Final Report Summary - SIMCHAIN (Development of physically based simulation chain for microstructure evolution and resulting mechanical properties focused on additive manufacturing processes)
Executive Summary:
Powder bed based additive manufacturing processes belong to the key technologies of the future. They allow the production of complex shaped components from the powder with nearly no waste. To improve the process control and to identify proper process operation-windows, ensuring part integrity and stable mechanical properties, further research on the interactions between the process and the specific material to be processed is required. Here, the SIMCHAIN-project wants to contribute with the development of a new multi-scale material-simulation approach.
The aim of the project is to establish a full software set, which allows the prediction of resulting mechanical properties of materials produced by additive manufacturing using Selective Electron Beam Melting as a function of the process parameters. In order to realize this goal, we couple three simulation modules covering all the essential physical mechanisms on relevant length- and time-scales: The melting, initial grain structure and orientation formation of the powder particles upon electron beam interaction will be simulated via Lattice-Boltzmann and Cellular Automata approaches; the initial microstructure formation during rapid dendritic solidification at micrometer-dendritic arm-spacing length and solidification time-scales will be covered by the phase-field module; the thermo-mechanical behavior of the resulting grain structure at heat-treatment-time-scales will be simulated using a crystal plasticity Finite Element simulation module.
Furthermore, the development of the simulation models will be accompanied by experiments to define essential material parameters and to calibrate, validate and optimize the derived models. SIMCHAIN is an innovative and unique approach to build a ready to use software set in order to predict the influence of various process parameters on the resulting mechanical properties during additive manufacturing using Selective Electron Beam Melting. SIMCHAIN prepares the ground for robust process design, as an important step towards design-driven manufacturing for future aero engines parts optimized in weight and function.
Project Context and Objectives:
The aim of the project SIMCHAIN is to provide a ready to use simulation software set which allows the prediction of mechanical properties of materials produced by additive manufacturing processes as a function of the various sensitive process parameters. Realizing physically based simulations with such kind of predictive power, we couple three well established simulation methods covering all the essential physical mechanisms on all the different relevant length- and time-scales: the melting and the initial grain structure and orientation formation of the powder particles upon laser or electron beam interaction, the initial microstructure formation during rapid dendritic solidification, and the thermo-mechanical behavior of the resulting grain structure of the solidified material at heat-treatment-time-scales. These modeling tasks will be accompanied by a number of appropriate experimental investigations, which are required for the physical understanding, the parameter identification and the individual module validation as well as for the validation of the combined simulation chain. Subsequently, the developed simulation chain will be applied to the Clean Sky demo part, which is the first experimental validation of the full simulation chain. On the basis of these results the individual modules will be optimized. Finally, the provided software set will be able to predict the resulting mechanical properties of manufactured parts.
The identified objectives of the SIMCHAIN-project are summarized as follows
• The theoretical setup of the simulation chain together with the detailed work flow analysis.
• The simulation chains Proof of Principle, which contains the first testing as well as the demonstration of principle applicability to the real physical conditions to be found in Inconel 718 during the SEBM process.
• The generation of the experimental data for calibration and validation of the modules
• The quantitative validation and calibration of the individual modules.
• The simulation chain application to the Clean Sky Demo Part and first software transfer.
• The optimization of the simulation chain with respect to the evaluation of the application to the Clean Sky Demo Part.
• Dissemination of the results and transfer of the final software set.
Project Results:
The aim of the project SIMCHAIN is to provide a ready to use simulation software set which allows the prediction of mechanical properties of materials produced by additive manufacturing processes as a function of the various sensitive process parameters. Realizing physically based simulations with such kind of predictive power, we couple three well established simulation methods covering all the essential physical mechanisms on all the different relevant length- and time-scales: the interaction of electron beam with the powder bed, meltpool dynamics and initial grain structure evolution, the initial microstructure formation during rapid dendritic solidification, and the thermo-mechanical behavior of the resulting grain structure of the solidified material at heat-treatment-time-scales. These modeling tasks will be accompanied by a number of appropriate experimental investigations, which are required for the physical understanding, the parameter identification and the individual module validation as well as for the validation of the combined simulation chain. Subsequently, the developed simulation chain will be applied to the Clean Sky demo part, which is the first experimental validation of the full simulation chain. On the basis of these results the individual modules will be optimized. Finally, the provided software set will be able to predict the resulting mechanical properties of manufactured parts.
The identified objectives of the SIMCHAIN-project are summarized as follows
• The theoretical setup of the simulation chain together with the detailed work flow analysis.
• The simulation chains Proof of Principle, which contains the first testing as well as the demonstration of principle applicability to the real physical conditions to be found in Inconel 718 during the SEBM process.
• The generation of the experimental data for calibration and validation of the modules
• The quantitative validation and calibration of the individual modules.
• The simulation chain application to the Clean Sky Demo Part and first software transfer.
• The optimization of the simulation chain with respect to the evaluation of the application to the Clean Sky Demo Part.
• Dissemination of the results and transfer of the final software set.
1.2 Description of the main S & T results/foregrounds
Please provide a description of the main S & T results/foregrounds. The length of this part cannot exceed 25 pages. The simulation chain: Workflow & Proof of Principle
The aim of the project SIMCHAIN was to establish and to provide a simulation software set, which allows the prediction of resulting mechanical properties of materials produced by the Selective Electron Beam Melting (SEBM) additive manufacturing process as a function of the various sensitive process parameters. The materials-simulation of mechanical properties of SEBM additive manufactured parts involves quit a number of physical aspects and mechanisms on many different length and time scales, which all need to be taken into account in a respective simulation. Therefore, the developed software set consists of three independently executable modules, where each of the modules addresses a different set of relevant physical aspects and mechanisms. All modules interact with each other via well defined interfaces, bridging the mutually different length and time scales.
The setup & workflow of the simulation chain including the interfaces, is illustrated in Fig. 2. The three individual modules are as follows
• The phase-field module for the microscopic simulation of rapid solidification and precipitation in the multi-component nickel-based Superalloy Inconel 718
• The Cellular-Automata-Lattice-Boltzmann (CALB) module for the Mesoscale simulation of the powder-melting and solidification grain structure formation upon the electron beam interaction
• The crystal plasticity module for the simulation-based prediction of the thermo-mechanical properties of the as heat-treated polycrystalline material
These three modules are coupled via the following four scale-bridging interfaces:
• Interface 1: The microscopic solidification tip-velocity as function of the local undercooling temperature calculated by the phase-field module is used for the mesoscale solidification simulation by the CALB-module.
• Interface 2: The solidification grain-structure calculated by the CALB-module is used for the simulation of the thermo-mechanical properties of the respective polycrystal by the crystal plasticity module.
• Interface 3: The microscopic precipitation state after heat-treatment can be integrated in the plastic deformation behavior of individual grains within the mesoscale simulation of crystal plasticity.
• Interface 4: The evolution of the local temperature and temperature gradient calculated by the mesoscale CALB-module can serve as the overall time-dependent temperature condition for the microscopic solidification simulation with the phase-field module.
However, with respect to SIMCHAIN's overall goal, a simulation-bases tailoring of the SEBM-additive manufacturing process, not all of these interfaces are equally important. The core-execution of the simulation chain for that purpose involves only the high priority interfaces 1 and 2. The execution of the core-chain starts with the microscopic simulations of dendritic solidification in Inconel 718 with the aim to provide the dendrite tip-velocity as function of the local undercooling within interface 1. Then based on these velocities the solidification simulation on the scale of the melt-pool, the electron beam and the powder particles is performed in order to obtain the grain structure of the part to be considered. After that the grain structure is transferred to the crystal plasticity simulation within interface 2, in order to predict the respective thermo-mechanical behavior of the respective polycrystal.
The phase field module
The phase-field module is setup as a generic command-line simulation program with an extensive parameter control, which allows the performance of very many different kinds of phase-field simulation configurations within the same program. So far the following physical effects have been implemented and can be freely switched on and of for the respective phase-field simulations to be performed
• Kinetic and capillary anisotropic transformations between two different phases
• temperature dependent chemical multi-component diffusion
• Inhomogeneous and anisotropic elasticity
With regard to the temperature dependent chemical multi-component diffusion, an automatized procedure for the respective generation of the temperature dependent thermodynamic and kinetic multi-component data has been created. The data generation procedure is based on the CALPHAD-methodology as well as on commercially available thermodynamic and kinetic data bases. especially for material-systems with a relatively high number of chemical components, such an automatization still allows the quick switch to different alloying systems. Within SIMCHAIN the module provides the physically based microscopic view onto the processing of the alloy Inconel 718. Within the framework of the simulation chain two microscopic mechanisms have been explicitly considered: The one is the rapid dendritic solidification and the other is the precipitation growth and coarsening during the subsequent heat treatment.
With regard to solidification, the module has been specifically applied to simulate columnar dendritic solidification in multi-component alloys, where one or a few dendritic tips with the same crystallographic orientation can grow. The alloy that has been considered, here is Inconel 718, which has been represented in the simulation with up to 8 explicit chemical components. The solidification is then driven by either a constant temperature below the liquidus temperature (free dendritic growth) or linear gradient temperature field (directional solidification).
With respect to the simulation of precipitation growth/coarsening elasto-mechanical contributions beside the multi-component thermodynamics are needed as well. In order to compute the elastic energy, values for the misfit between the phases and the elastic constants for all the relevant solid phases are required. Within the framework of the SIMCHAIN-project we have discussed the precipitation growth and coarsening of the cubic \gamma'-phase as well as the tetragonal \gamma''-phase. These two phases are the most important precipitation strengthening phases of Inconel 718. However, within the framework of the project it has been only the purpose to demonstrate the principle feasibility of this kind of simulation. Also the subsequent integration of the resulting precipitation state has only been rudimentarily discussed. Nevertheless, with regard to \gamma'-precipitation, we have demonstrated the feasibility of the simulation of a two-step precipitation heat treatment, where the first step takes 8 hours at 718°C and the second step takes another 8 hours 621°C. Similarly, also the principle feasibility of the simulation of \gamma''-precipitation during the heat treatment could be shown.
The CALB-module
The Cellular-Automata-Lattice-Boltzmann (CALB) module has been developed within SIMCHAIN to qualitatively predict the grain structure evolution during Selective Electron Beam Melting (SEBM). The model uses a cellular automata type algorithmic approach. The evolution of the temperature field during SEBM is calculated using a pre–existing Lattice Boltzmann method based tool. The temperature field calculated using the LB model is coupled to the CA model for grain structure simulation. For a given set of specific input process parameters the CALB module can be applied to study their influence on the final grain structure of the as–built part. The model can be applied to any material solidifying in Face Centered Cubic (FCC) structure. Within the SIMCHAIN-project the CALB module has been applied to study the grain structure evolution of IN718. Process parameters during the build process strongly affect the final quality of the SEBM built part, usually measured in term of internal porosity, grain texture and surface roughness, which has a strong influence on the mechanical properties of the part.
The CALB module includes all the physical aspects and mechanisms, which act on the scale of the initial powder particles as well as the resulting solidification grain structure, such as sub–grain structure in powder particles and competitive grain growth laws. Within the SIMCHAIN project the aspect of grain formation and growth has been successfully integrated and is validated at each step, e.g. the basic grain growth model with an analytical model for the growth of a single grain with varying initial orientation subjected to different pre–defined thermal conditions. Competitive grain growth laws have been validated by the simulation of directional solidification. The tool enables the simulation of complex geometries owing to its very flexible beam movement logic, e.g. the cup, bridge and tilted T shaped geometries. Furthermore, Matlab based post–processing tools have been developed to calculate the grain texture from the simulated results. The developed tool is very stable and can work under linux or windows based operating systems.
Data transfer from the grain structure module to the crystal plasticity simulation as well as to the phase field module have been performed successfully. Temperature evolution during the single line melting has been transferred to the phase field simulation module. Export tools to transfer the simulated grain structure in suitable format usable by the crystal plasticity model have been developed and successfully applied.
The developed tool is successfully applied to study the influence of process and model specific input parameters on the resulting grain structure during SEBM. For example many single line melting simulations have been performed to understand the sensitivity of the grain growth model towards the input dendritic tip velocity. Sensitivity studies over four orders of magnitude have been performed. From the single line melting simulation results it can be concluded that the choice of, v_tip = 0.1 v_{pureNi}, gives qualitatively correct grain structure within reasonable undercooling range and assures the faster computation times. A decrease of the input dendritic tip velocity by 20\,\% of the chosen default value does not have a significant influence on the final grain structure. However, an increase in input dendritic tip velocity by 20% with respect to the default value chosen for CALB leads to a finer grain texture. For multi–layer hatching a series of simulations have been performed to study the sensitivity of the final grain structure to (i) powder sub–grain structure (ii) input dendritic tip velocity (20% of the chosen default value) (iii) scanning strategy and (iv) beam return time.
For a well built part without bulk porosity inclusion of the powder sub–grain structure mainly influences the grain density at the side walls. Whereas if the chosen process parameters leads to a porous bulk region many new grains are formed within the bulk which interrupts the epitaxial growth of the grains originating from the baseplate. Depending on the geometry of the built part, wherever the solidification starts directly at the powder bed, e.g. the bottom of the horizontal section and the walls of the T–part, a higher grain density is observed. This may play a very significant role, when thin walled or cellular structures are built using SEBM.
The scanning strategy has a major influence on the final grain morphology and texture. In case of leftwards and rightwards only scanning by the beam the final grain boundaries are also tilted leftwards and rightwards, respectively. Whereas for symmetric hatching the grain boundaries in the final grain structure are aligned to the build direction. In case of a lateral shift of the scanning contours in each layer a much finer texture is observed compared to the case when the scanning contours lie exactly on top of each other at every layer.
A longer beam return time contributes to a higher spatial scale of grain boundary wiggling and broader texture. This is due to the fact that the final grain structure results from the solidification of the individual non–overlapping melt-pools. In such cases the contribution of the grains with higher misorientations from the side of the melt-pools reduces the overall texture of the sample. Whereas, for shorter beam return times consecutive melt-pools overlap and the final grain structure results from solidification of longer and flatter melt-pools, which also reduces the relative contribution of the grains from the sides of the melt-pool on the overall texture of the built part.
In light of the discussions presented above, it can be concluded that the CALB module developed within the SIMCHAIN project is a very valuable tool. At this development stage, it gives very good qualitative guidelines regarding the final grain structure in response to a change in the process parameters. However, there are still certain aspects of the grain growth, which are not yet included into the CALB model and affect the final predictive power of the entire simulation chain as a whole. Presence of gas porosity in the powder particles is neither included in the grain structure simulation module nor in the crystal plasticity module. The CALB tool is able to simulate epitaxial growth and the process of new grain nucleation occurs via partially molten powder particles only. For certain settings of the SEBM process parameters, it was also possible to produce equiaxed grain structures in build parts made of IN718 (see D6.1). Nevertheless, simulation results using the same process parameters as used in these experiments showed mainly columnar grain growth. Based on this outcome it is our speculation that the inclusion of a grain nucleation model should be an extremely important extension of the model for simulating equiaxed grains.
The crystal plasticity module
The aim of the crystal plasticity calculations is to predict the final thermo-mechanical behavior of the polycrystalline material. The module is implemented as user subroutine UMAT for the implicit finite element (FE) code Abaqus standard. The module is located at the end of the simulation chain. Therefore, the interfaces to the other modules define the input data for the simulations. One main input for the calculations is the grain-structure information of the material after solidification, i.e. interface 2. That includes morphology and texture information. With this data a statistically equivalent 3D grain structure has been build up for the actual crystal plasticity simulations. Of course, regarding the transfer from the grain growth simulation additional assumptions on the orientation in space have to be made. These assumptions for the 2D to 3D transfer of the orientation has a strong influence on the predicted mechanical properties. Nevertheless, there is a small range for the random values for the respective grain orientation angles, which should be used for grain structures with clearly columnar character. However, due to the small number of grains within the simulation domain the random value for each grain has a big influence. Therefore, a smaller range for the two Euler angles also leads to less variation of the results, when two sets of random values are used for the simulation.
Another input for the crystal plasticity simulation can be the precipitation state after the heat treatment. The information about size and distribution of precipitates can be also be provided by a respective microscopic phase field simulation, indicated by the interface 3 above. More precisely, the temperature-independent reference of the initial strength on the slip system can defined as a function of the precipitation state. However, here, the parameter describing the plastic behavior were calibrated by inverse simulation of tensile experiments in 0°-direction at room temperature and 650°C.
The output of the crystal plasticity simulation is the thermo-mechanical characteristics of the investigated material, such as the stress strain relation, the yield stress and the anisotropic elastic behavior. As the grain structure our SEBM fabricated samples typically showed a largely anisotropic crystal structure consisting of columnar shaped grains that are strongly elongated along the built direction, the resulting mechanical behavior is correspondingly anisotropic. The anisotropy of the mechanical properties was investigated for the respective columnar crystal structure. The elastic properties of pure Nickel single crystals were used to describe the elastic behavior. That leads to a macroscopic modulus of elasticity of 137 GPa in build-up direction and of 196 GPa in the transverse direction. Figure 5 shows exemplarily such crystal plasticity simulations as well as the resulting anisotropy in the stress strain curve.
The comparison between experimental and simulation results showed a good match of the anisotropy and the decrease of yield stress with temperature. For the investigated temperatures as well as the considered different orientations the thermo-mechanical behavior is well captured by the approach. Therefore, it can be noted that the approach is able to predict the elastic and plastic behavior at different temperatures for the three different directions of tensile specimens in good agreement with experimental results.
The Clean Sky Demo Part & respective optimizations of the simulation chain
In order to test the actual predictive power of the simulation chain as well as to identify weak points for the subsequent optimization, the so-called Clean Sky Demo Part has been setup within the SIMCHAIN-project. This part has been built in reality using SEBM additive manufacturing as well as virtually by a respective application of the simulation chain. The actual shape of the Clean Sky Demo Part has been defined in conjunction with the Topic Manager. It has been decided to build a rotated T-piece with the main arm to be rotated by an angle of 30° with respect to the building direction. Both main arm and roof will be of cylindrical shape in diameter (see Fig. 6).
Then, the virtual grain structure as well as the simulation-based predictions of the mechanical properties have been compared to the ones from the real part. On the basis of the performance of this comparison with real-life complexity, we could identify the necessary and most beneficial actions for the further optimization of the individual modules as well as the simulation chain as a whole. These actions have been in particular
1. A detailed review of the underlying scaling of the phase-field solidification simulation as well as a respective discussion of the received solidification velocities has be performed.
2. Performance of the pending phase-field simulations of the precipitation during the heat treatment.
3. The grain structure simulation module has be improved with regard to the control of the beam-logic.
4. It is interesting to analyze powder particles with a sub-grain structure and their influence on the final grain structure.
5. Adding a grain identification number as another column for the next grain-structure transfer is helpful for the crystal plasticity module.
6. Additionally provide a dislocation density based crystal plasticity formulation.
After the respective identification, it was of course also an important objective of the project to perform these actions and optimize the simulation chain on the basis of these findings. With regard to the phase-field module the focus has been on the optimized work flow of how the two different kinds of phase-field simulations, dendritic solidification as well as precipitation both in Inconel 718, are properly scaled. For the final scaling-concept within the phase-field module, we have suggested a fitting to the respective length-scales, which are experimentally accessible. For the respective calibration of the simulation of solidification, such a length-scale is the primary dendrite arm-spacing. Further, the precipitation simulation has been performed using one and the same phase-field code. Moreover, we have shown that a scaling of the respective simulation to typical particle diameters, as observed in the experiment, is in principle feasible. The grain structure simulation tool as individual module has been extended by the following two new features:
1. Flexible beam logic to simulate arbitrary geometries.
2. Sub-grain structure in powder particles.
The functionality of a flexible beam logic has been demonstrated by simulating a cup shaped part and a bridge shaped part. The influence of the powder particles sub-grain structure on the grain structure evolution has been illustrated by means of two respective simulations of a straight T-part. For the optimization of the crystal plasticity module the hardening description in the crystal plasticity model was adapted to incorporate dislocation densities and an initial strength on the slip system to account for other strengthening mechanisms as the presence of precipitates. The model was formulated temperature dependent and includes five parameters to model the thermo-elastic behavior and four parameters to describe the temperature dependent hardening. The respective calibration revealed that the four investigated temperatures and the three different orientations the thermo-mechanical behavior is well captured by the approach. Therefore, the approach is able to predict the plastic behavior at different temperatures and for different directions.
Concerning the optimization/improvement of the module-interfaces, the following actions have been performed during the projects optimization phase.
• Interface 1 (the transfer of the solidification velocity as function of the local undercooling): The provision of the additional functionality to scale the velocity-temperature function to absolute as well as relative undercooling temperatures
• Interface 2 (the transfer of the final grain structure to the subsequent crystal plasticity simulation): The extension of the grain-structure-data-format by the grain identification number (optimization action 5)
• Interface 2 (the transfer of the final grain structure to the subsequent crystal plasticity simulation): The provision of the functionality to choose the „region of interest“ within the provided grain structure data, which allows to study local mechanical properties as well.
Potential Impact:
The manufacturing sector is essential for economic prosperity and growth. Part of this sector is now undergoing a paradigm change. Component design driven by fabrication limitations changes to component design driven by functionality. Additive manufacturing is the key to accomplish this fundamental change and has a strong impact on the method of production of some companies. It boosts innovation and lays the technological basis for new products where an optimum component design reduces weight. Last but not least, additive manufacturing shortens the time to market. Especially additive manufacturing of products based on high performance metal alloys is a highly promising area where Europe has already a strong foundation. Nevertheless, much of the turnover is still in the field of rapid prototyping since reproducibility and part quality is yet often not sufficient. SIMCHAIN will lead to a reproducible and reliable prediction of part quality which has a high impact for additive manufacturing. This will help to accomplish the change from rapid prototyping to manufacturing of fully operational parts. Certainly, SIMCHAIN will strongly help to increase the growth rates of additive manufacturing. In addition, it will give strong support to break into completely new markets in the field of engineering. Furthermore, new products and components with optimized performance due to the design freedom of additive manufacturing will emerge. The design of these new products is not driven by manufacturing limitations but by performance. This will help to reduce weight and fuel in aerospace applications. Thus, SIMCHAIN contributes significantly to reduce the environmental footprint of aviation.
Although this project is focussed on aeronautical parts, there is a number of other markets for high performance metallic components, e.g. engine parts, sports automotive, power generation, food processing and chemical plant, where SIMCHAIN will have a strong impact.
Sustainability, preservation of raw materials and Reduction of the Scrap rate Today, one of the important issues in metallurgy is sustainability and the preservation of raw materials. Especially high performance metallic alloys contain a high amount of critical metals which Europe must preserve and have access to – Co, Ti, Nb, Ta, Pt, rare earth metals, refractory metals. Conventionally, high performance components are manufactured through machining of forged pieces or through machining from plate where most of the metal is wasted. Very often, less than 10 % of the purchased metal is in the final product and more than 90% is machined away. During additive manufacturing, more than 70% of the purchased raw material is used, which may reduce the total energy demands for recycling of raw materials. Furthermore, free material optimization modeling allows new and unique 3D designs, which lead to further chances in material and weight saving of parts.
Nevertheless, the scrap rate for parts produced by additive manufacturing is still rather high (> 25 – 30 %). The reasons for this high scrap rate are manifold. Besides problems already occurring during the assembly process a variety of other defects is observed. In order to reduce the scrap rate a scientific understanding of the types of defects and their origins has to be developed. SIMCHAIN will strongly increase the scientific understanding and will provide numerical tools to predict faults on a sound experimental basis. SIMCHAIN will bridge the length scale from the macroscopic scale of the component to the local microstructure. It will strongly increase the understanding of the correlation between the material and process parameters, the meso- and microstructure and the resulting material properties. Thus, SIMCHAIN will strongly contribute to sustainability and preservation of raw materials.
In a first step towards these potential impacts, major results of the single simulation modules in SIMCHAIN are already disseminated. Investigations on the predictions of the equilibrium and metastable phase/grain morphologies of Ni-base superalloys using the phase-field modeling developed during SIMCHAIN are exploited. A phenomenological dependency of the growth velocity on undercooling during the rapid solidification is evaluated. The predicted microstructure is in good agreement with experimental observations. The developed model during SIMCHAIN is validated against experimental findings of single track melting and solidification of a baseplate. The model is applied to simulate the grain structure produced during additive manufacturing. The major influence of the hatching strategy on the grain structure as well as stray grain formation resulting from partially molten powder particles is found. The columnar grains grow epitaxially from the baseplate. Many morphological characteristics observed experimentally, e.g. grain penetration from side walls and grain boundary wiggling are well reproduced by the model. The simulated grains structure and texture is in good qualitative agreement with the experimental observations. The evolution of the grain structure as a function of build height follows the classical rule of grain selection. Further SIMCHAIN results regarding the coupling of the simulation chain modules and the SIMCHAIN Demo Part will be disseminated.
List of Websites:
The projects website:
http://www.cleansky.uni-bayreuth.de
• University of Bayreuth, Germany
Prof. Dr. H. Emmerich
heike.emmerich@uni-bayreuth.de
• Friedrich Alexander Universität Erlangen-Nürnberg, Germany
Prof. Dr. C. Körner
carolin.koerner@ww.uni-erlangen.de
• Fraunhofer Institute for Mechanics of Materials, Freiburg, Germany
Dr. D. Helm
dirk.helm@iwm.fraunhofer.de
Powder bed based additive manufacturing processes belong to the key technologies of the future. They allow the production of complex shaped components from the powder with nearly no waste. To improve the process control and to identify proper process operation-windows, ensuring part integrity and stable mechanical properties, further research on the interactions between the process and the specific material to be processed is required. Here, the SIMCHAIN-project wants to contribute with the development of a new multi-scale material-simulation approach.
The aim of the project is to establish a full software set, which allows the prediction of resulting mechanical properties of materials produced by additive manufacturing using Selective Electron Beam Melting as a function of the process parameters. In order to realize this goal, we couple three simulation modules covering all the essential physical mechanisms on relevant length- and time-scales: The melting, initial grain structure and orientation formation of the powder particles upon electron beam interaction will be simulated via Lattice-Boltzmann and Cellular Automata approaches; the initial microstructure formation during rapid dendritic solidification at micrometer-dendritic arm-spacing length and solidification time-scales will be covered by the phase-field module; the thermo-mechanical behavior of the resulting grain structure at heat-treatment-time-scales will be simulated using a crystal plasticity Finite Element simulation module.
Furthermore, the development of the simulation models will be accompanied by experiments to define essential material parameters and to calibrate, validate and optimize the derived models. SIMCHAIN is an innovative and unique approach to build a ready to use software set in order to predict the influence of various process parameters on the resulting mechanical properties during additive manufacturing using Selective Electron Beam Melting. SIMCHAIN prepares the ground for robust process design, as an important step towards design-driven manufacturing for future aero engines parts optimized in weight and function.
Project Context and Objectives:
The aim of the project SIMCHAIN is to provide a ready to use simulation software set which allows the prediction of mechanical properties of materials produced by additive manufacturing processes as a function of the various sensitive process parameters. Realizing physically based simulations with such kind of predictive power, we couple three well established simulation methods covering all the essential physical mechanisms on all the different relevant length- and time-scales: the melting and the initial grain structure and orientation formation of the powder particles upon laser or electron beam interaction, the initial microstructure formation during rapid dendritic solidification, and the thermo-mechanical behavior of the resulting grain structure of the solidified material at heat-treatment-time-scales. These modeling tasks will be accompanied by a number of appropriate experimental investigations, which are required for the physical understanding, the parameter identification and the individual module validation as well as for the validation of the combined simulation chain. Subsequently, the developed simulation chain will be applied to the Clean Sky demo part, which is the first experimental validation of the full simulation chain. On the basis of these results the individual modules will be optimized. Finally, the provided software set will be able to predict the resulting mechanical properties of manufactured parts.
The identified objectives of the SIMCHAIN-project are summarized as follows
• The theoretical setup of the simulation chain together with the detailed work flow analysis.
• The simulation chains Proof of Principle, which contains the first testing as well as the demonstration of principle applicability to the real physical conditions to be found in Inconel 718 during the SEBM process.
• The generation of the experimental data for calibration and validation of the modules
• The quantitative validation and calibration of the individual modules.
• The simulation chain application to the Clean Sky Demo Part and first software transfer.
• The optimization of the simulation chain with respect to the evaluation of the application to the Clean Sky Demo Part.
• Dissemination of the results and transfer of the final software set.
Project Results:
The aim of the project SIMCHAIN is to provide a ready to use simulation software set which allows the prediction of mechanical properties of materials produced by additive manufacturing processes as a function of the various sensitive process parameters. Realizing physically based simulations with such kind of predictive power, we couple three well established simulation methods covering all the essential physical mechanisms on all the different relevant length- and time-scales: the interaction of electron beam with the powder bed, meltpool dynamics and initial grain structure evolution, the initial microstructure formation during rapid dendritic solidification, and the thermo-mechanical behavior of the resulting grain structure of the solidified material at heat-treatment-time-scales. These modeling tasks will be accompanied by a number of appropriate experimental investigations, which are required for the physical understanding, the parameter identification and the individual module validation as well as for the validation of the combined simulation chain. Subsequently, the developed simulation chain will be applied to the Clean Sky demo part, which is the first experimental validation of the full simulation chain. On the basis of these results the individual modules will be optimized. Finally, the provided software set will be able to predict the resulting mechanical properties of manufactured parts.
The identified objectives of the SIMCHAIN-project are summarized as follows
• The theoretical setup of the simulation chain together with the detailed work flow analysis.
• The simulation chains Proof of Principle, which contains the first testing as well as the demonstration of principle applicability to the real physical conditions to be found in Inconel 718 during the SEBM process.
• The generation of the experimental data for calibration and validation of the modules
• The quantitative validation and calibration of the individual modules.
• The simulation chain application to the Clean Sky Demo Part and first software transfer.
• The optimization of the simulation chain with respect to the evaluation of the application to the Clean Sky Demo Part.
• Dissemination of the results and transfer of the final software set.
1.2 Description of the main S & T results/foregrounds
Please provide a description of the main S & T results/foregrounds. The length of this part cannot exceed 25 pages. The simulation chain: Workflow & Proof of Principle
The aim of the project SIMCHAIN was to establish and to provide a simulation software set, which allows the prediction of resulting mechanical properties of materials produced by the Selective Electron Beam Melting (SEBM) additive manufacturing process as a function of the various sensitive process parameters. The materials-simulation of mechanical properties of SEBM additive manufactured parts involves quit a number of physical aspects and mechanisms on many different length and time scales, which all need to be taken into account in a respective simulation. Therefore, the developed software set consists of three independently executable modules, where each of the modules addresses a different set of relevant physical aspects and mechanisms. All modules interact with each other via well defined interfaces, bridging the mutually different length and time scales.
The setup & workflow of the simulation chain including the interfaces, is illustrated in Fig. 2. The three individual modules are as follows
• The phase-field module for the microscopic simulation of rapid solidification and precipitation in the multi-component nickel-based Superalloy Inconel 718
• The Cellular-Automata-Lattice-Boltzmann (CALB) module for the Mesoscale simulation of the powder-melting and solidification grain structure formation upon the electron beam interaction
• The crystal plasticity module for the simulation-based prediction of the thermo-mechanical properties of the as heat-treated polycrystalline material
These three modules are coupled via the following four scale-bridging interfaces:
• Interface 1: The microscopic solidification tip-velocity as function of the local undercooling temperature calculated by the phase-field module is used for the mesoscale solidification simulation by the CALB-module.
• Interface 2: The solidification grain-structure calculated by the CALB-module is used for the simulation of the thermo-mechanical properties of the respective polycrystal by the crystal plasticity module.
• Interface 3: The microscopic precipitation state after heat-treatment can be integrated in the plastic deformation behavior of individual grains within the mesoscale simulation of crystal plasticity.
• Interface 4: The evolution of the local temperature and temperature gradient calculated by the mesoscale CALB-module can serve as the overall time-dependent temperature condition for the microscopic solidification simulation with the phase-field module.
However, with respect to SIMCHAIN's overall goal, a simulation-bases tailoring of the SEBM-additive manufacturing process, not all of these interfaces are equally important. The core-execution of the simulation chain for that purpose involves only the high priority interfaces 1 and 2. The execution of the core-chain starts with the microscopic simulations of dendritic solidification in Inconel 718 with the aim to provide the dendrite tip-velocity as function of the local undercooling within interface 1. Then based on these velocities the solidification simulation on the scale of the melt-pool, the electron beam and the powder particles is performed in order to obtain the grain structure of the part to be considered. After that the grain structure is transferred to the crystal plasticity simulation within interface 2, in order to predict the respective thermo-mechanical behavior of the respective polycrystal.
The phase field module
The phase-field module is setup as a generic command-line simulation program with an extensive parameter control, which allows the performance of very many different kinds of phase-field simulation configurations within the same program. So far the following physical effects have been implemented and can be freely switched on and of for the respective phase-field simulations to be performed
• Kinetic and capillary anisotropic transformations between two different phases
• temperature dependent chemical multi-component diffusion
• Inhomogeneous and anisotropic elasticity
With regard to the temperature dependent chemical multi-component diffusion, an automatized procedure for the respective generation of the temperature dependent thermodynamic and kinetic multi-component data has been created. The data generation procedure is based on the CALPHAD-methodology as well as on commercially available thermodynamic and kinetic data bases. especially for material-systems with a relatively high number of chemical components, such an automatization still allows the quick switch to different alloying systems. Within SIMCHAIN the module provides the physically based microscopic view onto the processing of the alloy Inconel 718. Within the framework of the simulation chain two microscopic mechanisms have been explicitly considered: The one is the rapid dendritic solidification and the other is the precipitation growth and coarsening during the subsequent heat treatment.
With regard to solidification, the module has been specifically applied to simulate columnar dendritic solidification in multi-component alloys, where one or a few dendritic tips with the same crystallographic orientation can grow. The alloy that has been considered, here is Inconel 718, which has been represented in the simulation with up to 8 explicit chemical components. The solidification is then driven by either a constant temperature below the liquidus temperature (free dendritic growth) or linear gradient temperature field (directional solidification).
With respect to the simulation of precipitation growth/coarsening elasto-mechanical contributions beside the multi-component thermodynamics are needed as well. In order to compute the elastic energy, values for the misfit between the phases and the elastic constants for all the relevant solid phases are required. Within the framework of the SIMCHAIN-project we have discussed the precipitation growth and coarsening of the cubic \gamma'-phase as well as the tetragonal \gamma''-phase. These two phases are the most important precipitation strengthening phases of Inconel 718. However, within the framework of the project it has been only the purpose to demonstrate the principle feasibility of this kind of simulation. Also the subsequent integration of the resulting precipitation state has only been rudimentarily discussed. Nevertheless, with regard to \gamma'-precipitation, we have demonstrated the feasibility of the simulation of a two-step precipitation heat treatment, where the first step takes 8 hours at 718°C and the second step takes another 8 hours 621°C. Similarly, also the principle feasibility of the simulation of \gamma''-precipitation during the heat treatment could be shown.
The CALB-module
The Cellular-Automata-Lattice-Boltzmann (CALB) module has been developed within SIMCHAIN to qualitatively predict the grain structure evolution during Selective Electron Beam Melting (SEBM). The model uses a cellular automata type algorithmic approach. The evolution of the temperature field during SEBM is calculated using a pre–existing Lattice Boltzmann method based tool. The temperature field calculated using the LB model is coupled to the CA model for grain structure simulation. For a given set of specific input process parameters the CALB module can be applied to study their influence on the final grain structure of the as–built part. The model can be applied to any material solidifying in Face Centered Cubic (FCC) structure. Within the SIMCHAIN-project the CALB module has been applied to study the grain structure evolution of IN718. Process parameters during the build process strongly affect the final quality of the SEBM built part, usually measured in term of internal porosity, grain texture and surface roughness, which has a strong influence on the mechanical properties of the part.
The CALB module includes all the physical aspects and mechanisms, which act on the scale of the initial powder particles as well as the resulting solidification grain structure, such as sub–grain structure in powder particles and competitive grain growth laws. Within the SIMCHAIN project the aspect of grain formation and growth has been successfully integrated and is validated at each step, e.g. the basic grain growth model with an analytical model for the growth of a single grain with varying initial orientation subjected to different pre–defined thermal conditions. Competitive grain growth laws have been validated by the simulation of directional solidification. The tool enables the simulation of complex geometries owing to its very flexible beam movement logic, e.g. the cup, bridge and tilted T shaped geometries. Furthermore, Matlab based post–processing tools have been developed to calculate the grain texture from the simulated results. The developed tool is very stable and can work under linux or windows based operating systems.
Data transfer from the grain structure module to the crystal plasticity simulation as well as to the phase field module have been performed successfully. Temperature evolution during the single line melting has been transferred to the phase field simulation module. Export tools to transfer the simulated grain structure in suitable format usable by the crystal plasticity model have been developed and successfully applied.
The developed tool is successfully applied to study the influence of process and model specific input parameters on the resulting grain structure during SEBM. For example many single line melting simulations have been performed to understand the sensitivity of the grain growth model towards the input dendritic tip velocity. Sensitivity studies over four orders of magnitude have been performed. From the single line melting simulation results it can be concluded that the choice of, v_tip = 0.1 v_{pureNi}, gives qualitatively correct grain structure within reasonable undercooling range and assures the faster computation times. A decrease of the input dendritic tip velocity by 20\,\% of the chosen default value does not have a significant influence on the final grain structure. However, an increase in input dendritic tip velocity by 20% with respect to the default value chosen for CALB leads to a finer grain texture. For multi–layer hatching a series of simulations have been performed to study the sensitivity of the final grain structure to (i) powder sub–grain structure (ii) input dendritic tip velocity (20% of the chosen default value) (iii) scanning strategy and (iv) beam return time.
For a well built part without bulk porosity inclusion of the powder sub–grain structure mainly influences the grain density at the side walls. Whereas if the chosen process parameters leads to a porous bulk region many new grains are formed within the bulk which interrupts the epitaxial growth of the grains originating from the baseplate. Depending on the geometry of the built part, wherever the solidification starts directly at the powder bed, e.g. the bottom of the horizontal section and the walls of the T–part, a higher grain density is observed. This may play a very significant role, when thin walled or cellular structures are built using SEBM.
The scanning strategy has a major influence on the final grain morphology and texture. In case of leftwards and rightwards only scanning by the beam the final grain boundaries are also tilted leftwards and rightwards, respectively. Whereas for symmetric hatching the grain boundaries in the final grain structure are aligned to the build direction. In case of a lateral shift of the scanning contours in each layer a much finer texture is observed compared to the case when the scanning contours lie exactly on top of each other at every layer.
A longer beam return time contributes to a higher spatial scale of grain boundary wiggling and broader texture. This is due to the fact that the final grain structure results from the solidification of the individual non–overlapping melt-pools. In such cases the contribution of the grains with higher misorientations from the side of the melt-pools reduces the overall texture of the sample. Whereas, for shorter beam return times consecutive melt-pools overlap and the final grain structure results from solidification of longer and flatter melt-pools, which also reduces the relative contribution of the grains from the sides of the melt-pool on the overall texture of the built part.
In light of the discussions presented above, it can be concluded that the CALB module developed within the SIMCHAIN project is a very valuable tool. At this development stage, it gives very good qualitative guidelines regarding the final grain structure in response to a change in the process parameters. However, there are still certain aspects of the grain growth, which are not yet included into the CALB model and affect the final predictive power of the entire simulation chain as a whole. Presence of gas porosity in the powder particles is neither included in the grain structure simulation module nor in the crystal plasticity module. The CALB tool is able to simulate epitaxial growth and the process of new grain nucleation occurs via partially molten powder particles only. For certain settings of the SEBM process parameters, it was also possible to produce equiaxed grain structures in build parts made of IN718 (see D6.1). Nevertheless, simulation results using the same process parameters as used in these experiments showed mainly columnar grain growth. Based on this outcome it is our speculation that the inclusion of a grain nucleation model should be an extremely important extension of the model for simulating equiaxed grains.
The crystal plasticity module
The aim of the crystal plasticity calculations is to predict the final thermo-mechanical behavior of the polycrystalline material. The module is implemented as user subroutine UMAT for the implicit finite element (FE) code Abaqus standard. The module is located at the end of the simulation chain. Therefore, the interfaces to the other modules define the input data for the simulations. One main input for the calculations is the grain-structure information of the material after solidification, i.e. interface 2. That includes morphology and texture information. With this data a statistically equivalent 3D grain structure has been build up for the actual crystal plasticity simulations. Of course, regarding the transfer from the grain growth simulation additional assumptions on the orientation in space have to be made. These assumptions for the 2D to 3D transfer of the orientation has a strong influence on the predicted mechanical properties. Nevertheless, there is a small range for the random values for the respective grain orientation angles, which should be used for grain structures with clearly columnar character. However, due to the small number of grains within the simulation domain the random value for each grain has a big influence. Therefore, a smaller range for the two Euler angles also leads to less variation of the results, when two sets of random values are used for the simulation.
Another input for the crystal plasticity simulation can be the precipitation state after the heat treatment. The information about size and distribution of precipitates can be also be provided by a respective microscopic phase field simulation, indicated by the interface 3 above. More precisely, the temperature-independent reference of the initial strength on the slip system can defined as a function of the precipitation state. However, here, the parameter describing the plastic behavior were calibrated by inverse simulation of tensile experiments in 0°-direction at room temperature and 650°C.
The output of the crystal plasticity simulation is the thermo-mechanical characteristics of the investigated material, such as the stress strain relation, the yield stress and the anisotropic elastic behavior. As the grain structure our SEBM fabricated samples typically showed a largely anisotropic crystal structure consisting of columnar shaped grains that are strongly elongated along the built direction, the resulting mechanical behavior is correspondingly anisotropic. The anisotropy of the mechanical properties was investigated for the respective columnar crystal structure. The elastic properties of pure Nickel single crystals were used to describe the elastic behavior. That leads to a macroscopic modulus of elasticity of 137 GPa in build-up direction and of 196 GPa in the transverse direction. Figure 5 shows exemplarily such crystal plasticity simulations as well as the resulting anisotropy in the stress strain curve.
The comparison between experimental and simulation results showed a good match of the anisotropy and the decrease of yield stress with temperature. For the investigated temperatures as well as the considered different orientations the thermo-mechanical behavior is well captured by the approach. Therefore, it can be noted that the approach is able to predict the elastic and plastic behavior at different temperatures for the three different directions of tensile specimens in good agreement with experimental results.
The Clean Sky Demo Part & respective optimizations of the simulation chain
In order to test the actual predictive power of the simulation chain as well as to identify weak points for the subsequent optimization, the so-called Clean Sky Demo Part has been setup within the SIMCHAIN-project. This part has been built in reality using SEBM additive manufacturing as well as virtually by a respective application of the simulation chain. The actual shape of the Clean Sky Demo Part has been defined in conjunction with the Topic Manager. It has been decided to build a rotated T-piece with the main arm to be rotated by an angle of 30° with respect to the building direction. Both main arm and roof will be of cylindrical shape in diameter (see Fig. 6).
Then, the virtual grain structure as well as the simulation-based predictions of the mechanical properties have been compared to the ones from the real part. On the basis of the performance of this comparison with real-life complexity, we could identify the necessary and most beneficial actions for the further optimization of the individual modules as well as the simulation chain as a whole. These actions have been in particular
1. A detailed review of the underlying scaling of the phase-field solidification simulation as well as a respective discussion of the received solidification velocities has be performed.
2. Performance of the pending phase-field simulations of the precipitation during the heat treatment.
3. The grain structure simulation module has be improved with regard to the control of the beam-logic.
4. It is interesting to analyze powder particles with a sub-grain structure and their influence on the final grain structure.
5. Adding a grain identification number as another column for the next grain-structure transfer is helpful for the crystal plasticity module.
6. Additionally provide a dislocation density based crystal plasticity formulation.
After the respective identification, it was of course also an important objective of the project to perform these actions and optimize the simulation chain on the basis of these findings. With regard to the phase-field module the focus has been on the optimized work flow of how the two different kinds of phase-field simulations, dendritic solidification as well as precipitation both in Inconel 718, are properly scaled. For the final scaling-concept within the phase-field module, we have suggested a fitting to the respective length-scales, which are experimentally accessible. For the respective calibration of the simulation of solidification, such a length-scale is the primary dendrite arm-spacing. Further, the precipitation simulation has been performed using one and the same phase-field code. Moreover, we have shown that a scaling of the respective simulation to typical particle diameters, as observed in the experiment, is in principle feasible. The grain structure simulation tool as individual module has been extended by the following two new features:
1. Flexible beam logic to simulate arbitrary geometries.
2. Sub-grain structure in powder particles.
The functionality of a flexible beam logic has been demonstrated by simulating a cup shaped part and a bridge shaped part. The influence of the powder particles sub-grain structure on the grain structure evolution has been illustrated by means of two respective simulations of a straight T-part. For the optimization of the crystal plasticity module the hardening description in the crystal plasticity model was adapted to incorporate dislocation densities and an initial strength on the slip system to account for other strengthening mechanisms as the presence of precipitates. The model was formulated temperature dependent and includes five parameters to model the thermo-elastic behavior and four parameters to describe the temperature dependent hardening. The respective calibration revealed that the four investigated temperatures and the three different orientations the thermo-mechanical behavior is well captured by the approach. Therefore, the approach is able to predict the plastic behavior at different temperatures and for different directions.
Concerning the optimization/improvement of the module-interfaces, the following actions have been performed during the projects optimization phase.
• Interface 1 (the transfer of the solidification velocity as function of the local undercooling): The provision of the additional functionality to scale the velocity-temperature function to absolute as well as relative undercooling temperatures
• Interface 2 (the transfer of the final grain structure to the subsequent crystal plasticity simulation): The extension of the grain-structure-data-format by the grain identification number (optimization action 5)
• Interface 2 (the transfer of the final grain structure to the subsequent crystal plasticity simulation): The provision of the functionality to choose the „region of interest“ within the provided grain structure data, which allows to study local mechanical properties as well.
Potential Impact:
The manufacturing sector is essential for economic prosperity and growth. Part of this sector is now undergoing a paradigm change. Component design driven by fabrication limitations changes to component design driven by functionality. Additive manufacturing is the key to accomplish this fundamental change and has a strong impact on the method of production of some companies. It boosts innovation and lays the technological basis for new products where an optimum component design reduces weight. Last but not least, additive manufacturing shortens the time to market. Especially additive manufacturing of products based on high performance metal alloys is a highly promising area where Europe has already a strong foundation. Nevertheless, much of the turnover is still in the field of rapid prototyping since reproducibility and part quality is yet often not sufficient. SIMCHAIN will lead to a reproducible and reliable prediction of part quality which has a high impact for additive manufacturing. This will help to accomplish the change from rapid prototyping to manufacturing of fully operational parts. Certainly, SIMCHAIN will strongly help to increase the growth rates of additive manufacturing. In addition, it will give strong support to break into completely new markets in the field of engineering. Furthermore, new products and components with optimized performance due to the design freedom of additive manufacturing will emerge. The design of these new products is not driven by manufacturing limitations but by performance. This will help to reduce weight and fuel in aerospace applications. Thus, SIMCHAIN contributes significantly to reduce the environmental footprint of aviation.
Although this project is focussed on aeronautical parts, there is a number of other markets for high performance metallic components, e.g. engine parts, sports automotive, power generation, food processing and chemical plant, where SIMCHAIN will have a strong impact.
Sustainability, preservation of raw materials and Reduction of the Scrap rate Today, one of the important issues in metallurgy is sustainability and the preservation of raw materials. Especially high performance metallic alloys contain a high amount of critical metals which Europe must preserve and have access to – Co, Ti, Nb, Ta, Pt, rare earth metals, refractory metals. Conventionally, high performance components are manufactured through machining of forged pieces or through machining from plate where most of the metal is wasted. Very often, less than 10 % of the purchased metal is in the final product and more than 90% is machined away. During additive manufacturing, more than 70% of the purchased raw material is used, which may reduce the total energy demands for recycling of raw materials. Furthermore, free material optimization modeling allows new and unique 3D designs, which lead to further chances in material and weight saving of parts.
Nevertheless, the scrap rate for parts produced by additive manufacturing is still rather high (> 25 – 30 %). The reasons for this high scrap rate are manifold. Besides problems already occurring during the assembly process a variety of other defects is observed. In order to reduce the scrap rate a scientific understanding of the types of defects and their origins has to be developed. SIMCHAIN will strongly increase the scientific understanding and will provide numerical tools to predict faults on a sound experimental basis. SIMCHAIN will bridge the length scale from the macroscopic scale of the component to the local microstructure. It will strongly increase the understanding of the correlation between the material and process parameters, the meso- and microstructure and the resulting material properties. Thus, SIMCHAIN will strongly contribute to sustainability and preservation of raw materials.
In a first step towards these potential impacts, major results of the single simulation modules in SIMCHAIN are already disseminated. Investigations on the predictions of the equilibrium and metastable phase/grain morphologies of Ni-base superalloys using the phase-field modeling developed during SIMCHAIN are exploited. A phenomenological dependency of the growth velocity on undercooling during the rapid solidification is evaluated. The predicted microstructure is in good agreement with experimental observations. The developed model during SIMCHAIN is validated against experimental findings of single track melting and solidification of a baseplate. The model is applied to simulate the grain structure produced during additive manufacturing. The major influence of the hatching strategy on the grain structure as well as stray grain formation resulting from partially molten powder particles is found. The columnar grains grow epitaxially from the baseplate. Many morphological characteristics observed experimentally, e.g. grain penetration from side walls and grain boundary wiggling are well reproduced by the model. The simulated grains structure and texture is in good qualitative agreement with the experimental observations. The evolution of the grain structure as a function of build height follows the classical rule of grain selection. Further SIMCHAIN results regarding the coupling of the simulation chain modules and the SIMCHAIN Demo Part will be disseminated.
List of Websites:
The projects website:
http://www.cleansky.uni-bayreuth.de
• University of Bayreuth, Germany
Prof. Dr. H. Emmerich
heike.emmerich@uni-bayreuth.de
• Friedrich Alexander Universität Erlangen-Nürnberg, Germany
Prof. Dr. C. Körner
carolin.koerner@ww.uni-erlangen.de
• Fraunhofer Institute for Mechanics of Materials, Freiburg, Germany
Dr. D. Helm
dirk.helm@iwm.fraunhofer.de