is mitigated. Therefore, the downward transfer channel is widened, and its premature solidification is prevented. The numerical results are well validated by experimental measurements of metal/glass observation and X-ray fluorescence element mapping.
• Numerical analysis of the dependency of the weld pool shape on turbulence and thermodynamic activity of solutes in laser beam welding of unalloyed steelsArtinov, A.; Kising, P.; Bachmann, M.; Meng, X.; Rethmeier, M.; 10.3217/978-3-85125-968-1-09A three-dimensional numerical model was developed to accurately predict the steady-state weld pool shape in full penetration laser beam welding. The model accounts for the coupling between the heat transfer and the fluid dynamics by considering the effects of solid/liquid phase transition, thermo-capillary convection, natural convection, and phase-specific and temperature-dependent material properties up to the evaporation temperature. A fixed right circular cone was utilized as a keyhole geometry to consider the heat absorbed from the laser beam. The model was used to analyze the influence of the thermodynamic activity of solutes and turbulence on the weld pool shape. A mesh sensitivity analysis was performed on a hybrid mesh combining hexahedral and tetrahedral elements. For the case of full penetration laser beam welding of 8 mm thick unalloyed steel sheets, the dependence of the weld pool shape on the surface-active element sulfur was found to be negligible. The analysis of the results showed that a laminar formulation is sufficient for accurately predicting the weld pool shape since the turbulence has a minor impact on the flow dynamics in the weld pool. The validity of the numerical results was backed up by experimental measurements and observations, including weld pool length, local temperature history, and a range of metallographic cross-sections.
• FEM study of thermomechanical welding of austenitic stainless steel and experimental validationWang, P.; Szalowski, B.; Azuke, J. E.; Vallant, R.; Poletti, M. C.; Enzinger, N.; 10.3217/978-3-85125-968-1-10A coarse-grained microstructure is usually formed in the fusion zone (FZ) and heat-affected zone (HAZ) of TIG welds. The coarse grains usually degrade the service performance and difficult the inspection of welds. Since the austenitic stainless steels do not transform in solid-state, hammering applied on the hot already solidified weld during cooling can be used to promote grain refinement. In this study, large local plastic deformation on the solidified TIG welds of AISI 304L austenitic stainless steel was applied by frequent hammering (at 35 Hz) at a distance of 20 mm behind the welding heat source. The microstructure of welds with and without hammering are characterized using light optical microscopy. Additionally, the temperature, effective strain, and effective strain rate distributions were calculated using the finite element modelling, simulating the integrated arc welding and frequently plastic deformation via the SFTC Deform®-3D. The microstructures were significantly refined in FZ due to mechanical vibration of the solidifying phase and dynamic recrystallization of the solid phase due to large local strains and strain rates. The grains in the HAZ remain coarse because of the negligible plastic deformation. The fusion line can be simulated accurately, and larger temperature heterogeneity was observed in the hammered welds. Therefore, the coupled thermomechanical simulation revealed the correlation of microstructure refinement in the weld and the thermomechanical welding process.
• Study of resistance spot welding via experimental, numerical and advanced analytical methodsGao, H.; Zwart, R.; Aa, E. vd; Veldt, T. vd; 10.3217/978-3-85125-968-1-11In this work, finite element welding models are firstly validated with experimental measurements. Cases with different weld time, weld current and squeeze force are subsequently simulated in order to create a new database. Multiple linear regression, decision tree and random forest methods are used to train the analytical models. The trained analytical models are used to predict new experiments with reasonably good accuracy. In addition, weights for the input variables are explicitly ranked and discussed, which provides valuable information for optimizing the welding process. The main conclusions are summarized as follows, • The approaches proposed in this work to combine the experiments, numerical models and analytical models are proven to be reliable and can be extended to other materials and processes. • Increase in weld time or weld current, or decrease in squeeze force will increase the nugget size. With the studied material, multiple linear regression model provides two equations to calculate the nugget diameter and height with respect to weld time, weld current and squeeze force. Decision tree has a slightly better accuracy than multiple linear regression. Random forest provides the best predictions. • Weld current has a dominating weight (more than 0.85) to determine the final nugget size. Squeeze force has a weight of 0.12 for determining the nugget height.
• Establishing an automated heat-source calibration frameworkRissaki, D.K.; Vasileiou, A.N.; Smith, M.C.; Muránsky, O.; Benardos, P.G.; Vosniakos, G.-C.; 10.3217/978-3-85125-968-1-12Heat source model calibration is a critical step in the process of developing a weld model. This typically involves running a thermal analysis for various sets of welding input parameters (efficiency, heat source radii, etc), and comparing the simulation results against experimental data, including thermocouple traces and fusion zone boundaries. This trial-and-error approach takes time and requires user judgement. This work aims towards establishing an automation and optimisation framework for heat-source calibration. Exhaustive search, exploration of solution space, and identification of suitable metrics are implemented and applied for the automated heat source calibration of a previously validated arc welding benchmark.
• Numerical Analysis of the influence of an auxiliary oscillating magnetic field on suppressing the porosity formation in deep penetration laser beam welding of aluminum alloysYang, F.; Meng, X.; Bachmann, M.; Artinov, A.; Putra, S. N.; Rethmeier, M.; 10.3217/978-3-85125-968-1-13The contactless magnetohydrodynamic technology has been considered as a potential and promising method to improve the weld qualities of deep penetration laser beam welding. In this paper, numerical investigations are conducted to study the influence of the auxiliary oscillating magnetic field on the porosity suppression in laser beam welding of 5754 aluminum alloy. To obtain a deeper insight into the suppression mechanism, a three-dimensional transient multi-physical model is developed to calculate the heat transfer, fluid flow, keyhole dynamic, and magnetohydrodynamics. A ray tracing algorithm is employed to calculate the laser energy distribution on the keyhole wall. A time-averaged downward Lorentz force is produced by an oscillating magnetic field. This force acts in the molten pool, leading to a dominant downward flow motion in the longitudinal section, which blocks the bubble migration from the keyhole tip to the rear part of the molten pool. Therefore, the possibility for the bubbles to be captured by the solidification front is reduced. The electromagnetic expulsive force provides an additional upward escaping speed for the bubbles of 1 m/s ~ 5 m/s in the lower and middle region of the molten pool. The simulation results are in a good agreement with experimental measurements. Based on the results obtained in this study, a better understanding of the underlying physics in laser beam welding enhanced by an auxiliary oscillating magnetic field can be provided and thus the welding process can be further optimized reducing the porosity formation.
• Simulation of laser assisted double wire deposition welding with two different approaches with eulerian (FVM) and lagrangian (SPH) methodsReisgen, U.; Sharma, R.; Mokrov, O.; Emadmostoufi, S.; Kruska, J.; Hermsdorf, J.; Lammers, M.; Bokelmann, T.; 10.3217/978-3-85125-968-1-14Laser-assisted double-wire deposition welding is a welding process developed at LZH to increase the deposition rate with the minimum degree of dilution of the surfacing layer. The use of the laser beam serves to locally heat up the molten pool and substrate surface and, thanks to a smaller wetting angle, leads to improved bond to the substrate, which causes a wider and deeper weld pool. The aim of this paper is to investigate the basics of this process using numerical methods. The free surface of the deposition layer, the mass flow of the melting wires in the weld pool, as well as interphase mass exchange (e. g. evaporation), laser absorption and interphase heat balance were calculated. Two different simulation methods Eulerian finite volume method (FVM) and Lagrangian smooth particle hydrodynamics (SPH) were used to build the model. This paper deals with the model construction as well as the precision and computational effort of these methods. The results of both methods agree with each other. Conclusions were drawn about the advantages and limitations of both methods
• Simulation of Residual Stresses During the wire arc additive Manufacturing (WAAM) ProcessAlrumayh, A. A.; Nied, H. F.; 10.3217/978-3-85125-968-1-15This study examines the residual stresses and distortion that can occur during the Wire Arc Additive Manufacturing (WAAM) process. Of particular interest is the evolution of the residual stresses due to the repetitive heating and cooling associated with the moving arc heat source and deposition of hot metal during the layer-by-layer metal deposition process. The presented results show how the fundamental welding process parameters, thermomechanical material properties, and clamping/fixture conditions affect the post-processing residual stresses. For thin-walled structures, these residual stresses can result in significant warpage and/or premature cracking. This computational study is primarily based on the application of finite element models generated using the SYSWELD finite element software to simulate the coupled heat transfer and mechanical behavior during the layer-by-layer "printing" of a thin, 3-D rectangular plate. Of particular interest are the residual stress comparisons between an austenitic stainless steel, AISI 316L, and a low-carbon steel alloy, S355J2G3. The differences in residual stresses are closely related to the differences in the volumetric strains associated with the metallurgical phase changes between these two different steel alloys.
• Numerical analysis of welding process for distortion prediction of pipe structures for aerospace industryRasche, S.; Fischer, M.; Hildebrand, J.; Bergmann, J. P.; 10.3217/978-3-85125-968-1-16The preferred joining process for lightweight titanium alloyed pipe structures used in aerospace industry is Tungsten Inert Gas welding. The further development and automation of the currently time-consuming and cost-intensive manual production of titanium bleed-air tubes is addressed in the collaborative research project ASciE. The Production Technology Group at Technische Universität Ilmenau works together with PFW Aerospace GmbH to develop finite element simulation models for the distortion prediction of thin-walled welded pipe structures. In this project the influence of varying welding process parameters and heat source models as well as different clamping conditions and thermal boundary conditions on the welded geometry will be studied numerically. The real geometry of the circumferential welded components is measured by means of an optical 3D measurement system, which is well suited for the evaluation of spatial distortions of structures with high resolution. Therefore, the prediction quality of the numerical model can be validated. This conference contribution presents an example of a butt welded pipe component, the results of 3D measurement and the results of finite element analyses. The challenges that arise for a realistic prediction of the welding distortion of the thin-walled pipe component will be discussed.
• Effect of phase- and temperature- dependent strain-hardening slopes on the calculated welding residual stresses in S235 steelSun, J.; Nitschke-Pagel, T.; Dilger, K.; 10.3217/978-3-85125-968-1-17The purpose of this work is to systematically clarify the impact of phase- and temperature-dependent strain-hardening slopes on the calculated welding residual stresses in the commonly used structural steel S235. Both experimental methods and numerical simulation have been utilized for investigation. The results reveal that the temperature-dependent strain-hardening slopes of the generated phases (austenite and bainite here) have nearly no influence on the simulated welding residual stresses. The calculated magnitude of longitudinal residual stress in the base metal near the weld area is highly sensitive to the applied strain-hardening slopes of the initial microstructure (ferrite here), while that of transverse residual stress is not. Meanwhile, comparing to the strain-hardening slope of the initial microstructure at elevated temperatures, that at room temperature plays a critical role in the simulated longitudinal residual stress. In this study, the guidance on how to economically and reliably determine the temperature- and phase-dependent strain-hardening slopes of a given steel in numerical welding simulation is provided.
• Validation of Welding Structure SimulationsLoose, T.; Girresser, T.; Goldak, J.; 10.3217/978-3-85125-968-1-18The Welding structure simulation is a numerical method that predicts distortions, residual stresses and microstructure in welded structures. It enables design engineers to optimize the design and the manufacturing process for the strength and usability of the assembly. To trust the simulation software, it should be validated to demonstrate that the predictions provide a best fit with the reality. We want to prove, that our numerical model of simulation matches the physical behavior of the reality. We must ensure that the virtual experiment matches the physical experiment and that we compare the same sensor and virtual data: • Same time • Same location • Same state Previously, the comparison between the virtual and physical experiment was limited to the final result: final distortion, final residual stress and final residual strain. Now also the transient state during the process shall be considered: transient measurement of temperature field, strain field or deformation. This validates not only the final results but also the computational algorithm that leads to these results. This paper presents transient results of validation experiments with the scope on deformation. The experiment, welding of an orthotropic plate, was chosen in accordance with a published experiment from Murakawa [1]. Because the process is transient, it is important that the transient data be compared.
• Numerical simulation of L-PBF additive manufacturing of mediummanganese steel for automotive crash applicationsAbburi Venkata, K. ; Schob, B.; Kasprowicz, M.; 10.3217/978-3-85125-968-1-19Advanced High Strength Steels (AHSS) such as dual-phase steels are favoured for conventional crash box applications due to the excellent combination of strength and ductility. Generation three AHSS steels such as medium-Manganese Transformation Induced Plasticity (TRIP) steels are a possible alternative to fabricate prototype crash-boxes with equivalent properties of a conventional crash-box due to the TRIP effects. Laser Power Bed Fusion (L-PBF) can produce prototype crash boxes without the requirement of costly dies as in conventional manufacturing. This allows significant benefit in lead times and cost efficiency in manufacturing prototype crash boxes. A reliable numerical simulation tool can predict the L-PBF build process accurately while considering the thermo-metallurgical and mechanical behaviour of the material under multiple thermal cycles and aid the prototype design phase. In the current paper, an improved methodology for the simulation of L-PBF build process using finite element (FE) framework is presented. The proposed methodology provides better spatial resolution of the build process and considers the effects of phase transformation in the medium-Manganese TRIP steel during multiple thermal cycles thereby increasing the accuracy of numerical predictions. The model is set-up and analysed using commercial software Simufact Welding 2022 based on FE solver Marc 2021.2. A comparison of the simulation results with that of experimental analysis on a simple cantilever and a representative double-hat profile crash geometry indicates a very good agreement proving the suitability of the current approach for accurate simulation of L-PBF process whilst maintaining reasonable computational efficiency.
• A simulation approach for series production of plasma-based additive manufacturing of Ti-6Al-4V componentsBielik, M.; Neubauer, E.; Kitzmantel, M.; Neubauer, I.; Kozeschnik, E.; 10.3217/978-3-85125-968-1-20Plasma Metal Deposition (PMD®) is a variant of a directed energy deposition (DED) process that uses an arc welding process to additively manufacture metal components. This process is characterised by relatively high deposition rates, low restrictions regarding the build space, low investment, and operating costs, and is, therefore, predestined for the series production of large structural parts. Additionally, advances in machine development, path planning, and the use of structural welding simulations are bringing these additive manufacturing (AM) technologies into the focus of modern production strategies. To ensure the quality of additively manufactured components in an exactly reproducible manner is, however, a challenge. This requires the complete reproducibility of the manufacturing process and the materials used. This paper investigates the Plasma Metal Deposition manufacturing process of a small batch of Ti-6Al-4V components. Numerical approaches for predicting temperature fields, distortions and residual stresses are examined using the Finite Element (FE) software Simufact Welding 8.0. The focus of the investigations is on the numerical analysis of the influence of the manufacturing process on the mechanical behaviour for multi-layer components. As an outstanding example, the manufacturing of an aerospace bracket is presented.
• Extraction of process-structure-property linkage using deep learning methodsInoue, J.; Noguchi, S.; 10.3217/978-3-85125-968-1-21The establishment of the process–structure–property linkage is essential for designing new materials with desired properties. Based on the concept, the discovery of new materials has been accelerated in the field of functional and bio- materials by combining quantum and molecular modeling tools with efficient machine learning methods. However, in the case of structural materials, even though the development of Integrated Computational Materials Engineering (ICME), it is still difficult to efficiently design new materials because of the uncertainties within models and experimental data. In the present paper, our recent development of a general methodology for extracting the linkage between hierarchical microstructure and process conditions as well as properties will be reviewed. In the proposed method, the uncertainties will be captured in the form of probability density functions using deep learning methods. Since microstructures of typical structural materials are composed of finite kinds of dissimilar phases developing competitively with totally different physical processes, they are supposed to have different geometrical features while maintaining spatial orders. The framework, thus, has two functional components: one is for extracting geometrical features of material microstructures necessary to decompose each different microstructures, and the other is for clarifying spatial orders among the extracted characteristic components. The method was applied for generating virtual steel microstructures obtained after a certain continuous cooling process and those for desired mechanical properties. The obtained results show that the proposed methodology not only generates realistic microstructural images comparable to real experimental images but also clarifies a part of microstructures critically affecting the target property. The proposed approach has been developed to help designing the optimum welding parameters as well as structural materials with an improved weldability.
• Microstructure evolution subroutine for finite element analysisShan, Y. V.; Viernstein, B.; Kozeschnik, E.; 10.3217/978-3-85125-968-1-22Existing Finite Element Method software can be used in a broad field of material characterization, such as heat conduction, plasticity, electric conductivity or fluid mechanics. However, in terms of microstructure, there is a lack of sophisticated packages to thoroughly model the evolution of these parameters. In the present work, a simple but extensive subroutine is presented, to express the kinetics of precipitation and grain growth on the one hand, and the evolution of structural defects, such as dislocation density and vacancy concentration, on the other hand, in dependence of temperature and deformation rate. As a result, further technologically important material properties, such as yield strength, can be derived with the knowledge of aforementioned parameters. The basic functionality of the subroutine is outlined and the handling of the state parameters, which are used during calculation, are explicated.
• Stress-strain properties of HSS steel welded joint heterogeneous structure: experimental and numerical evaluationTomerlin, D.; Kozak, D.; Gubeljak, N.; Magić Kukulj, M.; 10.3217/978-3-85125-968-1-23Welded joints show significant heterogeneity as they are composed of base metal, heat affected zone and weld metal. Heat Affected Zone (HAZ) is further divided into characteristic segments. All the listed zones of the welded joints have certain microstructural differences and consequent differences in mechanical properties. Mechanical experimental examination and determination of stress-strain characteristics of such heterogeneous welded joints structure, especially in certain segments of the HAZ is very difficult. The classical approach to stress-strain testing using standard tensile specimens have only limited applicability, as even the subsize tensile specimens are difficult to position within the narrow HAZ segments. Difficulties in such experimental measurements and the possibility of testing the welded joints in full scope are the motivation for use of alternative experimental methods. The paper considers double-V butt joint made of High Strength Steel (HSS), welded with filler metal having approximately the same mechanical properties as base metal. Experimental work is focused on stress-strain properties determination with Mini tensile Specimens (MTS) used to determine the properties along the transverse weld line. The aim of the paper is the development of an appropriate computer model based on sufficient experimental data, describing the complete welded joint and its specific zones. The evolution of such model is done, starting with a simple model and refining it to a complex, fully segmented welded joint model. This final welded joint model is implemented into ASTM E8 large size specimen, oriented transverse to the welded joint, and covering all specific zones of the welded joint. Material behaviour is simulated using the ductile damage initiation criterion. Tensile test simulation results show good correlation between experimental data and numerical evaluation. Simulated tensile specimen fracture location matches the HAZ segment with decreased strength values, most prone to failure. The paper demonstrates the possibility of experimental determination of stress-strain mechanical properties throughout the heterogeneous welded joint regions, using the MTS specimens. These results can then be used to create a fully segmented welded joint model for tensile testing, or some similar applications.
• Analysis of Solidification Cracking Considering Mechanical and Metallurgical FactorsMaeda, S.; Kato, T.; Ikushima, K.; Shibahara, M.; 10.3217/978-3-85125-968-1-24The introduction of tack welding by large-heat-input welding instead of using an out-of-plane constraint has been considered for butt welding automation. However, the occurrence of solidification cracking in welding is a major issue in large-heat-input welding. Solidification cracking is a welding defect that can significantly reduce the strength of weld joints and structures; therefore, it is important to predict its occurrence for its prevention. Solidification cracking is a phenomenon caused by the interaction of mechanical and metallurgical factors and it is known to have a tendency to occur where columnar crystals collide. In the present study, we proposed a numerical analysis method to evaluate the occurrence of solidification cracking in welding using the thermal elastic-plastic analysis using finite element method (FEM) while considering mechanical and metallurgical factors. As a mechanical factor, a cracking evaluation index using the increment of plastic strain that occurs in the solidification brittle temperature range (BTR) during the cooling process is proposed. As a metallurgical factor, the solidification shrinkage strain and the strength of the solid-liquid coexistence region that changes with the solid fraction are modelled. In addition, we proposed a simplified method for predicting the direction of columnar crystal growth using temperature gradients in the BTR. The proposed method was applied to butt welding, and the effect of welding conditions on solidification crack generation was investigated. It was found that the direction of columnar crystal growth predicted by the simplified method using temperature gradient in the BTR was in good agreement with the experimental results. In addition, the crack generation positions obtained with an evaluation index using the plastic strain increment in the BTR were in good agreement with the cracking positions observed in experiments. The influence of the solidification shrinkage strain on solidification crack generation was examined in terms of the relation between the collision angle of columnar crystals and the restraining state around the melting part. Based on these results, it can be concluded that the effects of mechanical and metallurgical factors on solidification cracking in welding can be analysed using the proposed method.
• Numerical study of the TEKKEN welding testPaget, A. A. M.; Robin, V.; Draup, J.; Hendili, S.; Ufarté, R.; Khan, T.; Delmas, J.; Smith, M. C.; 10.3217/978-3-85125-968-1-25To improve the assessment methods of the cold cracking risk, the usage of the Tekken welding test is proposed. The test load is generated by the residual stresses, whose magnitude is controlled by the size of the Tekken mock-up. In order to conduct a subsequent test benchmarking campaign, the optimum dimensions providing a suitable, realistic load are sought. A 2D computational model is developed to answer this question. A range of thicknesses from 15 mm to 150 mm is investigated. An enrichment by experimental data allows predictions with a great level of confidence. For the welding parameters and metals selected, it is found that the level evolution of the self-restraint conditions reaches a stabilization above a thickness of 50 mm. Moreover, the non-negligible effect of the solid-solid phase transformation induced material property changes is highlighted.
• Assessment of fatigue behaviour of UHSS steel butt-welded joints by means of a fracture mechanics methodologySteimbreger, C.; Gubeljak, N.; Vuherer, T.; Enzinger, N.; Ernst, W.; Chapetti, M. D.; 10.3217/978-3-85125-968-1-26Advances in steel manufacturing technologies made possible the use of high-strength steel (HSS) and ultra-high strength steel (UHSS) in bridges, cranes, offshore structures, oil pipelines and automotive parts. Welding procedures had to be developed to join these materials successfully, but this is still a major issue in mechanical design of HSS elements. Particularly in welding codes and design documents, fatigue resistance of as-welded joints is normally considered to be independent of the base material (BM) static strength. However, cyclic loaded as-welded components with high quality or post-weld treated joints have shown improved performance when using HSS and UHSS as the base material. The present work aims to apply a fracture mechanics methodology to the analysis of fatigue behaviour of welded joints. The approach requires estimating the driving force available for subcritical crack growth at the location of maximum stress concentration. In this regard, stress intensity factor proved to be a sensible parameter that can account for loading scheme and local weld geometry. It can be determined by numerical modelling, which demands a change from continuum mechanics stress analysis to one that estimates fracture mechanics parameters, considering the existence of defects and cracks. Then, the total driving force applied to the crack can be compared to its threshold for propagation, resulting in the effective driving force for crack growth. Particularly, the effect of welding process on the fatigue behaviour of ultra-high strength steel butt-welded joints was studied. Sheets of steel S960MC and S960QL were joined with different welding techniques: Gas Metal Arc Welding (GMAW), Laser Hybrid Welding (LHW) and Electron Beam Welding (EBW). To validate the model, fatigue tests were performed with stress ratio R = 0.1, under four points bending loading. Joints manufactured with GMAW exhibited the highest fatigue strength of the three configurations. Compared to the fatigue limit of the BM, a decrease in fatigue strength around 60% was observed in welds jointed with LHW and EBW, although the latter showed longer fatigue lives for higher nominal stresses. Proposed methodology allows to assess the effect of microstructure, defect size, hardness, and joint geometry resulting from each welding technique. Results conservatively describes the fatigue behaviour of each weld configuration and highlights the relative influence of all factors considered in the assessment. Although the validated results request further studies to improve understanding of the acting mechanism, they also show the potential of welded HSS and UHSS joints compared to the standard design approaches.
• Efficient numerical analysis of directed energy deposition processesElsner, B. A. M.; Neubauer, I.; Raberger, L.; 10.3217/978-3-85125-968-1-27Directed energy deposition (DED) presents a versatile method in the field of Additive Manufacturing that allows to create complex structures by continuously welding new filler material to the underlying structure. The technology is developing quickly and DED structures with several kilometres of weld length are already engineering reality. Although numerical analysis has proven a valuable tool for the evaluation and understanding of different welding processes, typical transient simulations cannot handle such long weld tracks as time needs to be discretized in increments short enough to track the movement of the weld source in steps not larger than its own dimension. In this contribution, we present an Advanced Thermal Cycle (ATC) that allows to reduce the number of computational steps while still capturing the heat source’s movement, local temperature differences, and maintaining the correct energy balance. The approach is validated by comparing the simulated thermal profile with the thermal history measured both for the substrate and the deposited weld filler of a demonstrator part.
• Prognosis of seam geometry during laser beam weldingSchwarz, C.; Puschmann, M.; Mauermann, R.; 10.3217/978-3-85125-968-1-28Data evaluation is of great importance for quality assurance and control loops. Components and process parameters involved in the production process can be networked and evaluated using all relevant information and controlled in real time. Thermal joining processes are complex; so is laser beam welding (LBW). The numerical description of the processes provides good approximations for partial aspects. However, experiments are still the basis for determining optimal process parameters. This is time-consuming and cost-intensive. For the evaluation of experimental data there are some AI approaches; e.g. response surface method, Taguchi method, KNN models, Kriging models, principal component analysis (PCA). Systematic backup, analysis and visualization of welding and quality data using database systems and analysis algorithms is not currently taking place on a wide scale. Expert knowledge is a mandatory prerequisite for the preparation, execution and evaluation of LBW processes. Therefore, the potentials of the process are often not fully exploited. The consequences are e.g. long commissioning times, low flexibility for new tasks and missing objective knowledge management. In the lecture a tool based on PCA will be presented. The interdependencies between process parameters and the welding result in the form of the 2D weld geometry are mapped in a statistical model. The system is learned from experimental data sets. The weld geometry can contain further characteristic values; e.g. weld penetration depth, weld width or load-bearing weld cross-section. These characteristic values can be linked to the quality of the welded joint. The two-dimensional weld formation and all contained spatially resolved result variables can be represented; e.g. width of heat affected zones and grain size distribution. This requires the analysis of multivariate data; e.g. micrographs with a pixel number of several million, as dependent result variables with all nonlinear dependencies. To realize the spatial resolution of the result variables, the full pixel resolution is used for image analysis. From the formed statistical model, the seam geometry with all properties can be predicted ad hoc within the learned data space for arbitrary parameter combinations, or local target variables can be specified and an optimization algorithm searches for the best possible parameter combination from many model queries. Using an LBW task as an example, the evaluation principle and the GUI are shown. Key points are the permanent accumulation of knowledge, usable control strategies, quality proofs and thus time and cost savings.
• Numerical analysis of ultrasonic vibration enhanced friction stir welding of dissimilar Al/Mg alloysYang, C. L.; Wu, C. S.; Bachmann, M.; Rethmeier, M.; 10.3217/978-3-85125-968-1-29The ultrasonic vibration enhanced friction stir welding (UVeFSW) process has unique advantages in joining dissimilar Al/Mg alloys. While there are complex coupling mechanisms of multi fields in the process, it is of great significance to model this process, to reveal the influence mechanism of ultrasonic vibration on the formation of Al/Mg joints. In this study, the acoustic plastic constitutive equation was established by considering the influence of both ultrasonic softening and residual hardening on the flow stress at different temperatures and strain rates. And the ultrasonic induced friction reduction (UiFR) effect on friction coefficient in different relative directions at the FSW tool workpiece interface was quantitatively calculated and analyzed. The Al/Mg UVeFSW process model was developed by introducing the above acoustic effects into the model of Al/Mg friction stir welding (FSW). The ultrasonic energy is stronger on the aluminum alloy side. In the stirred zone, there is the pattern distribution of ultrasonic sound pressure and energy. The heat generation at the tool workpiece contact interface and viscous dissipation were reduced after applying ultrasonic vibration. Due to the UiFR effect, the projections of friction coefficient and heat flux distributions at the tool workpiece interface present a "deformed" butterfly shape. The calculated results show that ultrasonic vibration enhanced the material flow and promoted the mixing of dissimilar materials.
• A process modelling approach to the development of lap welding proceduresLewis, M.; Smith, S.; 10.3217/978-3-85125-968-1-30Process modelling (PM) is used in many industries to make considerable contributions to the development of safe and efficient engineering designs. The authors have experience of applying PM to a wide range of industries and applications; they are currently investigating the application of commercial CFD software to Friction Stir Welding (FSW). PM augments results generated by laboratory testing by providing full field and full duration predictions of important parameters such as material state. Analyses of engineering processes are increasingly conducted with commercial software, which attracts a significant license fee, but license fees may not be the only barrier to the utilisation of PM in engineering design. PM can be very complex and the requirements for an understanding of many advanced material behaviours may deter their use. It might fairly be expected that there could be a considerable investment needed before a desirable return is achieved. FSW has been the fastest growing joining technology over the last two decades and it now being used in many industries, including aerospace, automotive and shipping. The process has advantages over the more traditional techniques for joining metallic materials. Results show that FSW joints have a superior fatigue resistance when comparted to fusion joining, and the lower process temperature means that thinner sheet fabrications are less likely to suffer heat-related distortion. FSW procedure development is needed for each application to show that the proposed welding procedure for any application will robustly produce defect-free welds of suitable quality. The current paper extends a FSW PM technique which assumes that accuracy mostly depends upon the conditions that exist near the FSW tool plus some straightforward mechanics and thermodynamics remote from the tool. The technique can be used for a wide range of materials using properties that are generally available and easily understood by a process engineer. Comparisons are made here against test data of a series of lap welds for a range of conditions. Of particular interest is the effect of two different tool pin designs on the corresponding welds. The test data showed that welding parameters (rotation rate and welding speed) and tool pin design strongly affect the characteristics of the typical defect features (hooking and cold lap defects). The comparisons showed that the tool pin design including thread must be captured explicitly in the model, and that the relative position of the tool as it rotates must also be modelled. This results in quantitatively accurate predictions of torque and qualitatively realistic results for hooking and cold lap defects. The approach is based upon the use of general purpose CFD software combined with a set of easy to access materials data.
• Impact of activation in projection welding with capacitor discharge using multiphysics simulation and a process-integrated transition resistance measurementKoal, J.; Baumgarten, M.; Zschetzsche, J.; Füssel, U.; 10.3217/978-3-85125-968-1-31Capacitor discharge welding (CDW) is defined by a pulsed current profile. It is usually used for projection welding and is characterized by very high power density and very short welding time. The process can be classified into four phases: Contacting, Activating, Material Connection and Holding Pressure. High-speed images show the formation of metal vapor during the activation phase. This removes impurities and oxide layers from the contact zone and activates the surfaces. The material connection is achieved by pressing the activated surfaces together. The purpose of the investigation is to describe the impact of activation on the formation of the material connection. A capacitor discharge welding system at the Technische Universität Dresden has the unique feature of interrupting the current flow at a given time. That allows different current profiles with the same rate of current increase. This unique feature permits the experimental investigation of activation and their impact on the formation of the material connection. To achieve this, projection components with different interruption times were welded. The transition voltages were measured before, while and after welding at each contact area. The mechanical strength of the welded joints was measured and a metallographic investigation was performed. To evaluate the impact of the activation in time and location, the experimental results were combined with simulative investigations. The evaluation of temperature and current density distributions enables the physical description of the activation. An iteratively coupled electrical-thermal and mechanical-thermal FEM-model was used. The experimental measured process variables force and current were used as boundary conditions for the simulation. A contact theory was implemented. In addition, the model considers temperature-dependent material characteristics and a remeshing procedure to model high projection deformations. The validation of the simulation was done by comparing measured and simulated transition voltages at the contact areas.
• Simulation model for laser hardening of small-diameter holesEvdokimov, A.; Jasiewicz, F.; Doynov, N.; Ossenbrink, R.; Michailov, V.; 10.3217/978-3-85125-968-1-32The Gaussian laser beam can be transformed into a ring-shaped beam profile. The laser hardening can be performed by directing the obtained ring spot to the inner surface of the small-diameter holes. In this case, the laser irradiation falls inclined to the material surface. In this study, a simulation model was developed, which allows for the computation of temperature distribution during the laser hardening of the holes. The heat source model was developed to describe the intensity distribution on the inner surface of the holes as a function of the inclination, distance to the laser focal plane, and parameters of the laser beam. The intensity distribution changes during the hardening of complex holes (varied diameters along the hole). The developed simulation model accounts for the changes in the laser inclination, the distance to the focal plane, and the circumference of the irradiated area inside the hole during the processing. The simulations were performed using the open-source software FEniCSx. The developed model was validated by comparing the computed and experimentally measured temperature cycles.