Current Trends in Probabilistic Damage Tolerance with DARWIN

Transcription

1 Current Trends in Probabilistic Damage Tolerance with DARWIN P-SAR Conference March 31, 2015 R. Craig McClung Michael P. Enright Southwest Research Institute San Antonio, Texas Approved for Public Release

2 Acknowledgements: Collaborators Southwest Research Institute DARWIN team Jonathan Moody, Yi-Der Lee, Vikram Bhamidipati, James Sobotka, John McFarland, Kwai Chan Elder Research Inc. DARWIN team Simeon Fitch, Ben Guseman, Colin Thomas FAA Steering Committee and Subcontractors GE Aviation, Honeywell, Pratt & Whitney, Rolls-Royce Corp. Nikki Howard, Alonso Peralta, Mike McClure, Jon Dubke Scientific Forming Technologies Corporation Wei-Tsu Wu and Ravi Shankar Fracture Analysis Consultants, Inc. Paul Wawrzynek and Bruce Carter Copyright 2015 Southwest Research Institute Current Trends in Probabilistic Damage Tolerance 2

3 Acknowledgments Funding Agencies Federal Aviation Administration Tim Mouzakis and Dave Galella Air Force Research Laboratory Pat Golden Naval Air Systems Command Ray Pickering NASA Glenn Research Center Jack Telesman Copyright 2015 Southwest Research Institute Current Trends in Probabilistic Damage Tolerance 3

4 Background Sioux City and Hard Alpha Sioux City accident (1989) highlighted the threat of undetected inherent material anomalies in engine rotors Titanium Hard Alpha Rare events, but not addressed by safe life design practices Engine industry developed an enhanced life management process Requested by Federal Aviation Administration (FAA) following Sioux City accident Developed by industry through AIA Rotor Integrity Sub-Committee (RISC) Probabilistic damage tolerance methods and opportunity inspections Parallel improvement of melt practice and NDE inspections Copyright 2015 Southwest Research Institute Current Trends in Probabilistic Damage Tolerance 4

5 Background: Pensacola and Surface Damage Pensacola accident (1996) highlighted the threat of undetected anomalies induced during manufacturing or maintenance FAA/RISC activity expanded to broaden the application of probabilistic damage tolerance Initial focus on machined holes Parallel effort on robust manufacturing methods 5

6 Development of FAA Advisory Material Specific damage tolerance methods for specific anomaly types have been developed and documented by the FAA in a series of Advisory Circulars AC (2001) titanium hard alpha AC (2009) general guidance for life-limited parts AC (2009) hole features Copyright 2015 Southwest Research Institute Current Trends in Probabilistic Damage Tolerance 6

7 Origins of FAA Grants As RISC and the FAA developed the enhanced life management process, they saw that further research and development (R&D) was needed to address shortfalls in technology and data SwRI (guided by RISC) proposed to the FAA and was awarded a series of R&D grants to address these shortfalls Enhanced predictive tool capability (DARWIN) Supplementary material/anomaly behavior characterization and modeling Goal to provide direct support for implementation and improvement of advisory material such as AC and AC SwRI is program manager U.S. engine companies serve as steering committee, major subcontractors 7

8 Overall FAA/RISC Vision Damage Tolerance Methodology Current Focus Inherent Flaws (Melt related, etc) Induced Flaws Titanium Hard Alpha Ni Anomalies Manufacturing Maintenance/ Service Analytical Method: Probabilistic FM Risk Calc <DTR Analysis Tool calibrated by Test Case Criteria Calibrated by Experience Circular holes Attachment slots Smooth surfaces Other? Analytical Method: Probabilistic FM Risk Calc <DTR Analysis Tool calibrated by Test Case Criteria Calibrated by Experience Copyright 2015 Southwest Research Institute Current Trends in Probabilistic Damage Tolerance 8

9 Probabilistic Fracture Mechanics Methodology Copyright 2005, Southwest 9 Research Institute Copyright 2015 Southwest Research Institute DARWIN Overview 9

10 DARWIN Implementation of Probabilistic Damage Tolerance Analysis 10

11 DARWIN Analysis Modes Inherent Anomalies Surface Damage Random Anomaly Zone-Based Risk (Volume) Feature-Based Risk (Area) 11

12 Current Trends: Outline New FAA-funded developments in Probabilistic Damage Tolerance Analysis (PDTA) methods for inherent and induced anomalies Linking PDTA with Integrated Computational Materials Engineering (ICME) Linking PDTA with advanced 3D fracture models Expansion of PDTA to additional damage states 12

13 Current Trends: Outline New FAA-funded developments in PDTA methods for inherent and induced anomalies Linking PDTA with Integrated Computational Materials Engineering (ICME) Linking PDTA with advanced 3D fracture models Expansion of PDTA to additional damage states 13

14 Automation of Life and Risk Analysis for 2D Models Automatic Fracture Models FCG Life Contours Automatic Risk Zoning 14

15 Automation of Life and Risk Analysis for 3D Models Extend 2D algorithms to 3D models Improve speed for large models See Moody, et al., 10 am Wed, Probabilistic Methods 2 15

16 Surface and Line Zoning for Attachment Slots RISC and the FAA are currently developing PDTA methods to address manufacturing and maintenance damage at attachment slots Surface damage frequency is generally a function of the surface area Experience has shown that certain regions of the attachment geometry may be more susceptible to damage than others New methods have been developed to define surface and line zones that combine locations with similar risk levels 16

17 New Anomaly Appearance at Shop Visits Inherent material or induced manufacturing anomalies are present when component enters service Induced maintenance anomalies may be introduced on the surface throughout the component lifetime Some attachment slot anomalies are linked to shop visits New probabilistic algorithms have been developed for shop visit anomalies User can define a random damage occurrence time, which may be linked with shop visit inspections (before or after NDE) User defines shop visit anomaly distribution(s) Shop visit anomalies may have nonzero crack formation lives User is responsible for providing crack formation algorithm Framework can be used to address new anomalies from other in-service events (e.g., FOD) 17

18 Importing Shop Visit Anomaly Distributions 18 18

19 Defining Timetables for Shop Visit Anomalies 19

20 Current Trends: Outline New FAA-funded developments in PDTA methods for inherent and induced anomalies Linking PDTA with Integrated Computational Materials Engineering (ICME) Linking PDTA with advanced 3D fracture models Expansion of PDTA to additional damage states 20

21 Integrating PDTA with Manufacturing Process Simulation Link DEFORM output with DARWIN input Finite element geometry (nodes and elements) Finite element stress, temperature, and strain results Residual stresses at the end of processing / spin test Location specific microstructure / property data Tracked location and orientation of material anomalies 21

22 DARWIN-DEFORM Links Residual Stresses Microstructure Anomaly Tracking and Deformation 22

23 Recent ICME Activities Linking DARWIN with DEFORM for deterministic and random Bulk residual stress Microstructure Average and ALA grain size Forging strains See Enright et al., 11:00, this session da/dn scaling factors for random grain size variations 23

24 DARWIN and NDE Simulation DARWIN can address the effects of NDE on fracture risk via POD curves Currently it is the responsibility of the DARWIN user to provide the appropriate POD curve for a given location & inspection type Physics-based software is available to simulate NDE inspections and determine POD Integration of DARWIN and XRSIM software has recently been explored in a new AFRL SBIR (Phase I) Partner: NDE Technologies Copyright 2015 Southwest Research Institute Current Trends in Probabilistic Damage Tolerance 24

25 Current Trends: Outline New FAA-funded developments in PDTA methods for inherent and induced anomalies Linking PDTA with Integrated Computational Materials Engineering (ICME) Linking PDTA with advanced 3D fracture models Expansion of PDTA to additional damage states 25

26 Linking PDTA with Advanced 3D Fracture Models Traditional engineering approaches to FCG employ simple rectangular fracture models with elliptical cracks CC11 CC09 SC30 SC31 EC05 EC04 EC05 EC04 CC08 SC18 TC13 CC10 SC29 TC11 TC12 Copyright 2015 Southwest Research Institute Current Trends in Probabilistic Damage Tolerance 26

27 Fracture Modeling for Complex Stress Gradients The weight function method provides a practical way to address complex univariant or bivariant stress gradients in the uncracked body 27

28 Extension of WF Methods to More Complex Geometries New weight function solutions for cracks at angled or chamfered corners in bivariant stress fields y CC12 CC18 σ(x,y) a-tip d y c-tip x σ(x,y) for x/d x +y/d y 1 was excluded d x 28

29 Integration of PDTA with 3D Numerical Fracture Methods For even more complex crack and component geometries, DARWIN has been linked to the 3D fracture software FRANC3D See Wawrzynek, et al., this session, 11:30 29

30 Current Trends: Outline New FAA-funded developments in PDTA methods for inherent and induced anomalies Linking PDTA with Integrated Computational Materials Engineering (ICME) Linking PDTA with advanced 3D fracture models Expansion of PDTA to additional damage states 30

31 Time-Dependent Crack Growth Address competing time-dependent and cycle-dependent crack growth mechanisms Address microstructural effects on damage growth rates See Chan, et al., 11:00 Wednesday, Session on Materials & Manufacturing: Modeling and Analysis 31

32 Pit and Crack Growth During Model corrosion pit formation and growth Determine when corrosion pits transition to form growing fatigue cracks Model resulting fatigue crack growth Hot Corrosion 32

33 Some Future Directions Continue improving the robustness and computational efficiency of life and risk automation methods for inherent anomalies Develop advanced methods for residual stress effects Extend these PDTA methods to other life-limited parts beyond engine rotors Address microstructure, anomaly, and residual stress effects in structural castings 33