2018 Example of a Digital Twin for 3 D printing
0 likes148 views
The document discusses advanced manufacturing through 3D printing and digitalization, emphasizing the importance of e-supply chain tools, automated defect analysis, and deep learning for achieving zero defects in additive manufacturing. It highlights the goals of the mansys project, which focuses on decision-making and supply chain management systems in metal additive manufacturing for industries like medical and aviation. Additionally, it covers the integration of various ICT tools for process optimization, including product design and predictive maintenance.
2018 Example of a Digital Twin for 3 D printing
- 1. ADVANCED MANUFACTURING WITH 3D PRINTING DIGITALIZATION: A KEY ELEMENT FOR THE NEXT LEVEL OF AM Jan Eite Bullema | Senior Scientist | TNO jan_eite.bullema@tno.nl
- 2. AM SYSTEMS The AMSYSTEMS Center is a joint innovation center of TNO and the High Tech Systems Center of Eindhoven University of Technology (TU/e HTSC) to accelerate (new ways of) additive manufacturing in diverse industries
- 3. CONTENT - e-supply chain tools for additive manufacturing - automated root cause analyses of printing defects - use of deep learning towards Zero Defects
- 4. MANSYS The aim of the ManSYS project (2013 – 2016) was to develop a complete decision making system and robust supply chain management system for metal additive manufacturing; especially for medical / aviation / dental jan_eite.bullema@tno.nl
- 5. MANSYS: E-SUPPLY CHAIN FOR QUALITY PARTS The aim of the ManSYS project (2013 – 2016) was to develop a complete decision making system and robust supply chain management system for metal additive manufacturing; especially for medical / aviation / dental Three key elements were defined -Decision support software -Supply chain management -Optimization of new products jan_eite.bullema@tno.nl
- 6. SUPPLY CHAIN MANAGEMENT AND DECISION MAKING SOFTWARE Logistics End-User Decision making software Material Supplier Service Bureau ManSYS office ManSYS Cloud ManSYS Webservice jan_eite.bullema@tno.nl
- 7. E-SUPPLY CHAIN TOOLS FOR ADDITIVE MANUFACTURING jan_eite.bullema@tno.nl
- 8. OPTIMIZING OF NEW PRODUCTS/ ASTM F42 JG TEST ARTIFACTS
- 9. ICT TOOLS FOR AM INTEGRATION Additive Manufacturing is a vital enabler for Industry 4.0 jan_eite.bullema@tno.nl
- 10. ICT TOOLS FOR AM INTEGRATION Modelling and Simulation Design Materials Build Process Post Processing Product End of Life Cloud Enabled jan_eite.bullema@tno.nl
- 11. ICT TOOLS FOR AM INTEGRATION AM Stage Software Tool Modelling and Simulation SolidWorks 2018, Autodesk Netfabb, Digimat-AM, SimcenterTM Design SolidWorks 2018, Autodesk Netfabb, Magics, Digimat-AM Materials Digimat-AM, Senvol Database Build Process SolidWorks 2018, Autodesk Netfabb Post Processing Simufact Product Magics Reporting, TeamcenterTM End of Life TeamcenterTM Cloud Enabled SolidWorks 2018, Autodesk Netfabb, Cloud, Mindsphere Autodesk, Dassault, Materialise, Siemens are important solution providers jan_eite.bullema@tno.nl
- 12. CONTENT - e-supply chain tools for additive manufacturing - automated root cause analyses of printing defects - use of deep learning towards Zero Defects
- 13. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 14. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 15. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 16. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 17. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 18. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 19. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 20. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 21. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 22. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 23. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 24. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 25. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 26. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 27. AUTOMATED ROOT CAUSE ANALYSES OF PRINTING DEFECTS jan_eite.bullema@tno.nl
- 28. CONTENT - e-supply chain tools for additive manufacturing - automated root cause analyses of printing defects - use of deep learning towards Zero Defects
- 29. USE OF DEEP LEARNING TOWARDS ZERO DEFECTS With Deep Learning models are generated from (Big) Data In general: the more data available the better the models perform A caveat: developing useful models requires process knowledge jan_eite.bullema@tno.nl
- 30. USE OF DEEP LEARNING TOWARDS ZERO DEFECTS With Deep Learning models are generated from (Big) Data In general: the more data available the better the models perform A caveat: developing useful models requires process knowledge Product Control Process Control Design Control jan_eite.bullema@tno.nl
- 31. DEEP LEARNING FOR PRODUCT CONTROL With deep learning defect classification can be implemented jan_eite.bullema@tno.nl
- 32. DEEP LEARNING FOR PRODUCT CONTROL With deep learning defect classification can be implemented jan_eite.bullema@tno.nl
- 33. DEEP LEARNING FOR PRODUCT CONTROL With deep learning defect classification can be implemented jan_eite.bullema@tno.nl
- 34. DEEP LEARNING FOR PRODUCT CONTROL With deep learning defect classification can be implemented jan_eite.bullema@tno.nl
- 35. DEEP LEARNING FOR PRODUCT CONTROL With deep learning defect classification can be implemented jan_eite.bullema@tno.nl
- 36. DEEP LEARNING FOR PROCESS CONTROL A simple form of process control with Deep Learning is anomaly detection, ultimately the goal is closed loop process control with Deep Learning jan_eite.bullema@tno.nl
- 37. DEEP LEARNING FOR PROCESS CONTROL Example Anomaly detection for predictive maintenance on a motor https://www.oreilly.com/ideas/anomaly-detection-with-apache-mxnet
- 38. DEEP LEARNING FOR PROCESS CONTROL Anomaly predictions on a test data set to simulate failure https://www.oreilly.com/ideas/anomaly-detection-with-apache-mxnet
- 39. DEEP LEARNING FOR PROCESS CONTROL Anomaly detection can help the 3D printing process jan_eite.bullema@tno.nl
- 40. DEEP LEARNING FOR DESIGN CONTROL Currently some very basic design rules are implemented. With Deep Learning it is possible to not only validate a design but also optimize a design for optimal 3D printing jan_eite.bullema@tno.nl
- 41. DEEP LEARNING FOR DESIGN CONTROL Design Control will enable design of optimized products Ultimately coupling (1) manufacturing, (2) field performance and (3) user experience https://www.autodesk.com/solutions/generative-design Autodesk Generative Design is an example how 3D product designers can be helped in optimizing product design for 3D printing jan_eite.bullema@tno.nl
- 42. CONCLUSION Advanced Manufacturing with 3D Printing Digitalization: A key element for the next level of AM
- 43. THANK YOU FOR YOUR ATTENTION Innovation by TNO and TU/e High Tech Systems Center Explore more on amsystems.com