Final Report Summary - INTFOP (Integrating Forging and Process Simulation for turbine disks)
Executive Summary:
The strong need for higher efficiency, reduced CO2 and NOX emissions, weight and noise reduction in aircraft engines leads to a demand of optimized lightweight engine disks. The new generation of aircraft engines with their geared turbofan concept fuels this demand. The special design of such engines with their faster rotating low pressure turbine leads to higher loads and therefore requires parts with tailored mechanical properties.
The objective of the project INTFOP was the implementation of an integrated simulation chain for aircraft engine disks including pre-material characteristics, thermo-mechanical processing (i.e. forging and heat treatment) and assessment of microstructure and functional properties. The integration of different computational results within an interdisciplinary and intercompany design-chain was a challenging task. A fully integrated simulation chain along the supply chain gives unique benefit for product development with respect to time, cost, and quality. This allows an optimization of the forgings in respect of weight, efficiency and CO2 reduction. Böhler Schmiedetechnik GmbH & Co KG (BSTG) has developed a model for determining the microstructure (e.g. local grain size distribution in a component) from forging and heat treatment simulations. This model is in daily use for thermo-mechanical process design and optimization of turbine disks. In past R&D projects, Böhler Schmiedetechnik has developed models for determining the local functional properties depending on the microstructure. In this project, the simulation chain was integrated into an industrialized useful format, target-aimed to customer needs. Suitable strategies and methods for integration of the pre-material characteristics were assessed. In this part of the project, the project partner Bohler Edelstahl GmbH & Co KG, which produces amongst others billets as well as milled semi-final products made of different double or triple melted nickel base alloys, generated valuable input. Furthermore, the quality of the existing models was evaluated together with the customer. Comprehensive statistic investigations work was necessary in order to make the usage of the simulated data possible at the customer. Detailed investigations were executed to improve and generate functional properties models like yield stress, low cycle fatigue and creep models. Interfaces between pre-material supplier, forging house and customer were defined and generated. A validation was examined with forged LPT (low pressure turbine) demonstrator disks which confirmed the effectiveness and capabilities of the implemented simulation chain.
Project Context and Objectives:
Work package 1:
Effective project coordination
Project risk management & management of project success criteria
Work package 2:
Report on quality of all models used
Coordinated plan for model development and industrialization (applicability at the customer)
Generate important answers to the questions as far decoupled post-processor statistical analysis can be used for the coupled nonlinear simulation process
Work package 3:
Well-defined improvement possibilities determined
Possibilities to industrialize the improvements defined
Work package 4:
Interface to raw material supplier in the manner of a detailed statistical preparation of the microstructural distribution generated
Work package 5:
Simulation data has to be submitted to the customer on the one hand for information purpose (microstructure, fatigue and creep distributions) and on the other hand as direct input for further simulation activities at the customer
A detailed definition of the data submitted has to be executed
The interface has to be programmed and test runs have to be verified
Work package 6:
Improve microstructure model in the matter of grain size distribution and precipitation consideration
Develop statistic variation consideration
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Define data and methods how to use the models in the customer industry (question of industrialization)
Prepare a microstructure “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 7:
Coordination of developing a yield strength model (incl. industrialization and statistics aspects)
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Prepare a yield strength (YS) “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 8:
Improve an existing approach of an LCF/HCF (Low/High Cycle Fatigue) model (incl. industrialization and statistics aspects)
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Prepare a LCF/HCF “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 9:
Coordination of developing a creep model (incl. industrialization and statistics aspects)
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Prepare a creep model “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 10:
The simulation of the improved thermo-mechanical process and functional properties is finished and evaluated at MTU
Statistical computation of the whole simulation chain finished
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Documentation of the integrated design approach is finished
Work package 11:
Machining, UT testing and sending demonstrator disks to MTU
Manufacturing and testing of two different disks defined together with MTU
Validation work and definition of future activities finished
Project Results:
Work package 1:
Effective coordination of project team, partners and other relevant projects
Good communication between all involved parties
Organization of meeting and workshops
Project controlling
Financial and technical reporting
Project marketing
Work package 2:
The available INTFOP project background of Böhler Schmiedetechnik was analyzed with respect to microstructure modeling, yield-strength modeling, creep modeling, and low-cycle-fatigue (LCF) modeling. These models were described and evaluated in a report (D2.2). Using the example of LCF, a first assessment on through-the-process statistical integration was given based on existing LCF data (see D2.2). This input regarding statistical aspects has been considered in other work packages too. Finally, a plan how to develop a physically motivated computational model was developed (cf. D2.3).
Work package 3:
A state-of-the-art thermo-mechanical process was developed for reference and demonstrator disks. Extensive finite-element (FE) analyses of the thermo-mechanical processing steps were performed, resulting in a deep understanding of the influencing variables and their effects. Statistical effects were also considered. In addition, different process layouts were designed to study the influence of, for example, reheating on the mechanical properties. The disks were sent to Westmoreland Testing and Research for cut-up testing and all test results were discussed together with the customer.
Work package 4:
The question of how far more detailed characterization possibilities of all specific pre-material batches used for production of disks and therefore improved and optimized disks are in contrast with the collateral higher costs was discussed with Böhler Edelstahl GmbH & Co KG. A forging simulation with a non-uniform initial grain size distribution was performed based on statistical information derived from the YS data base (cf. WP 7). Additional information was generated by analyzing different batches of triplemelt Alloy 718 regarding microstructure. A detailed statistical analysis of the microstructure results was also performed.
Work package 5:
A detailed definition of the transferred data was done at the beginning of this work package. The interface for CAD data was tested by sharing CAD data with the customer during the development of the UT geometry for the demonstrator disk. The definition of the interface between BSTG and MTU regarding thermo-mechanical simulation results was discussed during the project meetings. A postprocessing tool has been developed accordingly. A first test run of this post-processing procedure at MTU was completed successfully. Additional work was necessary to update the postprocessor with respect to prediction of mechanical properties. Again, the subsequent test run together with MTU was completed successfully. To assure compatibility with the latest Deform version, the Deform user meeting was visited in 01/2013. Furthermore, the usage of grain flow information from FE simulation at BSTG as input for US simulation at MTU was defined and complemented by the development of a converting tool (pat2igs).
Work package 6:
The influence of variations of the simulation inputs (for example, furnace temperatures, resting and transfer times, etc.) on the simulation outputs (effective strain, temperature, and grain size distributions) was studied. The result of this sensitivity analysis was that the resting times are of especial importance. Based on the YS data base (WP7), a statistical analysis of actual vs. predicted grain sizes was performed, showing that the model has to be extended to higher temperatures to predict the abnormal grain growth due to delta dissolution. BSTG decided to set up a new research project, which focuses on this topic. The implementation of the DA effect in Deform based on dislocation density has been proofed to be a challenging task due to lack of experimental data (dislocation density cannot be reliably measured in Alloy 718). Instead, a phenomenological parameter (Oberwinkler-ratio) was developed to account for the DA effect.
Work package 7:
An extensive YS data base was build based on existing production and testing results of several disks. This data base was a valuable input in other work packages (WP4, WP6) and was used for an assessment of the new developed YS models. Furthermore, a statistical analysis of the results showed interesting trends of grain size, testing temperature and yield strength. The coordination with the CD-Lab project worked well. New challenges have been addressed in the CD-Lab project (for example, the consideration of the elliptical shape of precipitations) and all new derived models were implemented in MatCalc. A MatCalc work shop was organized in 02/2013 in Munich, where the YS modeling was tested together with MTU.
Work package 8:
Additional LCF tests were performed in the frame of a MCL project. The parameters were adapted in accordance with MTU. A comprehensive statistical analysis of existing and additional LCF results was performed. This analysis led to an alternative way of modeling the LCF strength. In discussions with MTU it was agreed to further elaborate this “statistical” approach for LCF modeling. Thereby, valuable statistical input was delivered by a subcontractor. Finally, a statistically sound model was presented, which describes the low cycle fatigue in respect of testing temperature, strain amplitude and yield strength.
Work package 9:
Creep specimens were manufactured from a direct-aged air cooled Alloy 718 disk. Adapters were manufactured from Udimet 720 material (higher creep resistance than Alloy 718), for testing the Alloy 718 specimens at the lab of TU Graz. The microstructural characterization was performed at BEG. In addition, existing creep test data curves were requested from Westmoreland Mechanical Testing & Research. A literature survey was done. Thereby, two interesting models – Han Yafang & M.C. Chaturvedi (1989), and E.S. Huron et. al. (2012) – were identified. They were combined to a new approach that is able to describe the creep behavior in respect of both, grain size and precipitations. The test results mentioned before were used for the verification of this new approach. The validation was done based on test results of the validation disks.
Work package 10:
Based on the results of the previous work packages, the mechanical properties (yield strength, LCF, creep) of validation disks was simulated. The results were discussed together with the customer. Optimization potential in the process design was identified.
Work package 11:
Two demonstrator forgings were manufactured for MTU. The disks were forged, pre-machined and ultrasonic-tested at Böhler Schmiedetechnik GmbH & Co KG and delivered to MTU for final machining and testing. Furthermore, the cut-up plan for two validation disks were defined by Böhler Schmiedetechnik GmbH & Co KG and checked by MTU Aero Engines. Two validation disks were forged by BSTG. Testing were carried out by Westmoreland Mechanical Testing and Research Inc. and included 136 hot tensile tests, 17 LCF tests, 16 creep tests and 2 grain flows. The test results were summarized and discussed together with the customer. The comparison of actual and predicted results for these two forgings showed a good correlation and confirmed the successful implementation of integrated forging and process simulation into SAGE4 GTF LPT rotor design.
Potential Impact:
The results of this project will be used for future simulation-based process design of forded engine parts made of Alloy 718 to achieve optimized mechanical properties. Therefore, the main impact of the project is the possibility to optimize engine disks with respect to their mechanical properties to enable lightweight design. As a consequence, a reduction of weight of the engine becomes possible. This is the basis for future aircraft engines with higher efficiency and lower emissions. Finally, this will lead to a better competitiveness of the European aircraft industry.
List of Websites:
Bernd Oberwinkler, Böhler Schmiedetechnik GmbH & Co KG, 8605 Kapfenberg, Austria
The strong need for higher efficiency, reduced CO2 and NOX emissions, weight and noise reduction in aircraft engines leads to a demand of optimized lightweight engine disks. The new generation of aircraft engines with their geared turbofan concept fuels this demand. The special design of such engines with their faster rotating low pressure turbine leads to higher loads and therefore requires parts with tailored mechanical properties.
The objective of the project INTFOP was the implementation of an integrated simulation chain for aircraft engine disks including pre-material characteristics, thermo-mechanical processing (i.e. forging and heat treatment) and assessment of microstructure and functional properties. The integration of different computational results within an interdisciplinary and intercompany design-chain was a challenging task. A fully integrated simulation chain along the supply chain gives unique benefit for product development with respect to time, cost, and quality. This allows an optimization of the forgings in respect of weight, efficiency and CO2 reduction. Böhler Schmiedetechnik GmbH & Co KG (BSTG) has developed a model for determining the microstructure (e.g. local grain size distribution in a component) from forging and heat treatment simulations. This model is in daily use for thermo-mechanical process design and optimization of turbine disks. In past R&D projects, Böhler Schmiedetechnik has developed models for determining the local functional properties depending on the microstructure. In this project, the simulation chain was integrated into an industrialized useful format, target-aimed to customer needs. Suitable strategies and methods for integration of the pre-material characteristics were assessed. In this part of the project, the project partner Bohler Edelstahl GmbH & Co KG, which produces amongst others billets as well as milled semi-final products made of different double or triple melted nickel base alloys, generated valuable input. Furthermore, the quality of the existing models was evaluated together with the customer. Comprehensive statistic investigations work was necessary in order to make the usage of the simulated data possible at the customer. Detailed investigations were executed to improve and generate functional properties models like yield stress, low cycle fatigue and creep models. Interfaces between pre-material supplier, forging house and customer were defined and generated. A validation was examined with forged LPT (low pressure turbine) demonstrator disks which confirmed the effectiveness and capabilities of the implemented simulation chain.
Project Context and Objectives:
Work package 1:
Effective project coordination
Project risk management & management of project success criteria
Work package 2:
Report on quality of all models used
Coordinated plan for model development and industrialization (applicability at the customer)
Generate important answers to the questions as far decoupled post-processor statistical analysis can be used for the coupled nonlinear simulation process
Work package 3:
Well-defined improvement possibilities determined
Possibilities to industrialize the improvements defined
Work package 4:
Interface to raw material supplier in the manner of a detailed statistical preparation of the microstructural distribution generated
Work package 5:
Simulation data has to be submitted to the customer on the one hand for information purpose (microstructure, fatigue and creep distributions) and on the other hand as direct input for further simulation activities at the customer
A detailed definition of the data submitted has to be executed
The interface has to be programmed and test runs have to be verified
Work package 6:
Improve microstructure model in the matter of grain size distribution and precipitation consideration
Develop statistic variation consideration
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Define data and methods how to use the models in the customer industry (question of industrialization)
Prepare a microstructure “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 7:
Coordination of developing a yield strength model (incl. industrialization and statistics aspects)
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Prepare a yield strength (YS) “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 8:
Improve an existing approach of an LCF/HCF (Low/High Cycle Fatigue) model (incl. industrialization and statistics aspects)
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Prepare a LCF/HCF “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 9:
Coordination of developing a creep model (incl. industrialization and statistics aspects)
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Prepare a creep model “industrialization document” that shows the background and the complete procedure of how using the results at the customer
Work package 10:
The simulation of the improved thermo-mechanical process and functional properties is finished and evaluated at MTU
Statistical computation of the whole simulation chain finished
Execute cost-benefit analysis and coordination on most useful degree of accuracy
Documentation of the integrated design approach is finished
Work package 11:
Machining, UT testing and sending demonstrator disks to MTU
Manufacturing and testing of two different disks defined together with MTU
Validation work and definition of future activities finished
Project Results:
Work package 1:
Effective coordination of project team, partners and other relevant projects
Good communication between all involved parties
Organization of meeting and workshops
Project controlling
Financial and technical reporting
Project marketing
Work package 2:
The available INTFOP project background of Böhler Schmiedetechnik was analyzed with respect to microstructure modeling, yield-strength modeling, creep modeling, and low-cycle-fatigue (LCF) modeling. These models were described and evaluated in a report (D2.2). Using the example of LCF, a first assessment on through-the-process statistical integration was given based on existing LCF data (see D2.2). This input regarding statistical aspects has been considered in other work packages too. Finally, a plan how to develop a physically motivated computational model was developed (cf. D2.3).
Work package 3:
A state-of-the-art thermo-mechanical process was developed for reference and demonstrator disks. Extensive finite-element (FE) analyses of the thermo-mechanical processing steps were performed, resulting in a deep understanding of the influencing variables and their effects. Statistical effects were also considered. In addition, different process layouts were designed to study the influence of, for example, reheating on the mechanical properties. The disks were sent to Westmoreland Testing and Research for cut-up testing and all test results were discussed together with the customer.
Work package 4:
The question of how far more detailed characterization possibilities of all specific pre-material batches used for production of disks and therefore improved and optimized disks are in contrast with the collateral higher costs was discussed with Böhler Edelstahl GmbH & Co KG. A forging simulation with a non-uniform initial grain size distribution was performed based on statistical information derived from the YS data base (cf. WP 7). Additional information was generated by analyzing different batches of triplemelt Alloy 718 regarding microstructure. A detailed statistical analysis of the microstructure results was also performed.
Work package 5:
A detailed definition of the transferred data was done at the beginning of this work package. The interface for CAD data was tested by sharing CAD data with the customer during the development of the UT geometry for the demonstrator disk. The definition of the interface between BSTG and MTU regarding thermo-mechanical simulation results was discussed during the project meetings. A postprocessing tool has been developed accordingly. A first test run of this post-processing procedure at MTU was completed successfully. Additional work was necessary to update the postprocessor with respect to prediction of mechanical properties. Again, the subsequent test run together with MTU was completed successfully. To assure compatibility with the latest Deform version, the Deform user meeting was visited in 01/2013. Furthermore, the usage of grain flow information from FE simulation at BSTG as input for US simulation at MTU was defined and complemented by the development of a converting tool (pat2igs).
Work package 6:
The influence of variations of the simulation inputs (for example, furnace temperatures, resting and transfer times, etc.) on the simulation outputs (effective strain, temperature, and grain size distributions) was studied. The result of this sensitivity analysis was that the resting times are of especial importance. Based on the YS data base (WP7), a statistical analysis of actual vs. predicted grain sizes was performed, showing that the model has to be extended to higher temperatures to predict the abnormal grain growth due to delta dissolution. BSTG decided to set up a new research project, which focuses on this topic. The implementation of the DA effect in Deform based on dislocation density has been proofed to be a challenging task due to lack of experimental data (dislocation density cannot be reliably measured in Alloy 718). Instead, a phenomenological parameter (Oberwinkler-ratio) was developed to account for the DA effect.
Work package 7:
An extensive YS data base was build based on existing production and testing results of several disks. This data base was a valuable input in other work packages (WP4, WP6) and was used for an assessment of the new developed YS models. Furthermore, a statistical analysis of the results showed interesting trends of grain size, testing temperature and yield strength. The coordination with the CD-Lab project worked well. New challenges have been addressed in the CD-Lab project (for example, the consideration of the elliptical shape of precipitations) and all new derived models were implemented in MatCalc. A MatCalc work shop was organized in 02/2013 in Munich, where the YS modeling was tested together with MTU.
Work package 8:
Additional LCF tests were performed in the frame of a MCL project. The parameters were adapted in accordance with MTU. A comprehensive statistical analysis of existing and additional LCF results was performed. This analysis led to an alternative way of modeling the LCF strength. In discussions with MTU it was agreed to further elaborate this “statistical” approach for LCF modeling. Thereby, valuable statistical input was delivered by a subcontractor. Finally, a statistically sound model was presented, which describes the low cycle fatigue in respect of testing temperature, strain amplitude and yield strength.
Work package 9:
Creep specimens were manufactured from a direct-aged air cooled Alloy 718 disk. Adapters were manufactured from Udimet 720 material (higher creep resistance than Alloy 718), for testing the Alloy 718 specimens at the lab of TU Graz. The microstructural characterization was performed at BEG. In addition, existing creep test data curves were requested from Westmoreland Mechanical Testing & Research. A literature survey was done. Thereby, two interesting models – Han Yafang & M.C. Chaturvedi (1989), and E.S. Huron et. al. (2012) – were identified. They were combined to a new approach that is able to describe the creep behavior in respect of both, grain size and precipitations. The test results mentioned before were used for the verification of this new approach. The validation was done based on test results of the validation disks.
Work package 10:
Based on the results of the previous work packages, the mechanical properties (yield strength, LCF, creep) of validation disks was simulated. The results were discussed together with the customer. Optimization potential in the process design was identified.
Work package 11:
Two demonstrator forgings were manufactured for MTU. The disks were forged, pre-machined and ultrasonic-tested at Böhler Schmiedetechnik GmbH & Co KG and delivered to MTU for final machining and testing. Furthermore, the cut-up plan for two validation disks were defined by Böhler Schmiedetechnik GmbH & Co KG and checked by MTU Aero Engines. Two validation disks were forged by BSTG. Testing were carried out by Westmoreland Mechanical Testing and Research Inc. and included 136 hot tensile tests, 17 LCF tests, 16 creep tests and 2 grain flows. The test results were summarized and discussed together with the customer. The comparison of actual and predicted results for these two forgings showed a good correlation and confirmed the successful implementation of integrated forging and process simulation into SAGE4 GTF LPT rotor design.
Potential Impact:
The results of this project will be used for future simulation-based process design of forded engine parts made of Alloy 718 to achieve optimized mechanical properties. Therefore, the main impact of the project is the possibility to optimize engine disks with respect to their mechanical properties to enable lightweight design. As a consequence, a reduction of weight of the engine becomes possible. This is the basis for future aircraft engines with higher efficiency and lower emissions. Finally, this will lead to a better competitiveness of the European aircraft industry.
List of Websites:
Bernd Oberwinkler, Böhler Schmiedetechnik GmbH & Co KG, 8605 Kapfenberg, Austria