Periodic Reporting for period 2 - Ascent AM (Adding Simulation to the Corporate ENvironmenT for Additive Manufacturing)
Periodo di rendicontazione: 2018-02-01 al 2020-01-31
The achievement of the Clean Sky 2 (CS 2) objectives, concerning the reduction of CO2 and fuel burn by up to 30 % as well as the reduction of NOX by up to 40 %, is strongly associated with the availability of innovative technologies and adapted structural design. Additive manufacturing (AM) is one of the key enablers and thus inalienable to reach these goals. First simple aero engine parts manufactured by AM processes, like laser beam melting (LBM) have successfully been introduced into mass production. To use all possible benefits of the technology like e. g. design freedom, cost efficiency and a reduced time to market, first time right production is essential. One major issue are the distortions after production. Employing manufacturing simulation to support process development is a promising approach to realize first time right production. Current simulation models however, fail to calculate distortions within an acceptable time frame and do not respect post-processes such as stress relief annealing. Thus, the goal of this project is to integrate a suitable simulation-based process chain into the tool landscape for AM process preparation that enables a prediction of distortions with high accuracy in reasonable time to generate benefit for AM users. The main focus lies on the metal-based layered manufacturing process of LBM as this is the most widespread approach to generating three-dimensional parts from powder material.
The following four objectives are set out for the proposed work:
Consideration of real life AM
Current simulation systems for the LBM process cover only the actual build-up process for simplified geometries. The proposed project aims to incorporate a creep model to enable the simulation of the stress relief an-nealing and a digital data chain for the predeformation of the input design both with measured and simulated results in order to complete the process chain and to increase shape accuracy. Additionally, the simulation of industry relevant geometries as well as current and future developments of LBM systems is realized.
Acceleration of simulation
With simulation results of small scale parts ranging in the order of days, the practical benefit of simulation is limited. The goal of this project is to use modelling techniques, adapted numerical methods as well as an efficient automation of data handling to decrease the duration of LBM simulations to be in the range or lower than the manufacturing time. With the implementation in an open source environment, all auxiliary processes can be tailored to the specific needs of the tool chain which increases the overall efficiency.
Increasing usability of AM
The LBM process itself is highly automated. However, certain downstream and upstream processes require a lot of manual work by an expert. Therefore, a digital data chain will be established in the simulation tool chain to support the worker in these processes. To increase usability of AM, the goal is that after the provided training sessions every participant is able to run the system without further expertise. The flexibility of open source programs will be used to automate all processes as far as possible. An exhaustive documentation is made available in addition to the source code itself, providing users with full insight and traceability.
Increase fidelity of predictions
The predictions of the simulation tool developed in this research project show very good results. The first-time-right manufacturing could be realized for most of the geometries. In addition, the complete process chain could be realized including the thermal post-processing. For this purpose, a separate creep subroutine for the material IN718 was determined experimentally. Especially the prediction accuracy for parts without support structure is very good. In order to improve the prediction accuracy for the support structures, a Multy Fidelity approach will be used in this research project. This approach combines the numerical (low fidelity) and experimental (high fidelity) results obtained in this research project.
The following four objectives are set out for the proposed work:
Consideration of real life AM
Current simulation systems for the LBM process cover only the actual build-up process for simplified geometries. The proposed project aims to incorporate a creep model to enable the simulation of the stress relief an-nealing and a digital data chain for the predeformation of the input design both with measured and simulated results in order to complete the process chain and to increase shape accuracy. Additionally, the simulation of industry relevant geometries as well as current and future developments of LBM systems is realized.
Acceleration of simulation
With simulation results of small scale parts ranging in the order of days, the practical benefit of simulation is limited. The goal of this project is to use modelling techniques, adapted numerical methods as well as an efficient automation of data handling to decrease the duration of LBM simulations to be in the range or lower than the manufacturing time. With the implementation in an open source environment, all auxiliary processes can be tailored to the specific needs of the tool chain which increases the overall efficiency.
Increasing usability of AM
The LBM process itself is highly automated. However, certain downstream and upstream processes require a lot of manual work by an expert. Therefore, a digital data chain will be established in the simulation tool chain to support the worker in these processes. To increase usability of AM, the goal is that after the provided training sessions every participant is able to run the system without further expertise. The flexibility of open source programs will be used to automate all processes as far as possible. An exhaustive documentation is made available in addition to the source code itself, providing users with full insight and traceability.
Increase fidelity of predictions
The predictions of the simulation tool developed in this research project show very good results. The first-time-right manufacturing could be realized for most of the geometries. In addition, the complete process chain could be realized including the thermal post-processing. For this purpose, a separate creep subroutine for the material IN718 was determined experimentally. Especially the prediction accuracy for parts without support structure is very good. In order to improve the prediction accuracy for the support structures, a Multy Fidelity approach will be used in this research project. This approach combines the numerical (low fidelity) and experimental (high fidelity) results obtained in this research project.
The following results are achieved in the project AscentAM so far:
• A simulation tool for laser beam melting with the open source tool CalculiX is established. A sequential coupled thermomechanical model is implemented in to order to calculate the deformation and the induced residual stresses of a part after the build-up process.
• A CAD-based data chain is developed with a maximal deviation of the FEM mesh on the range of the real layer height. The data chain includes a reengineering of the CAD model with the predeformed mesh from the simulation.
• The calculation is accelerated by the development and implementation of a LBM-specific direct solver and an adapted multirate method which takes advantage of the layerwise partitioning of the part. In addition, the required time steps for the structural calculation are reduced without loss of accuracy.
• A characteristic time scale is introduced in order to allow the transfer of the thermal history from a layer scale to a part scale.
• Heat management strategies are implemented to identify heat accumulation during the build-up process and to adjust the energy input in those regions.
• The continuous verification and validation methodology is set up as a continuous integration process directly in gitlab.
• A new creep subrutine was experimentally determined and implemented for the material IN 718.
• The process step stress relief annealing was added.
• The process step of direction dependent separation from the building plate was added.
• The validation of the results was successfully completed.
• Experimental and numerical results were obtained for the substitute material of the support structure.
• Simulations with support structures have been enabled.
• A simulation tool for laser beam melting with the open source tool CalculiX is established. A sequential coupled thermomechanical model is implemented in to order to calculate the deformation and the induced residual stresses of a part after the build-up process.
• A CAD-based data chain is developed with a maximal deviation of the FEM mesh on the range of the real layer height. The data chain includes a reengineering of the CAD model with the predeformed mesh from the simulation.
• The calculation is accelerated by the development and implementation of a LBM-specific direct solver and an adapted multirate method which takes advantage of the layerwise partitioning of the part. In addition, the required time steps for the structural calculation are reduced without loss of accuracy.
• A characteristic time scale is introduced in order to allow the transfer of the thermal history from a layer scale to a part scale.
• Heat management strategies are implemented to identify heat accumulation during the build-up process and to adjust the energy input in those regions.
• The continuous verification and validation methodology is set up as a continuous integration process directly in gitlab.
• A new creep subrutine was experimentally determined and implemented for the material IN 718.
• The process step stress relief annealing was added.
• The process step of direction dependent separation from the building plate was added.
• The validation of the results was successfully completed.
• Experimental and numerical results were obtained for the substitute material of the support structure.
• Simulations with support structures have been enabled.
A fast and accurate simulation tool for laser beam melting is established which takes into account the partspecific thermal history for the calculation of the distortion and induced residual stresses of a part after the build-up process.
Support structures are indispensable for the manufacturing of complex parts. There are different types of support structures, most of them are filigree. The aim is to derive a fast and efficient consideration of those structures in simulation in the further project. The focus will be on modelling of support structures, but also general filigree areas which belongs to the actual part will be regarded.
Creep experiments are conducted in order to generate viscoelasticity models for the simulation of stress relief annealing. As the proposed work also targets to determine whether creep is relevant for the build-up process itself, an analysis of relevant creep mechanisms is conducted. If applicable, a combination of different models for the different process steps is employed.
This project possesses direct impact for the AM industry that enables e. g. shorter design cycles and less material waste. Additionally, the project will contribute to the establishment of AM as an alternative form of production to conventional casting and machining that provides advantages not only in the manufacturing of single parts, but also largescale production. The direct impact of the project as well as the progression of additive manufacturing in general contributes to the objectives set out in Clean Sky 2 programme for a more environmentally friendly air traffic and an increasing European competitiveness.
Support structures are indispensable for the manufacturing of complex parts. There are different types of support structures, most of them are filigree. The aim is to derive a fast and efficient consideration of those structures in simulation in the further project. The focus will be on modelling of support structures, but also general filigree areas which belongs to the actual part will be regarded.
Creep experiments are conducted in order to generate viscoelasticity models for the simulation of stress relief annealing. As the proposed work also targets to determine whether creep is relevant for the build-up process itself, an analysis of relevant creep mechanisms is conducted. If applicable, a combination of different models for the different process steps is employed.
This project possesses direct impact for the AM industry that enables e. g. shorter design cycles and less material waste. Additionally, the project will contribute to the establishment of AM as an alternative form of production to conventional casting and machining that provides advantages not only in the manufacturing of single parts, but also largescale production. The direct impact of the project as well as the progression of additive manufacturing in general contributes to the objectives set out in Clean Sky 2 programme for a more environmentally friendly air traffic and an increasing European competitiveness.