Search

Mixing in extrusion is a vital part of achieving consistent and high-quality extrudates, with residence time being an elucidative measure of the mixing performance. Recent studies around numerical modeling of residence time distributions in single-screw extruders appear to consider flooded extruders mainly. This paper introduces a new and general CFD model to characterize the extruder fill length and residence time distribution for a viscoplastic ceramic material in a starve-fed extruder, including free surface tracking. The CFD model simulates a pulse-injection test, where a fluid parcel is injected at the inlet, with subsequent outlet concentration measured over time. The study includes material characterization and model validation based on laboratory tests. Results quantify the impact of accounting for the partially filled extruder instead of assuming it to be flooded, addressing the potential error when only considering simple analytical approximations to calculate system average residence times. Results further show the ability to fit simulation results to more simple analytical models. This underlines the importance of including the entire extrusion system and forming the basis for further work toward enabling real-time model predictions in starve-fed extrusion systems.

Metal additive manufacturing is increasingly used as a complementary manufacturing technique in industrial settings and slowly moving from pure prototyping applications toward full production. In parallel, there is an emergence of Industry 4.0, where the applicability of concepts such as digital twins of manufacturing machines and components are being investigated. Compared to conventionally manufactured parts, typical quality metrics of metal additively manufactured components such as dimensions, roughness, porosity, and hardness are underperforming in an as-built state. As a mitigation strategy, the build chamber variables are often measured and logged by the metal additive manufacturing system to maintain a stable production environment. Thus, proper insight into the expected responses in part quality from changes in those build chamber variables is important in the pursuit of digital twins and process improvement. This sheds more light on the influence of the gas flow variables, namely gas flow speed, relative pressure, and oxygen content on the metal additive manufacturing quality metrics, specifically channel roughness, bulk porosity, average diameter, the equivalent diameter of the unobstructed cross-sectional area, and hardness of the bulk. A Design of Experiments was implemented on two laser powder bed fusion systems, namely an SLM 280 processing 316L stainless steel and an SLM 500 processing Ti6Al4V. The current work found that surface oxidation of 316L and Ti6Al4V components may be classified based on simple red, green, and blue (RGB) color constituent analysis. The influence of gas flow variables was found to be different in the two investigated SLM systems, suggesting a high dependency on the processed material. Oxygen content in the build chamber had the highest standalone effect on the selected quality metrics, while the gas flow speed had the lowest standalone effect. The second-order effects were found to be, in general, more significant than the main effects. The findings of the current work is a step towards an improved understanding of the interaction effects of gas flow conditions on typical quality metrics of metal additive manufactured components. By the creation of simple but computationally fast response surface models, in-line assessments may be carried out and the effect of process variability on component quality may be evaluated in-situ while being one step away from the full feedback control implementation in the digital twin. Following the methodology of the current work for other laser powder bed fusion systems enables the generation of 3D point cloud visualizations for decision making under uncertainty.

Grain refinement of additively manufactured titanium and titanium alloys can be promoted via adding foreign elements or particles, but it may lead to a reduction in ductility due to the formation of brittle intermetallic compounds. The present study shows that in-situ grain refinement of commercially pure titanium (CP-Ti) can be achieved through properly controlling the selective laser melting (SLM) parameters. It was found that higher input energy density worked in favour of grain refinement. Detailed microstructural characterization coupled with multi-physics simulation were performed to reveal the grain refinement mechanism. This was attributed to the intrinsic heat treatment (IHT) effect which resulted from the cyclic reheating inherent to the SLM process. As a result, the refined CP-Ti exhibited an exceptionally high ductility of 34.3 ± 0.5% without notable mechanical anisotropy. This work demonstrates the feasibility of utilizing thermal cycling of additive manufacturing (AM) to refine grains of metals without changing the composition.

Purpose Conformal cooling channels in additively manufactured molds are superior over conventional channels in terms of cooling control, part warpage and lead time. The heat transfer ability of cooling channels is determined by their geometry and surface roughness. Laser powder bed fusion manufactured channels have an inherent process-induced dross formation that may significantly alter the actual shape of nominal channels. Therefore, it is crucial to be able to predict the expected surface roughness and changes in the geometry of metal additively manufactured conformal cooling channels. The purpose of this paper is to present a new methodology for predicting the realistic design of laser powder bed fusion channels. Design/methodology/approach This study proposes a methodology for making nominal channel design more realistic by the implementation of roughness prediction models. The models are used for altering the nominal shape of a channel to its predicted shape by point cloud analysis and manipulation. Findings A straight channel is investigated as a simple case study and validated against X-ray computed tomography measurements. The modified channel geometry is reconstructed and meshed, resulting in a predicted, more realistic version of the nominal geometry. The methodology is successfully tested on a torus shape and a simple conformal cooling channel design. Finally, the methodology is validated through a cooling test experiment and comparison with simulations. Practical implications Accurate prediction of channel surface roughness and geometry would lead toward more accurate modeling of cooling performance. Originality/value A robust start to finish method for realistic geometrical prediction of metal additive manufacturing cooling channels has yet to be proposed. The current study seeks to fill the gap.

Conformal cooling channels, made by the laser powder bed fusion process, are a promising design strategy for optimising cooling and process control of manufacturing tools and injections moulds. The...

This study presents a holistic segmentation procedure, which can be used to obtain individual fibre inclination angles from X-ray computed tomography. The segmentation approach is based on principal component analysis and was successfully applied for a unidirectional and an air-textured glass fibre–reinforced composite profile. The inclination results show a weighted mean fibre inclination of 2.1° and 8.0° for the unidirectional and air-textured profile, respectively. For the air-textured composite, fibre inclinations of up to 55° were successfully segmented. The results were verified by comparative analysis with equivalent results obtained from structure tensor analysis – showing no notable deviation. The comparable characteristics in combination with the distinct differences of the two material systems make this case study ideal for verification and validation of idealized models. It is shown how this approach can provide fast, accurate and repeatable inclination estimates with a high degree of automation.

In this paper, a 2D numerical framework for the simulation of flows in porous media that are characterised by a sharp transition between the saturated and unsaturated zone is presented. Using a finite volume scheme and the level-set method, the framework is derived based on a conventional, implicit solution of Darcys law for the pressure field, while the level-set function is advected explicitly locally at the flow front. With the main application to liquid composite moulding (LCM) in mind, the numerical framework is verified against analytical solutions, experiments, and other benchmark cases for a variety of situations that occur in this composite manufacturing process. The cases include local changes in permeability related to edge-effects, merging of flow fronts, and continuous manufacturing processes. It is highlighted that the numerical framework achieves convergence with a spatial accuracy between 1st and 2nd-order; the level-set operations only take about 20% of the total CPU time; and the propagation of the resin front can be achieved with CFL-sized time steps without overfilling cells and without applying any artificial smoothing.

Molded pulp products are experiencing increased importance due to their environment-friendly characteristics with regards to both production and disposal. In this study, the thermoforming process for the production of molded pulp products was examined by means of two experimental campaigns, with the scope of distinguishing between the water removal due to only compression (better known as cold pressing) and the water removal under additional thermal conditions. The underlying multiphysics of the processes were identified by comparison with similar manufacturing processes found in available literature, and then implemented using a 1-dimensional finite element model in COMSOL software. This fast, simple model was able to predict the water removal for a specific set of process conditions and the results were supported by corresponding experimental observations. Within a process time of up to 15 s, temperatures ranging from 80 °C to 120 °C, and a final dryness between 40% and 65%, the simulations gave accurate estimates, with a difference in prediction down to 0.3%. The paper thus documents the first combined experimental-simulation study for thermoforming of molded pulp products.

Laser-based powder bed fusion (L-PBF) is a branch of additive manufacturing technology which is considered to be a superior process due to its capability of producing complex designs with low material waste. Despite L-PBFs various unique characteristics, manufactured parts still suffer from a wide variety of defects, among which porosity is one of the most important. In this paper, a multiphysics numerical model for the multi-track/multi-layer L-PBF is developed and used for analysing the formation and evolution of voids caused by lack of fusion and improper melting. The multiphysics model is in meso-scale and is used to track and observe the formation of porosities, and considers phenomena such as multi-phase flow, melting/solidification, radiation heat transfer, capillary and thermo-capillary (Marangoni effect) forces, recoil pressure, geometry dependant absorptivity and finally evaporation and evaporative cooling. A novel methodology has been introduced to model the two subsequent powder-laying and fusion processes, for each layer, by means of a discrete element method (DEM) in a Lagrangian framework and a computational fluid dynamics (CFD) model, both implemented in Flow-3D. The results for the investigated process parameters indicate that the porosities (voids) are mainly formed in between the tracks, largely due to improper fusion of the particles. Moreover, it is observed that the pores are mostly elongated in the direction parallel to the laser scanning paths, as expected. The probability of the presence of pores is also observed to be higher in the first layer, where the average layer temperature is lower as well. Furthermore, the lack of fusion zones are seen to become smaller in the subsequent layers, largely due to better fluid flow and higher temperatures, because of heat accumulation in those layers.

Temperature and relative humidity are crucially important for the reliability of electronic devices in outdoor applications. For small enclosures, thermal control by a heat sink that is cooled by natural convection is the most widely used method. In this work, the functionality of a heat sink as a thermal mass for partially storing the heat that is generated by the electronics inside a typical electronics enclosure is studied using transient 3D CFD simulations. The stored energy is released with a time lag for the purpose of enhancing the stability of the local climate inside the enclosure. A typical enclosure is assumed to be exposed to Copenhagen outdoor conditions over a two-day time period. For fast and long term predictions, a Cauer ladder RC model is developed based on CFD simulation results. A correlation for calculating the Nusselt number representing heat transfer between the trapped air inside the enclosure and the heat sink is derived. Compared with CFD simulations, the RC model is found to predict slightly lower values for temperature, and slightly higher values for RH – given that the value of absolute humidity is quite constant over the timescales here, then the higher values of RH are largely due to the lower values of temperature. The lumped model correctly predicts the trend for temperatures and RH variations, however, suggesting that a properly calibrated RC model can be used for scoping calculations, and more importantly for long-term predictions. The method can be used for any other ambient and working conditions, as long as the temperature difference in the air does not exceed 25 [°C].

Channels manufactured by laser powder bed fusion have an inherent process-induced dross formation and surface texture that require proper characterization for design and process optimization. This work undertakes surface texture characterization of AlSi10Mg channels of nominal diameter sizes ranging from 1 mm to 9 mm using X-ray computed tomography. Profile parameters, including Pa, Pz, and Pq, were found to be interchangeable for qualitative characterization of surface texture variation. Psk, Pvv, and the fractal dimension could identify the presence of extreme dross and sintered particles on the measured profiles. A method for predicting the equivalent diameter of the unobstructed cross-sectional area (Deq) was presented and its reduction was found to follow a logarithmic trend, as a function of channel length. An empirical model Pa (β, D), as a function of local angular position (β) and channel diameter (D), was demonstrated on a perfect channel geometry, resulting in well-predicted roughness and internal geometry.

This paper concerns non-isothermal flow in a thermoset resin-injection pultrusion process. Supported by temperature measurements from an industrial pultrusion line and a material characterisation study (curing kinetics, chemorheology, and permeability), the material flow was analysed for the manufacture of a thick glass-fibre profile saturated with a pultrusion-specific polyurethane resin. A central finding is that the heating configuration, together with the strongly convective flow near inlets resulted in phase transitions that were both concave and convex-shaped. This is different from existing literature that commonly describes curing being initiated from die-walls, resulting in the concave phase-transitions.

This study presents a detailed material characterization study of a pultrusion specific polyurethane resin system. Firstly, the chemical behaviour was characterized by utilizing differential scanning calorimetry. The cure kinetics was fitted well to an autocatalytic cure kinetics model with Arrhenius temperature dependency. The resin system did not show any inhibition time. Then, the viscosity evolution was observed using a rheometer. From these measurements, the gelation point was estimated from the criterion of cross-over point of storage and loss modulus curves. The corresponding cure degree of gelation point was found to be 0.79 using the cure kinetics model. The complex viscosity evolution through dynamic scans was predicted well with a cure degree and temperature dependent viscosity model. The elastic modulus and glass transition temperature of fully and partially cured samples were measured by means of DMA in tension mode. A five step cure hardening instantaneous linear elastic model was fitted well to a wide range of cure degree values after gelation. Lastly, the fitted material models were employed in a case study of a typical pultrusion process to visualize the effects of pulling speed on the material property evolution.

Laser-based powder bed fusion, due to its layer-by-layer nature, results in a unique stress profile in a part after the primary production process. The residual stresses are typically tensile near the top, while they are compressive near the bottom of the part. When it is removed without proper precautions, the part will bend excessively. In order to alleviate this deformation, a stress relief heat treatment can be applied. In this paper, such a stress relaxation heat treatment is modelled to investigate the effect of the post-processing parameters. The model uses an Arrhenius-type creep equation to simulate the influence of the heat treatment temperature and dwell time on the stress field in a relatively simple cantilever beam produced in Ti-6Al-4V. Via validation of the simulations, the effect of the heat treatment is shown to be represented accurately. The validated model is used to predict the deformation that results from the residual stresses after heat treating the part under various conditions. The results from the simulations ultimately allow choosing the optimal heat treatment conditions to obtain a given reduction in the residual stress level, while reducing the need for extensive experimental investigations.

Process-induced stress and deformation are critical factors when ensuring product quality and structural integrity of composite profiles manufactured using thermoset pultrusion processes. In this paper, we present a new steady-state 3D-Eulerian numerical framework that enables 9–35 times faster computations compared to the current state-of-the-art quasi-static 3D-methods. In addition, we show how process-induced effects from the profile-advancing pulling force and an initial compressive stress state can be modelled. We demonstrate in theoretical parameter studies that the pulling force advances die-detachment and reduces die-swelling, while the initial compressive stress state has the opposite but a more pronounced effect.

In spite of many years of research, the physical phenomena leading to the evolution of compacted graphite (CG) during solidification is still not fully understood. In particular, it is unknown how highly branched CG aggregates form and evolve in the semi-solid, and how local microstructural variations at micrometer length scale affect this growth process. We present here the first time-resolved synchrotron tomography combined with a bespoke high-temperature environmental cell that allows direct observation of the evolution of CG and relates this dynamic process to the local surrounding microstructures in a cast iron sample during repeated melting and solidification. Distinct processes are identified for the formation of CG involving the nucleation, growth, development of branches and interconnection of graphite particles, ultimately evolving into highly branched graphite aggregates with large sizes and low sphericities. CG is found to nucleate with a spheroidal or a plate-like shape, developing branches induced by high carbon concentration, e.g. thin melt channels. Additionally, CG grows much faster than spheroidal graphite during subsequent cooling in solid state. The direct visualization of the dynamic solidification process provides unprecedented new insights into formation mechanisms of CG and correlating factors such as local microstructural variations, and guides the development of CG iron solidification models.

Several different interfacial forces affect the free surface of liquid metals during metal additive manufacturing processes. One of these is thermo-capillarity or the so-called Marangoni effect. In this work, a novel framework is introduced for unraveling the effects of thermo-capillarity on the melt pool morphology/size and its thermo-fluid conditions during the Laser Powder Bed Fusion (L-PBF) process. In this respect, a multi-physics numerical model is developed based on the commercial software package Flow-3D. The model is verified and validated via mesh-independency analysis and by comparison of the predicted melt pool profile with those from lab-scale single-track experiments. Two sets of parametric studies are carried out to find the role of both positive and inverse thermo-capillarity on the melt pool shape and its thermal and fluid dynamics conditions. The thermo-fluid conditions of the melt pool are further investigated using appropriate dimensionless numbers. The results show that for the higher Marangoni number cases, the melt pool temperature drops, and at the same time, the temperature field becomes more uniform. Also, it is shown that at higher Marangoni numbers, temperature gradients decrease, thus reducing the role of conduction in the heat transfer from the melt pool. Furthermore, for the first time, a novel methodology is introduced for the calculation of the melt pool's average Nusselt number. The average Nusselt numbers calculated for the positive and inverse thermo-capillarity are then used for finding the effective liquid conductivity required for a computationally cheaper pure heat conduction simulation. The results show that the deviation between the average melt pool temperature, using the pure conduction model with effective conductivity, and the one obtained from the advanced fluid dynamics model is less than 2%.

Through various examples, this short review presents the main X‐ray based techniques that are available to characterize nodular cast iron at the microstructural level. Emphasis is placed on the enormous potential offered by the recent developments in X‐ray tomography, X‐ray diffraction and digital volume correlation, which allow collecting microstructural and micromechanical information in 4D (3D plus time) during both casting and subsequent mechanical loading. The goal is to demonstrate that for nodular cast iron, which has an inherently three‐dimensional, composite microstructure, X‐ray based techniques provide some significant advantages over conventional microscopy. For this reason, these techniques can be instrumental in unveiling the mechanisms controlling both the formation of the microstructure as well as its micro‐mechanical behavior during in‐service loading, thus paving the way to the development of improved process‐structure‐property relations.

This chapter presents the various techniques used to model the additive manufacturing of metal components. The concepts of length and time scales of different physical phenomena are discussed, followed by a classification of the different modelling techniques along these lines. The potential applications of these different models towards reducing defects and increasing the precision of metal additive manufacturing are put forward. The chapter rounds off with the introduction of the concepts of uncertainty in models and the need for proper verification and calibration procedures. The particular challenges faced when trying to simulate the metal additive manufacturing process are highlighted, together with the corresponding solution approaches propounded by various researchers.

There is a growing interest for integrating additive manufacturing (AM) technology in different manufacturing processes such as injection moulding (IM) due to the possibility of achieving shorter manufacturing times and increased cost effectiveness. This paper evaluates IM inserts fabricated by the AM vat photopolymerisation method. The inserts are directly manufactured with a photopolymer material, integrated on an injection moulding tool and subsequently used for IM. Therefore, particular attention has to be paid in order to develop the soft tooling process chain and the IM experimental procedure as detailed in this study. Different combinations of IM parameters are investigated in this work in order to determine the influence of the various process settings on the inserts’ performance (lifetime, crack propagation, consistency of the mould surface features). The mould inserts were analysed by three-dimensional optical metrology and evaluated with regard to the different surface features that were affected by the IM process. A three-dimensional thermo-mechanical with phase change model for the analysis of the effects of the IM process on the additive manufactured tools was accomplished in the FE software COMSOL Multiphysics. The potential causes for the insert failure are identified both by means of the IM experiments and the numerical model. The developed model could also predict the thermally induced deformations produced in the mould and identify where this phenomenon would eventually lead to defects in the shape of the parts. The influence of three different temperatures of the insert at 25 °C, 50 °C and 100 °C on the failure of the insert was investigated. Also a detailed discussion about the solidification and temperature changes is given.

Friction stir welding (FSW) is a new welding technique which has been applied to lightweight materials such as polymers. In this work, the feasibility of FSW on poly methyl methacrylate (PMMA) was investigated. For this purpose, an appropriate friction stir welding tool is used to make butt joint welded specimens. Experimental tests were carried out according to an L16 Taguchi design of experiments in which tool rotational speed, tool transverse speed, and tool penetration depth were applied as FSW parameters. In addition, the ductile behavior and weld appearance of welded samples have been studied. In order to determine the effects of FSW parameters on the weld strength, the signal-to-noise (S/N) ratio and analysis of variance (ANOVA) were utilized. According to the ANOVA results, the tool transverse speed has a significant effect on the weld strength. Also, the contribution of each FSW main factor and their interaction have been determined. This study demonstrates that friction stir welding of PMMA is feasible and the range of FSW parameters that provide acceptable welds are designated.

Today, dimensional validation of polymer parts with micrometer level accuracy is performed long time after production because the part needs time to “settle” (which for some polymer materials takes weeks). However, manufacturing industry requires the measurements to be performed before part stabilization, to reduce the waiting time for quality control without compromising the level of accuracy. This work proposes an analytical solution for predicting the reference length of a classical 32 mm polymer part with an uncertainty of less than 10 μm when measured at non-reference conditions. Creep, moisture uptake and temperature are here the main issues to consider for polymer parts. The present study focuses on the dimensional changes governed by moisture uptake and creep with model parameters separately obtained by experimental studies combined with numerical simulations. Finally, the obtained analytical solution is used to predict the reference length of stabilized and non-stabilized polymer parts measured continuously over two months. The prediction shows good agreement with the reference values for settled parts at reference conditions.

In the present work a new method for thermal compensation in dimensional metrology of polymer parts in a production environment based on 3D thermomechanical simulations is developed. A fixture for measuring the length dimension of a classical polymer part is placed in a production environment and equipped with sensors in terms of thermocouples for temperature measurements of the part and the fixture and a contact probe for measuring the dimension. A 3D thermomechanical model is developed in ABAQUS, emulating the thermoelastic conditions of the polymer part when placed in the fixture. Knowledge from classical heat transfer and elasticity theory is then applied to derive a more generic, yet simple expression for the compensation from the transient 3D temperature and displacement fields, based on dimensionless values, which makes applicable for a wide range of initial and surrounding conditions found in a production environment. The developed expression is then used for length compensation on 24 samples measured inline ten minutes after production. The results reveals a significant improvement in capturing the transient behavior of the part with a reduced error from 13 μm to 3 μm, applying the developed formula instead of using more classical 1D standard thermal compensation.

In the present work a 2D axisymmetric thermomechanical model of the direct-drive friction welding process is developed, taking the temperature dependent shear yield stress into account in the description of the heat generation, utilizing a recent thermal pseudo mechanical model originally developed for the friction stir welding (FSW) process. The model is implemented in ABAQUS/Explicit via a subroutine. The application in this case is joining of austenitic stainless steel rods with an outer diameter of 112 mm, used for manufacturing of exhaust gas valves for large two stroke marine engines. The material properties in terms of the temperature dependent flow stress curves used both in the thermal and the mechanical constitutive description are extracted from compression tests performed between 20 °C and 1200°C on a Gleeble 1500 thermomechanical simulator. Comparison between measured and simulated transient temperatures shows relatively good agreement and furthermore, the simulated deformations in terms of upsetting length and flash formation are also in good agreement with the observations from the experiment.

Ductile cast irons (DCIs) are of increasing importance in the renewable energy and transportation sectors. The distribution and morphology of the graphite nodules, in particular the formation of degenerate features during solidification, dictate the mechanical performance of DCIs. In situ high-speed synchrotron X-ray tomography was used to capture the evolution of graphite nodules during solidification of DCI, including degenerate features and the effect of the carbon concentration field. The degeneration of nodules is observed to increase with re-melting cycles, which is attributed to Mg-loss. The dendritic primary austenite and carbon concentration gradients in the surrounding liquid phase were found to control nodule morphology by locally restricting and promoting growth. A coupled diffusion-mechanical model was developed, confirming the experimentally informed hypothesis that protrusions form through liquation cracking of the austenite shell and subsequent localised growth. These results provide valuable insights into the solidification kinetics of cast irons, supporting the design of advanced alloys.

We present an Eulerian free-surface flow solver for incompressible pseudoplastic and viscoelastic non-Newtonian fluids. The free-surface flow solver is based on the streamfunction flow formulation and the volume-of-fluid method. The streamfunction solver computes the vector potential of a solenoidal velocity field, which ensures by construction the mass conservation of the solution, and removes the pressure unknown. Pseudoplastic liquids are modelled with a Carreau model. The viscoelastic fluids are governed by differential constitutive models reformulated with the log-conformation approach, in order to preserve the positive-definiteness of the conformation tensor, and to circumvent the high Weissenberg number problem. The volume fraction of the fluid is advected with a geometric conservative unsplit scheme that preserves a sharp interface representation. For the sake of comparison, we also implemented an algebraic advection scheme for the liquid volume fraction. The proposed numerical method is tested by simulating the planar extrudate swell with the Carreau, Oldroyd-B and Giesekus constitutive models. The swell ratio of the extrudates are compared with the data available in the literature, as well as with numerical simulations performed with the open-source rheoTool toolbox in OpenFOAM®. While the simulations of the generalized Newtonian fluids achieved mesh independence for all the methods tested, the flow simulations of the viscoelastic fluids are more sensitive to mesh refinement and the choice of numerical scheme. Moreover, the simulations of Oldroyd-B fluid flows above a critical Weissenberg number are prone to artificial surface instabilities. These numerical artifacts are due to discretization errors within the Eulerian surface-capturing method. However, the numerical issues arise from the stress singularity at the die exit corner, and the unphysical predictions of the Oldroyd-B model in the skin layer of the extrudate after the die exit, where large extensional deformations occur.

This paper gives an update on current work concerning the development of the method of Dynamic Length Metrology (DLM) to obtain accurate traceable measurements directly in the production. Three case studies are described dealing with diameter measurements with sub-micrometer uncertainty on metallic parts, and dimensional measurements on polymer parts, where the achievable uncertainty of after molding predictions is of a few micrometers.

The growing demands for quality and flexibility and at the same time production speed challenges conventional metrology. The future tendency is that metrology is an integrated part of the production line and thus is placed in a production environment. This is a challenge since dimensional metrology in a production environment might lead to higher uncertainties due to dynamic variations both in the conditions of the environment and in the conditions of the produced parts. However, many of these effects can be treated as systematic errors if the physical phenomena leading to the deviations can be described. Today, it is very common to compensate for the variations in temperature in a classical 1D manner. However, when temperature gradients and very complex part geometries exist the deformation pattern might not at all follow a linear path. Instead, more advanced three-dimensional thermomechanical numerical models should be applied taking the inherent build-up of residual stresses and warpage into account.

Numerical simulations have recently shown their potential as a robust, cheap and reliable tool for predicting the quality of components produced by metal additive manufacturing (MAM) processes. Despite the advantages of the MAM processes over conventional manufacturing methods, there is still a lack of thorough understanding on how different defects can form and originate during MAM processing. In that respect, advanced numerical techniques recently developed have the ability to predict the occurrence of such defects. These techniques have paved the way to efficiently obtain the optimal processing window for targeted mechanical properties to satisfy the end-use design requirements. The aim of this review paper is hence to present and classify numerical simulations of MAM, not solely based on their length-scale as often seen, but also based on the involved physics, as well as the modeling strategies at both the meso-scale and part-scale. The paper is arranged in the following way: First, literature describing purely conduction-based heat transfer simulations at meso-scale are presented. This is followed by a review of fluid-based simulations of increasing complexity based on the treatment of free surface of the melt pool at meso-scale. Finally, contributions based on different part-scale modeling approaches with a focus on thermo-mechanical behavior are reviewed.