On the measurement of fracture toughness to understand the cracking resistance of advanced high strength steel sheets

On the measurement of fracture toughness to understand the cracking resistance of Advanced High Strength Steel sheets David Frómeta Gutiérrez Doctoral thesis On the measurement of fracture toughness to understand the cracking resistance of Advanced High Strength Steel sheets David Frómeta Gutiérrez Thesis by compendium of publications submitted to Universitat Politècnica de Catalunya in partial fulfilment of the requirements for the degree of Doctor of Philosophy Thesis supervisor Jessica Calvo Muñoz Thesis supervisor Daniel Casellas i Padró To my brother, Daniel, who always believed in me i Abstract Automotive designers are constantly facing new challenges to meet the more and more stringent safety and CO2 emission legislations. Concerning the latter, vehicle lightweighting has become one of the main goals of the automotive industry, not only to reduce fuel consumption in fuelpowered cars but also to enhance the battery range in electric vehicles. At the same time, weight reduction cannot be attained at the expense of passenger’s safety in case of a crash. Hence, it is important to select the best-suited strategies to find the optimum balance between weight reduction and crashworthiness. In this sense, Advanced High Strength Steels (AHSS) have been positioned as one of the most effective solutions to this demand. AHSS present very high strength and high crash performance, which allows reducing vehicle mass while maintaining the safety of the occupants. These outstanding mechanical properties have promoted their widespread implementation for structural and crash-relevant automobile components. However, the application of AHSS have introduced new challenges related to their limited ductility and cracking resistance. Premature cracking during edge forming operations (edge cracking) or the occurrence and propagation of cracks under impact loading are some of the common cracking related issues in processing and implementation of AHSS. To face these problems, the development of new approaches to properly characterize the cracking resistance of AHSS has become unavoidable since conventional failure criteria based on uniaxial tensile properties and forming limit curves fail to describe cracking related phenomena. In this thesis, a fracture mechanics-based approach is proposed to rationalize and understand the crack initiation and propagation resistance of AHSS. Results have been correlated with edge cracking resistance and crash behaviour of a broad range of advanced high strength sheet steels. Fracture toughness is evaluated in the frame of fracture mechanics through different testing methods, such as the essential work of fracture, the J-integral and the Kahn-type tear tests. The relationship between the obtained fracture toughness parameters as well as the limitations of the different methods have been discussed. High-resolution video extensometry and Digital Image Correlation (DIC) techniques were used to investigate the fracture behaviour of the different steels. Edge cracking resistance is characterized by standard hole expansion tests and DIC- assisted hole tension tests. Crashworthiness is assessed through laboratory impact resistance tests. The influence of microstructural constituents on the crack propagation resistance of AHSS is also assessed. The results show that fracture toughness, in particular the specific essential work of fracture (we), is a suitable material property to understand the cracking behaviour of AHSS and rank the material’s resistance to different crack-related failures, such as edge fracture or crack propagation during a crash event. These conclusions are based on the good correlation established between we and the results from edge cracking and impact resistance tests. On the other hand, the experimental observations show that we can be used to discern the role of microstructural constituents on the fracture behaviour of AHSS. It is pointed out that proper microstructural design cannot be only focused on tensile properties since they do not inform about cracking resistance. According to all the experimental findings, the fracture toughness is considered as a relevant material property for AHSS design and performance classification. In line with this, a new classification system, considering global ductility and fracture toughness, is proposed for a more comprehensive description of the overall formability and fracture behaviour of AHSS. ii iii Resumen Los diseñadores de automóviles se enfrentan constantemente a nuevos desafíos para cumplir con las cada vez más estrictas legislaciones de seguridad y emisiones de CO2. Con respecto a esto último, el aligeramiento de los vehículos se ha convertido en uno de los principales objetivos de la industria automotriz, no solo para reducir el consumo en los automóviles de combustión interna, sino también para mejorar la autonomía de los vehículos eléctricos. Al mismo tiempo, la reducción de peso no se puede lograr a expensas de la seguridad del pasajero en caso de accidente. Por lo tanto, es importante seleccionar las estrategias más adecuadas para encontrar el equilibrio óptimo entre reducción de peso y resistencia al impacto. En este sentido, los aceros avanzados de alta resistencia (AHSS) se han posicionado como una de las soluciones más efectivas. Los AHSS presentan una elevada resistencia y un buen comportamiento en caso de impacto, lo que permite reducir el peso del vehículo manteniendo la seguridad de los ocupantes. Estas excepcionales propiedades mecánicas han contribuido a su extensa implementación en componentes estructurales y de seguridad en el automóvil. Sin embargo, estos aceros también han introducido nuevos problemas relacionados con su limitada ductilidad y resistencia la fisuración, como la aparición prematura de fisuras durante el conformado (edge cracking) o la generación de fisuras durante el impacto. Para hacer frente a estos problemas, se ha hecho inevitable el desarrollo de nuevos enfoques para caracterizar la resistencia a la fisuración de los AHSS, ya que los criterios convencionales basados en ensayos de tracción y curvas límite de conformabilidad no son adecuados. En esta tesis doctoral se propone un enfoque basado en la mecánica de la fractura para explicar este tipo de fracturas relacionadas con la resistencia a la iniciación y propagación de grietas en el material. Con este fin, se investiga la correlación entre las mediciones de tenacidad de fractura y la resistencia al edge cracking y el comportamiento en caso de impacto en una amplia gama de chapas de acero avanzado de alta resistencia. La tenacidad de fractura se evalúa en el marco de la mecánica de la fractura mediante distintos métodos como el trabajo esencial de fractura, la integral J o los ensayos tipo Kahn y se discute la relación entre los parámetros obtenidos, así como las limitaciones de los diferentes métodos. Se utilizan técnicas de video de alta resolución y correlación de imágenes digitales para investigar el comportamiento de fractura de los diferentes aceros. La resistencia edge cracking se caracteriza mediante ensayos de expansión de orificios (hole expansion tests). La resistencia al impacto se evalúa mediante ensayos de impacto de laboratorio. Finalmente, se analiza brevemente la influencia de la microestructura en la resistencia a la propagación de grietas de los AHSS. Los resultados muestran que la tenacidad de fractura, en concreto el trabajo esencial de fractura (we) es una herramienta útil para comprender fenómenos de fisuración en los AHSS. Estas conclusiones se basan en la buena correlación establecida entre we y los resultados de las pruebas de resistencia al impacto y al edge cracking. Por otro lado, las observaciones experimentales muestran el gran potencial del parámetro we para discernir el efecto de la microestructura en la resistencia a la fractura de los AHSS. Se destaca que el diseño microestructural no debe centrarse sólo en las propiedades de tracción, ya que éstas no aportan información sobre la resistencia a la propagación de fisuras. De acuerdo con esto, la tenacidad de fractura se considera una propiedad del material relevante para el diseño y clasificación de los AHSS y se propone un nuevo método de clasificación para una descripción más completa de la conformabilidad y la resistencia a la fractura de los aceros AHSS. x xi Table of Contents Abstract .......................................................................................................................................... i Resumen ....................................................................................................................................... iii Preface ........................................................................................................................................... v Acknowledgements ..................................................................................................................... vii List of publications ....................................................................................................................... ix Table of Contents ......................................................................................................................... xi List of figures .............................................................................................................................. xv List of tables ............................................................................................................................... xxi List of symbols and abbreviations ............................................................................................ xxiii Chapter 1 Introduction .................................................................................................................. 1 1.1 Advanced High Strength Steels ........................................................................................... 1 1.1.1. Definition and applications ......................................................................................... 1 1.1.2. Classification ............................................................................................................... 1 1.1.2.1. Dual Phase ............................................................................................................ 3 1.1.2.2. Complex Phase .................................................................................................... 4 1.1.2.3. TRIP .................................................................................................................... 4 1.1.2.4. Ferritic-Bainitic steel ............................................................................................ 5 1.1.2.5. Martensitic steel ................................................................................................... 5 1.1.2.6. Press Hardened steels ........................................................................................... 6 1.1.2.7. TWIP .................................................................................................................... 7 1.1.2.8. TRIP-assisted bainitic ferritic ............................................................................... 8 1.1.2.9. Quenched and partitioned ..................................................................................... 8 1.1.2.10. Medium Mn steels .............................................................................................. 9 1.1.2.11. δ-TRIP .............................................................................................................. 10 1.2 Global and local formability ............................................................................................. 10 1.2.1. Global formability ..................................................................................................... 11 1.2.2. Local formability ....................................................................................................... 12 1.2.2.1. Hole expansion tests ........................................................................................... 12 1.2.2.2. V-bending tests ................................................................................................... 12 1.2.2.3. Local fracture strain measurements .................................................................... 13 1.2.3. Global vs local formability maps .............................................................................. 14 1.3 Local formability and crack-related failures ..................................................................... 15 1.3.1 Edge cracking ............................................................................................................. 15 1.3.1 Crash fracture behaviour ............................................................................................ 17 1.4 Fracture toughness of advanced high strength steels ........................................................ 19 1.4.1. Introduction to fracture mechanics ............................................................................ 20 xii 1.4.2. Linear Elastic Fracture Mechanics ............................................................................ 20 1.4.3. Loading modes .......................................................................................................... 21 1.4.4. Crack tip triaxiality: plane stress vs plane strain ....................................................... 22 1.4.5. Influence of thickness on fracture toughness ............................................................ 23 1.4.6. Fracture toughness evaluation in the frame of linear elastic fracture mechanics ...... 23 1.4.7.1. Crack Tip Opening Displacement ...................................................................... 25 1.4.7.2. Crack Tip Opening Angle .................................................................................. 26 1.4.7.3. J-integral ............................................................................................................. 26 1.4.7.4. Relationship between J and CTOD .................................................................... 27 1.4.8.1. J-integral and CTOD .......................................................................................... 28 1.4.8.2. CTOA and δ5 ...................................................................................................... 31 1.4.8.3. Kahn-type tear tests ............................................................................................ 32 1.4.8.4. Essential Work of Fracture ................................................................................. 33 Chapter 2 Objectives and scope .................................................................................................. 37 2.1 Objectives of the work ...................................................................................................... 37 2.2 Scope of the research......................................................................................................... 38 Chapter 3 Materials and experimental methods .......................................................................... 41 3.1 Materials ............................................................................................................................ 41 3.2 Fracture toughness measurements ..................................................................................... 42 3.2.1.1 Notch preparation method ................................................................................... 44 3.2.1.2 Determination of the critical crack tip opening displacement. ............................ 45 3.2.1.3 Thickness strain measurements in DENT specimens .......................................... 45 3.3 Edge cracking resistance characterization ......................................................................... 47 3.3.1.1 Thickness strain measurements in HET specimens ............................................. 48 3.4 Crash behaviour characterization ...................................................................................... 50 3.5 Microstructural characterization ....................................................................................... 52 Chapter 4 Results ........................................................................................................................ 55 4.1.1.1. Specific essential work of fracture, we ............................................................... 55 4.1.1.2. Fracture toughness at crack initiation, wei .......................................................... 57 4.1.1.3. Critical crack tip opening displacement, δc ........................................................ 58 4.1.1.4. Thickness strain at crack initiation and propagation .......................................... 59 4.1.1.5. Specimens with EDM notches ........................................................................... 61 4.1.1.6. Specimens with sheared notches ........................................................................ 62 4.3.1.1 Hole Expansion Ratio .......................................................................................... 67 4.3.1.2 Thickness strain ................................................................................................... 67 Chapter 5 Discussion ................................................................................................................... 73 5.1.1.1 Crack initiation and propagation resistance, wei and we. ..................................... 78 xiii 5.1.1.2 Necking contribution to we .................................................................................. 78 5.1.1.3 Relationship between we and δc ........................................................................... 80 5.1.1.4 Influence of notch preparation method on EWF ................................................. 81 5.1.2.1 Influence of specimen geometry ......................................................................... 85 5.1.2.2 Relationship between we and Jc ........................................................................... 85 5.1.2.3 Thickness independent fracture toughness validation criterion .......................... 86 Chapter 6 Conclusions and future work .................................................................................... 107 References ................................................................................................................................. 113 Appendix A ............................................................................................................................... 123 Paper I. .................................................................................................................................. 125 Paper II. ................................................................................................................................. 145 Paper III ................................................................................................................................. 169 Paper IV ................................................................................................................................ 185 Appendix B ............................................................................................................................... 213 Paper A .................................................................................................................................. 215 Paper B .................................................................................................................................. 225 Paper C .................................................................................................................................. 233 Paper D .................................................................................................................................. 243 xiv xv List of figures Figure 1.1. Implementation of AHSS in the body structure of a modern car. Reference for a NAFTA mid-size sedan. Source: ArcelorMittal ............................................................................ 1 Figure 1.2. a) Average AHSS utilization in North America light passenger cars. Forecasts from 2015 to 2025. Image adapted from [4]. b) Body and closure metallic material content by type. Image adapted from [1]. ................................................................................................................ 2 Figure 1.3. Strength ductility diagram for various types type of steels, including conventional and AHSS grades [1] ........................................................................................................................... 3 Figure 1.4. a) Typical microstructure of DP steels (SEM micrograph). F: ferrite; M: martensite b) Time-temperature cycle applied to obtain the DP microstructure. Ms: martensite start temperature. ....................................................................................................................................................... 3 Figure 1.5. a) Microstructure of a CP steel (SEM micrograph). B: bainite; TM: tempered martensite b) Time-temperature cycle applied to obtain the CP microstructure. .......................... 4 Figure 1.6. a) Microstructure of TRIP steel. F: ferrite; B: bainite; RA: retained austenite. b) Engineering stress-strain curves for DP and TRIP steel of similar strength. ................................ 5 Figure 1.7. a) Microstructure of a FB 450/600 [1]. F: ferrite; B: bainite. b) Tensile strengthelongation diagram showing the range of mechanical properties covered by FB steels [1]. ........ 5 Figure 1.8. a) Microstructure of a MS (SEM micrograph). M: martensite. b) Time-temperature cycle applied to obtain martensitic microstructures. ..................................................................... 6 Figure 1.9. Schematic representation of the press hardening process [25]. .................................. 6 Figure 1.10. SEM micrographs of different press-hardened microstructures. a) martensite, b) bainite, c) bainite + martensite and d) ferrite + bainite [26].......................................................... 7 Figure 1.11. a) Microstructure of TWIP steels. A: austenite. b) Illustration of the dynamical Hall- Petch effect [27]. ........................................................................................................................... 7 Figure 1.12. SEM micrograph of TBF steel. B: bainite, F: Ferrite; RA: retained austenite. b) Thermal cycle for TBF production [9]. ......................................................................................... 8 Figure 1.13. a) Microstructure of Q&P steel. RA: retained austenite; TM: tempered martensite; LB: lower bainite. b) Schematic representation of the quenching and partitioning process [9]. .. 9 Figure 1.14. a) SEM micrograph of a medium Mn TRIP steel with a chemical composition of Fe– 7Mn–0.14C–0.23Si [38]. b) processing routine of austenite reverted transformation (ART) annealing applied for the production of ultrafine-grained medium-Mn TRIP steels [38]. ............ 9 Figure 1.15. SEM micrographs of: a) Fe-0.4C-1.5Mn-5.2Al [39], b) Fe-0.4C-2.5Mn-5.2Al [39] and c) Fe-0.39C-0.5Mn-3.8Al [41]. ............................................................................................ 10 Figure 1.16. a) Typical stress-strain curve obtained from a uniaxial tensile test [53]. b) Schematic representation of an FLC [56]. .................................................................................................... 11 Figure 1.17. a) Geometry of the punch used for Nakajima tests. b) Tool used for Marciniak tests [57]. ............................................................................................................................................. 12 Figure 1.18. a) Experimental procedure of the HET and HET sample after the test. b) Experimental setup for the 3-point V-bending tests according to VDA 238-100 at voestalpine Stahl [47]. .................................................................................................................................... 13 xvi Figure 1.19. a) Schematic representation of the parabolic fracture surface profile of a uniaxial tensile specimen with rectangular cross-section [52]. b) Fracture surface of a flat sheet tensile specimen. Microscope image. ..................................................................................................... 14 Figure 1.20. Local/global formability maps for AHSS classification. Global (G) formability is represented by the true uniform strain (εu) from uniaxial tensile tests. Local (L) formability is expressed in terms of a) True Fracture Strain (TFS) [52] and b) True thickness strain at fracture (ε3f) [66]. ...................................................................................................................................... 15 Figure 1.21. Examples of edge cracking in cold formed automotive components. a) body in white stamped part. Material: DP600 [45]. b) Automotive seat component. Material: DP980 [67]. c) Under seat beam made of DP980. Images courtesy of Centro Ricerche Fiat (CRF). ................. 16 Figure 1.22. SEM images of the surface of two mechanically sheared edges. CP steel (upper row) and DP steel (lower row). ............................................................................................................ 17 Figure 1.23. Main crash energy management areas in a vehicle. Adapted from [1]. .................. 17 Figure 1.24. Experimental setup and specimen geometry for laboratory impact resistance tests at voestalpine Stahl: a) axial impact tests [81]. b) bending impact tests [82]. ................................ 18 Figure 1.25. Examples of cracks in axial (a, b) and bending (c) crash tested samples of AHSS and PHS. Images from: a) [81], b) [49] and c) [82] ........................................................................... 19 Figure 1.26. Science and engineering fields covered by fracture mechanics [92]. ..................... 20 Figure 1.27. Loading modes that can be applied to a crack ........................................................ 22 Figure 1.28. Three-dimensional deformation at the crack tip of a cracked plate subjected to inplane loading (left) and schematic variation of transverse stress through the thickness at a point near the crack tip (right) [101]. ................................................................................................... 22 Figure 1.29. Three-dimensional plastic zone at the crack tip in a finite plate [102]. .................. 23 Figure 1.30. a) Thickness-dependence of fracture toughness. b) Effect of specimen thickness on ductile fracture surface morphology. Redrawn from [101]. ........................................................ 23 Figure 1.31. Standardized specimens for fracture toughness determination: CT specimen (left) and SENB specimen (right) [105] ............................................................................................... 24 Figure 1.32. Initially sharp crack blunts before fracture due to plastic deformation, causing the displacement δ in the crack tip (CTOD) [101]. ........................................................................... 25 Figure 1.33. CTOD defined as the displacement at the intersection of a 90° vertex with the crack faces [101]. .................................................................................................................................. 26 Figure 1.34. CTOA (ψ) definition. .............................................................................................. 26 Figure 1.35. Arbitrary contour around the tip of a crack [101]. .................................................. 27 Figure 1.36. J determination through partial unloadings following the compliance method [105]. ..................................................................................................................................................... 29 Figure 1.37. J values against crack extension for the J-R curve determination [111]. ................ 30 Figure 1.38. Compact tension specimen with an anti-buckling system [113] (left) and example of CTOA and ψc determination by direct optical measurements (right). ........................................ 32 Figure 1.39. Typical specimen geometry and load–displacement curve of a Kahn Tear Test. ... 33 Figure 1.40. DENT specimen used for the evaluation of the EWF (left) and definition of the Fracture Process Zone (right). ..................................................................................................... 34 xvii Figure 1.41. Experimental procedure for the determination of the essential work of fracture, we and the specific work for fracture initiation, wei. ......................................................................... 34 Figure 1.42. Valid ligament lengths for the evaluation of the EWF [139] .................................. 35 Figure 3.1. DENT specimen used for EWF tests and detail of the fatigue pre-crack at the notch root. ............................................................................................................................................. 43 Figure 3.2. a) Resonance fatigue machine. b) Universal testing machine INSTRON 5585H used for EWF tests. c) Full-field strain analysis in the ligament area of a DENT specimen. d) Schematic representation of the experimental setup of the DIC equipment. ................................................ 44 Figure 3.3. Different notch configurations investigated. a) Fatigue pre-crack, b) EDM notch and c) mechanically sheared notches. Optical microscope images. ................................................... 44 Figure 3.4. Example of determination of the critical crack tip opening displacement (δc). ........ 45 Figure 3.5. a) Experimental setup for the evaluation of the J-integral in CT specimens. b) Image from the video camera used for measuring the crack extension. ................................................ 46 Figure 3.6. a) Specimen geometry used for the Kahn-type tear tests.b) Determination of the true major strain at fracture in the notch tip by means of DIC. .......................................................... 46 Figure 3.7. a) Specimen geometry used for hole expansion tests. b) HER specimen before (left) and after (right) the test. .............................................................................................................. 47 Figure 3.8. Experimental setup for HER determination. ............................................................. 48 Figure 3.9. Digital images used for the evaluation of the HER. ................................................. 48 Figure 3.10. Longitudinal section of a crack in a HET specimen after the test and location of the thickness measurements performed. SAZ: shear affected zone. ................................................. 49 Figure 3.11. a) Specimen geometry used for Hole Tension Tests. b) DIC image showing the major strain at fracture. .......................................................................................................................... 49 Figure 3.12. DIC strain mapping and sections used for εf HTT determination .............................. 50 Figure 3.13. a) Geometry of axial crash samples with 1.5 mm blank thickness. b) Force vs impactor displacement curves for two axial impact tests with DP1000. ..................................... 50 Figure 3.14. Geometry of bending impact test samples with 1.5 mm blank thickness. Length: 900 mm. ............................................................................................................................................. 51 Figure 3.15. Crack location and definition of crash index for bending impact test. ................... 52 Figure 3.16. a) Scanning Electron Microscope (SEM). b) Formation of diffraction patterns (EBSP)......................................................................................................................................... 53 Figure 3.17. a) iNano nanoindenter. b) SEM image showing a nanoindentation array. ............. 53 Figure 3.18. a) Experimental setup for Synchrotron X-Ray Powder Diffraction measurements in the MSPD line of Alba Synchrotron. b) Location of the measurements in the specimen surface. c) Different stages of deformation investigated. ......................................................................... 54 Figure 4.1. wf as a function of the initial ligament length (l0). a) 780 MPa steel grades. b) 1500 MPa steel. .................................................................................................................................... 56 Figure 4.2. wf as a function of the initial ligament length (l0). Results for 1000 MPa steel grades. ..................................................................................................................................................... 56 Figure 4.3. wf as a function of the initial ligament length (l0). Results for 1200 MPa steel grades. ..................................................................................................................................................... 56 xviii Figure 4.4. PHS1500. Equivalent Mises strain at crack initiation. Small (left) and large (right) ligament length. ........................................................................................................................... 57 Figure 4.5. CP1200. Equivalent Mises strain at crack initiation. Small (left) and large (right) ligament length. ........................................................................................................................... 57 Figure 4.6. TBF. Equivalent Mises strain at crack initiation. Small (left) and large (right) ligament length ........................................................................................................................................... 57 Figure 4.7. wei results for all the studied steel grades. ................................................................. 58 Figure 4.8. df as a function of the initial ligament length (l0). a) 780 MPa steel grades. b) 1500 MPa steel ..................................................................................................................................... 58 Figure 4.9. df as a function of the initial ligament length (l0). Results for 1000 MPa steel grades. ..................................................................................................................................................... 59 Figure 4.10. df as a function of the initial ligament length (l0). Results for 1200 MPa steel grades. ..................................................................................................................................................... 59 Figure 4.11. Fracture surface of DENT specimens. Images from optical microscope. a) CP1000, b) Q&P, c) DP1000B, d) DP780. ................................................................................................ 60 Figure 4.12. True thickness strain (TTSDENT) as a function of the distance from the crack tip. 780 MPa steel grades ......................................................................................................................... 60 Figure 4.13. True thickness strain (TTSDENT) as a function of the distance from the crack tip. 1000 MPa steel grades. ........................................................................................................................ 60 Figure 4.14. True thickness strain (TTSDENT) as a function of the distance from the crack tip. 1200 MPa steel grades. ........................................................................................................................ 61 Figure 4.15. EWF results for specimens with EDM notches (ρ= 150 µm). wf as a function of the initial ligament length (l0). .......................................................................................................... 62 Figure 4.16. wei results for the 4 investigated steel grades. Results for EDM notched specimens. ..................................................................................................................................................... 62 Figure 4.17. EWF results for specimens with sheared notches (ρ≈ 2-3 µm). wf as a function of the initial ligament length (l0). .......................................................................................................... 63 Figure 4.18. J-R curves for the 4 investigated AHSS grades: a) CP, b) DP, c) TBF and d) Q&P. Figure from Paper I. .................................................................................................................... 64 Figure 4.19. Load-displacement curves from KTT. Only one representative curve for each material is shown. ........................................................................................................................ 65 Figure 4.20. Results of KTT. a) UIE, UPE and εf KTT. b) Tear strength and TYR. ....................... 65 Figure 4.21. True fracture strain (TFS) and true thickness strain (TTS) measured from the fracture surface of uniaxial tensile specimens. ......................................................................................... 66 Figure 4.22. HER for the studied AHSS grades. True thickness strain measured from hole expansion test specimens (TTSHET) is also indicated for the steels investigated in Paper II. ...... 67 Figure 4.23. Major strain at fracture at the hole edge (εf HTT) for the four investigated steel grades. Results for different punch-to-die clearances. ............................................................................. 68 Figure 4.24. a) Crash boxes of CP1000 (top) and DP1000 (bottom) after axial impact resistance tests at different speeds. b) Examples and location of cracks observed in axial crash tests. ....... 70 xix Figure 4.25. Variation of Crash Index (CI) as a function of the intrusion in axial impact tests for the studied steels. Energy absorbed is also indicated in the upper x-axis. Figure from Paper III. ..................................................................................................................................................... 70 Figure 4.26. Variation of Crash Index (CI) as a function of the intrusion in bending impact tests. ..................................................................................................................................................... 71 Figure 5.1. Typical load-displacement curves obtained from EWF tests. a)TRIP780, b) DP1000, c) 3rd Gen 1180Q&P, d) PHS1500. ............................................................................................. 75 Figure 5.2. σmax as a function of the ligament length for: a) TRIP780, b) DP1000, c) 3rd Gen 1180Q&P, d) PHS1500. The black dashed line represents the Hill’s criterion, 1.15 σYS. The red dashed line represents 1.15σY ...................................................................................................... 77 Figure 5.3. Stress criterion based on an average value of σmax ( σmean). The data below 0.9σmean and above 1.1σmean are excluded for EWF calculation. ...................................................................... 77 Figure 5.4. we and wei for the AHSS investigated in this work. ................................................... 78 Figure 5.5. a) True thickness strain (TTSDENT) as a function of the distance from the crack tip (image from Paper I). b) we and wei. ............................................................................................ 79 Figure 5.6. Comparison of essential work of fracture (we) and fracture toughness at crack initiation (wei) with true thickness strain measured in DENT specimens (TTSDENT) ................................... 80 Figure 5.7. Relationship of δc with we and flow properties. ........................................................ 81 Figure 5.8. EWF results for specimens with EDM notches (open symbols) and specimens with fatigue pre-cracks (black squares). a) CP1000 (results of specimens with sheared notches are also given, grey triangles). b) DP1000. c) TBF and d) Q&P. ............................................................. 82 Figure 5.9. EWF results for specimens with sheared notches (grey triangles) and specimens with fatigue pre-cracks (black squares). a) DP1000A, b) DP1000B and c) 3rd GenTBF1180. ........... 83 Figure 5.10. Comparison of EWF and J-integral results. J values obtained from DENT specimens. ..................................................................................................................................................... 86 Figure 5.11. Comparison between KTT and EWF results. ......................................................... 88 Figure 5.12. Correlation between we and: a) UPE, b) εf KTT......................................................... 88 Figure 5.13. Relationship between we and different tensile parameters. ..................................... 89 Figure 5.14. a) Correlation between we and HER. Figure adapted from Paper II. b) Fracture strains measured in HTT specimens (εf HTT) for the 4 steel grades investigated in Paper B with different punch to die clearance. we are also plotted. Figure from Paper B. .............................................. 91 Figure 5.15. Thickness strain measurements performed in DENT and HET specimens. ........... 92 Figure 5.16. Edge cracks observed in the component manufactured with DP steel grade (Paper A) ..................................................................................................................................................... 93 Figure 5.17. Results from mechanical characterizations for the investigated CP and DP grades. a) Engineering stress-strain curves, b) FLCs, c) HER and d) EWF. Image from Paper A. ............ 93 Figure 5.18. Results of essential work of fracture (we) against CIDR and energy absorbed at maximum intrusion (CI=20%). Image from Paper III. ............................................................... 94 Figure 5.19. Correlation between fracture toughness (we) and cracking behaviour in bending impact resistance tests (CIDR). CIDR values of 0 correspond to steel grades that did not showed cracking during crash tests. ......................................................................................................... 95 1.1 Advanced High Strength Steels 2 carefully adjusted by controlling the chemical compositions and the thermomechanical processing routes, have been developed in the last two decades. Figure 1.2. a) Average AHSS utilization in North America light passenger cars. Forecasts from 2015 to 2025. Image adapted from [4]. b) Body and closure metallic material content by type. Image adapted from [1]. In general, AHSS can be divided into three big groups or generations: 1st, 2nd and 3rd generation AHSS. Dual-Phase (DP), Complex Phase (CP), Martensitic (MS), Press-hardened (PHS) or Hot formed (HF), and Transformation-Induced Plasticity (TRIP) steels are part of the 1st generation of AHSS. This generation is characterized by showing better formability than single-phase high strength low alloyed (HSLA) steels of similar strength [5]. The 2nd generation of AHSS includes Twinning-Induced Plasticity (TWIP) and austenitic stainless steels. These steels present very high strength and extremely high ductility when compared to the 1st generation of AHSS. However, their production complexity and elevated costs, together with other problems of delayed cracking and poor weldability, have limited their application [6]. The 3rd generation was developed to cover the gap between the 1st and the 2nd generation of AHSS. 3rd generation AHSS present superior strength and enhanced formability than 1st generation AHSS but at significantly lower production costs [1]. Such AHSS family comprises steels with ultrafine microstructural constituents, such as martensite or bainite, produced in non-equilibrium conditions, in combination with retained austenite (RA) [7-9]. Bainite and martensite contribute to increasing the strength, whereas the stress-induced transformation of RA (TRIP effect) contributes to further optimize ductility and strength [9]. Some of the steels developed under this classification are TBF (TRIP-aided bainitic ferritic) and Q&P (quenching and partitioning) steels. Other 3rd generation TRIP-assisted steels, such as medium-Mn [9,10] or δ-TRIP steels [11], and nanoprecipitation steels [12] are currently under development. Figure 1.3 illustrates the wide range of mechanical properties covered by AHSS in the well-known strength-ductility diagram for steels (also known as “banana” plot). The main characteristics of the above mentioned AHSS, in terms of microstructure, thermomechanical process and mechanical properties are described below. Chapter 1. Introduction 3 Figure 1.3. Strength ductility diagram for various types type of steels, including conventional and AHSS grades [1] 1.1.2.1. Dual Phase Dual phase (DP) steels have a microstructure basically consisting of a soft ferritic matrix with hard martensite islands embedded (Figure 1.4a). They are produced by controlled cooling from the austenite phase (Figure 1.4b, DP1) or from the ferrite + austenite phase (Figure 1.4b, DP2) to transform some austenite to ferrite before a rapid cooling transforms the remaining austenite to martensite. Due to the production process, a small amount of other secondary phases, such as bainite or retained austenite, can also be present in different proportions [13-15]. DP steels are characterized by showing high strength, low yield strength to tensile strength (YS/TS) ratio, high strain hardening and high ductility compared to high strength low-alloy steels [1,15]. The strength level of DP steels is mainly governed by the martensite volume fraction (typical martensite volume fractions are in the range of 20-50% for tensile strengths of 600-1200 MPa). Their high strain hardening is caused by the strain gradients between the soft ferritic matrix and the hard martensite islands during forming, which generate dislocations pile-ups (geometrically necessary dislocations) in the soft matrix to accommodate the plastic incompatibility between the two phases [16]. Such deformation mechanism contributes to increase the work-hardening rate and to delay the onset of localized necking, thus improving the formability. Nevertheless, DP steels are known to be more sensitive to edge fractures [17,18] and show rather low hole expansion capacity [18-20]. Figure 1.4. a) Typical microstructure of DP steels (SEM micrograph). F: ferrite; M: martensite b) Timetemperature cycle applied to obtain the DP microstructure. Ms: martensite start temperature. 1.1 Advanced High Strength Steels 4 1.1.2.2. Complex Phase The microstructure of complex phase (CP) steels consists of ferrite, bainite, martensite and tempered martensite (Figure 1.5a). CP steels are produced by intercritical annealing followed by fast cooling to a temperature above Ms, where it is isothermally held for some time followed by cooling to room temperature (Figure 1.5b). During the first cooling, the intercritical austenite transforms to bainite, and the austenite remaining after the isothermal holding transforms to martensite after the final cooling step [21]. These steels are characterized by high yield strength to tensile strength ratio and low strain hardening. Compared to DP, CP steels exhibit higher yield strengths at equal tensile strengths (≥800 MPa) and lower elongation. On the other hand, they have a great energy absorption capacity, good hole expansion and excellent bendability, which makes them especially suitable for crash-resistant parts [20-22]. Figure 1.5. a) Microstructure of a CP steel (SEM micrograph). B: bainite; TM: tempered martensite b) Timetemperature cycle applied to obtain the CP microstructure. 1.1.2.3. TRIP Transformation-induced plasticity (TRIP) steels have a primarily ferritic matrix with some amounts of bainite and a significant retained austenite (RA) volume fraction (5-20%). Figure 1.6a shows the microstructure of a TRIP steel with a tensile strength of 800 MPa. TRIP steels are produced using a heat treatment similar to that of CP grades (Figure 1.5b). The higher carbon and manganese content of these steels contributes to stabilizing the RA at room temperature [1, 9] and the addition of silicon or aluminium, helps to suppress carbide formation during bainitic transformation. The suppression of carbide formation in the austenite facilitates the carbon enrichment of austenite and increases its stability at room temperature [22]. The bainitic holding time and temperature determine the amount of bainite, as well as the amount and stability of RA. The main advantage of TRIP steels is their increased ductility and strain hardening compared to DP steels of similar strength level (Figure 1.6b). Their high formability and work hardening rates are attributed to the austenite to martensite transformation during deformation (TRIP effect). The beneficial influence of the TRIP effect in mechanical properties is associated with the formation of additional mobile dislocations in ferrite in the vicinity of strain-induced martensite, which increases work hardening and delays the onset of necking. Due to the accompanying volume change, the shear strain leads to an additional increase of the dislocation density. With further straining the fresh formed strain-induced martensite generates more geometrically necessary Chapter 1. Introduction 5 dislocations [9]. The additional strain hardening provides TRIP steels enhanced ductility and strength. Figure 1.6. a) Microstructure of TRIP steel. F: ferrite; B: bainite; RA: retained austenite. b) Engineering stress-strain curves for DP and TRIP steel of similar strength. 1.1.2.4. Ferritic-Bainitic steel Ferritic-Bainitic (FB) steels are a variation of DP steels. They have a ferritic matrix with bainite as a secondary phase in substitution of martensite (Figure 1.7a). The thermal cycle for processing FB steels is similar to that of CP steels (Figure 1.5b). The isothermal holding time determines the amount of bainite present in the final microstructure. Increasing the isothermal holding time increases the bainite volume fraction and decreases the amount of martensite formed during the final cooling step [23]. FB steels show slightly lower strength values and lower strain hardening than ferrite-martensite DP steels. They cover a tensile strength range of approximately 500–900 MPa, with total elongation values from 10 to 30% (Figure 1.7b). The main advantage of FB steels with respect to DP and TRIP steels is the improved edge stretchability. This makes FB steels especially suitable for applications where high stretch flangeability or hole expansion capability is required. Figure 1.7. a) Microstructure of a FB 450/600 [1]. F: ferrite; B: bainite. b) Tensile strength-elongation diagram showing the range of mechanical properties covered by FB steels [1]. 1.1.2.5. Martensitic steel Martensitic steels (MS) are characterized by a martensitic matrix containing small amounts of ferrite and/or bainite (Figure 1.8a). They are produced by quenching at very high cooling rates 1.1 Advanced High Strength Steels 6 from the austenite region. The hard martensitic matrix provides very high yield and tensile strength (up to 1700 MPa) and lower elongation. Post quench tempering treatments improve ductility maintaining high strength. Generally, this kind of steels presents low stretch flangeability and fracture toughness. They are ideal for components of the passenger compartment where high resistance to intrusion is required. Figure 1.8. a) Microstructure of a MS (SEM micrograph). M: martensite. b) Time-temperature cycle applied to obtain martensitic microstructures. 1.1.2.6. Press Hardened steels Press hardened steels (PHS) are extensively used in body-in-white applications, especially for structural components with high anti-intrusion requirements (impact beams, bumper beams, A- and B-pillars, etc.). The press-hardening technique, also known as hot stamping, takes advantage of low flow stress of boron-alloyed steel (22MnB5) in the austenitic phase at elevated temperature and allows the manufacturing of parts with ultrahigh strength and minimum springback issues [24]. The process is schematized in Figure 1.9. The blanks are austenitized at temperatures between 900 and 950 °C for 4 to 10 min inside a furnace and subsequently transferred to a cooled die via an automated transfer system. The blanks are formed at high temperature (650-850 ºC) in a single stroke and cooled down under pressure for a specific amount of time (5-15 s). During this period, the formed part is quenched in the closed die at a cooling rate of 50 to 100 °C/s. The total cycle time for transferring, stamping, and cooling in the die is 15 to 25 s [24]. Figure 1.9. Schematic representation of the press hardening process [25]. Usually, the final microstructure is martensite (Figure 1.10a) with high ultimate tensile strength (1400-1700 MPa) and yield strength (1000-1200 MPa). However, as shown in Figure 1.10b-d, a wide range of final microstructures and mechanical properties can be obtained by controlling the cooling rate in the quenching process [26]. Chapter 1. Introduction 7 Figure 1.10. SEM micrographs of different press-hardened microstructures. a) martensite, b) bainite, c) bainite + martensite and d) ferrite + bainite [26] 1.1.2.7. TWIP Twinning-Induced plasticity (TWIP) steels present a fully austenitic microstructure at room temperature (Figure 1.11), caused by their high manganese content (15-30%). Additional alloying elements such as Si and/or Al are needed to obtain the high strength and large uniform elongation associated with the strain-induced twinning. Al increases the stacking-fault energy (SFE) and, therefore, suppresses the austenite to martensite transformation so that the formation of deformation twins is favoured [27]. In contrast, Si decreases the SFE and sustains austenite to martensite transformation during cooling and deformation [28]. The main deformation mechanisms in TWIP steel are dislocation glide and deformation twinning. Mechanical twins are formed due to the low stacking fault energy. The formation of deformation-induced twins, gradually reduces the dislocation mean free path (Figure 1.11), resulting in a very high strain hardening rate due to the dynamic Hall-Petch effect. This high strain hardening rate allows for the combination of higher strengths and higher uniform elongations [27]. Figure 1.11. a) Microstructure of TWIP steels. A: austenite. b) Illustration of the dynamical Hall-Petch effect [27]. 1.1 Advanced High Strength Steels 8 1.1.2.8. TRIP-assisted bainitic ferritic TRIP-assisted bainitic ferritic (TBF) steels present a multiphase matrix, basically consisting of ferrite and bainite, with significant amounts of metastable RA (5-15%). The replacement of the soft single-phase matrix present in 1st generation TRIP steels by harder ferrite/bainite matrix, allows attaining higher strengths, whereas the good ductility and formability are maintained thanks to the TRIP effect. Figure 1.12a shows the microstructure of a TBF steel. The thermal cycle applied for the production of TBF steels is shown in Figure 1.12b. After full austenitization, the material is cooled down to the overageing temperature (usually about 400°C), at which the isothermal bainitic transformation takes place. The steel composition should contain Si and/or Al in order to prevent the carbide formation in the RA. TBF steels generally exhibit great strain hardening and large elongation values (both uniform and fracture) which makes them suitable for forming operations involving global deformation, such as deep drawing. However, their DP-like microstructures usually show limited local formability (hole expansion, edge stretching) compared to more homogeneous CP-type microstructures [20]. Figure 1.12. SEM micrograph of TBF steel. B: bainite, F: Ferrite; RA: retained austenite. b) Thermal cycle for TBF production [9]. 1.1.2.9. Quenched and partitioned The microstructure of quenched and partitioned (Q&P) steels consists of a tempered martensite/lower bainite matrix (ferrite can be also present in case of partial austenitization) and metastable RA (Figure 1.13a). The Q&P process involves quenching to a temperature (TQ) under the Ms temperature, followed by a ‘partitioning’ treatment either at the initial quench temperature or above. This partitioning step is designed to enrich the remaining untransformed austenite with carbon, escaping from the supersaturated martensite, thus stabilizing RA at room temperature [29, 30]. The presence of hard martensite instead of bainitic ferrite promotes higher strength levels than the ones attained for TBF steels but lower elongation. The more homogeneous CP-like microstructure of Q&P steels makes them more suitable for forming operations with localized deformation, such as bending or edge stretching. These steels represent a good option for antiintrusion structural parts. Chapter 1. Introduction 9 Figure 1.13. a) Microstructure of Q&P steel. RA: retained austenite; TM: tempered martensite; LB: lower bainite. b) Schematic representation of the quenching and partitioning process [9]. 1.1.2.10. Medium Mn steels The first investigations on the development of medium-Mn TRIP steels were carried out by Miller in 1972 [31]. The microstructural constituents of Fe–0.11C–5.7Mn consisted of ferrite and 29 vol.% austenite with a tensile strength of 878 MPa and a total elongation of 34%. Motivated by this success, recent research has focused on medium-Mn alloy design to obtain optimum strength and ductility combinations [32-37]. Medium Mn TRIP steels contain typically from 5-7 wt-% Mn. Their microstructure consists of an ultrafine-grained ferritic matrix with a grain size typically less than 1μm, and a high volume fraction of RA, usually up to 30 vol.-% (Figure 1.14a). The tensile strength of these steels commonly exceeds 1000 MPa along with total elongations in the range of 25 – 40 %. The refined microstructure of medium Mn TRIP steels is obtained by the austenite-reverted- transformation (ART) annealing process (Figure 1.14b). The ART treatment is performed by reheating a strip with initial martensitic microstructure to a certain temperature (Intercritical annealing temperature, TIA) between Ac1 and Ac3 to allow the formation of austenite, followed by quenching to room temperature. C and Mn diffuse from martensite to austenite during isothermal holding, contributing to the strong hardenability of austenite. In addition, Mn segregation at the martensite/austenite interface promotes the growth of austenite [38]. Figure 1.14. a) SEM micrograph of a medium Mn TRIP steel with a chemical composition of Fe–7Mn– 0.14C–0.23Si [38]. b) processing routine of austenite reverted transformation (ART) annealing applied for the production of ultrafine-grained medium-Mn TRIP steels [38]. 1.2 Global and local formability 10 1.1.2.11. δ-TRIP The so-called δ-TRIP steels are a novel concept of TRIP-assisted steel. These steels have a multiphase microstructure consisting of δ-ferrite, bainitic ferrite and RA (Figure 1.15). They have a typical chemical composition of 0.3 - 0.4 wt-% C, 2 - 6 wt-% Al, 0.2 - 0.8 wt-% Si and 0.5 - 1.6 wt-% Mn [39]. The high aluminium addition makes possible a density reduction of up to approximately 5% without sacrificing the Young's modulus [40]. δ-TRIP steels have excellent mechanical properties with tensile strengths of around 800-1000 MPa and elongations between 25- 40% [39-42]. a) b) c) Figure 1.15. SEM micrographs of: a) Fe-0.4C-1.5Mn-5.2Al [39], b) Fe-0.4C-2.5Mn-5.2Al [39] and c) Fe- 0.39C-0.5Mn-3.8Al [41]. 1.2 Global and local formability The continuous development of new complex multiphase AHSS grades has led to the need for alternative formability and fracture performance classification criteria. Owing to their complex microstructures and limited ductility compared to mild steels, AHSS are more susceptible to the occurrence of cracks during cold forming (edge cracking, limited hole expansion ability, etc. [43- 45]) or in situations of severe deformation such as in crash scenario [46-49]. Unfortunately, this kind of fractures cannot be described by traditional ductility and formability definitions based on elongation values from uniaxial tensile tests and limit strains from Forming Limit Curves (FLC), as has been evidenced by several authors [43-51]. Therefore, extensive research is currently being conducted on the identification of different material parameters for a better description of the overall formability and fracture resistance of AHSS. In the last years, different approaches have been proposed and an increasing interest has arisen in the classification of AHSS according to their global and local formability [52]. The term “global” is used to define the deformation modes where relatively large regions of material are deformed simultaneously (stretch forming, drawing, etc.) and strain localization occurs due to the application of a uniform deformation. Therefore, global formability refers to the material’s resistance against the formation of localized necking. On the other hand, “local” formability is more related to the fracture resistance of the material when the deformation is applied in a localized zone (tight-radius bending, stretch flanging, hole expansion, etc.). Chapter 1. Introduction 11 1.2.1. Global formability Global formability can be described by classical uniaxial tensile parameters (true uniform strain, elongation at fracture, n-value) and FLCs. The true uniform strain (εu) is the true strain value corresponding to the percent uniform elongation (UE) in a conventional uniaxial tensile stressstrain curve (Figure 1.16a), and it is calculated according to Equation (1): 𝜀𝑢=ln(1+ 𝑈𝐸 100) (1) The UE corresponds to the elongation at the ultimate tensile strength (UTS). Before the UE, the deformation is homogeneously distributed throughout the tensile specimen. When the UTS is reached, the deformation starts to localize over a length of the order of the specimen width (“diffuse necking”) [54]. For materials that follow a power-law relationship between stress and strain (Equation (2)): 𝜎=K𝜀𝑛 (2) where σ is the true stress, ε is the true strain and K and n are two material constants (strength coefficient and strain hardening exponent, respectively); it is found that the corresponding strain at the onset of necking (εu) is equal to the strain hardening exponent [55]: 𝜀𝑢=𝑛 (3) Accordingly, both parameters εu and n inform about the material’s ability to uniformly distribute the strains and are used to describe the global ductility of a material; the higher they are, the higher the global ductility. Figure 1.16. a) Typical stress-strain curve obtained from a uniaxial tensile test [53]. b) Schematic representation of an FLC [56]. The FLC defines the deformation limits of a material for multiple strain paths, represented by different combinations of major (ε1) and minor (ε2) strains in the so-called Forming Limit Diagram (FLD, Figure 1.16b). These critical strains represent the onset of localized necking and determine the limits below which safety margins are calculated. The strain points below the FLC correspond to the safe zone, where the sheet metal can be formed without risk of localized necking. The points located above the FLC indicate risk of failure. 1.3 Local formability and crack-related failures 18 To characterize the crash energy absorption and anti-intrusion characteristics of AHSS, two main laboratory tests are currently used in the automotive industry, axial impact tests and bending (or side) impact tests, respectively (Figure 1.24). However, these tests are expensive and timeconsuming. Moreover, the results are very sensitive to the experimental setup and the crash sample geometry. This can provide contradictory crash failure behaviours for the same material with different test coupon design and makes difficult the comparison of results [81]. Thus, aimed at improving material selection and optimizing new material design, several attempts have been made to develop alternative small-scale tests and material parameters for crash resistance estimation. V-bending tests according to VDA 238-100 [63] are state of the art to characterize AHSS crash bendability [47-49, 82]. Other approaches based on plane strain fracture strain measurements from V-bending or notched specimens have been also found to be useful for ranking the crash folding ability of AHSS and PHS [48, 83]. Nevertheless, an additional complexity emerges from the crash failure behaviour of very high strength steels (UTS > 800 MPa), which is strongly affected by the nucleation and propagation of cracks (Figure 1.25). Consequently, these parameters are not accurate enough to describe the overall crash resistance of AHSS, since they do not take into account the non-negligible energy spent in crack propagation. Larour et al suggested that the overall crash failure behaviour of AHSS and PHS was mainly dominated by the bendability (resistance to crack initiation) and the fracture toughness (how rapid these cracks propagate through the material). The relationship between fracture toughness and the cracking behaviour of AHSS under impact loading is other of the main points analysed in the present project. Figure 1.24. Experimental setup and specimen geometry for laboratory impact resistance tests at voestalpine Stahl: a) axial impact tests [81]. b) bending impact tests [82]. Chapter 1. Introduction 19 Figure 1.25. Examples of cracks in axial (a, b) and bending (c) crash tested samples of AHSS and PHS. Images from: a) [81], b) [49] and c) [82] 1.4 Fracture toughness of advanced high strength steels The fracture toughness, from a fracture mechanics point of view, is the property that controls the crack initiation and propagation resistance of a material. It is important to differentiate this definition from the conventional use of the term ‘toughness’, referring to the area under the stressstrain curve of a uniaxial tensile test or the product of the ultimate tensile strength by the total elongation (UTSxTE), which is not suitable to describe the material resistance in the presence of pre-existing cracks or defects. Fracture toughness is considered as a key design parameter for engineering applications where structural integrity is of primary importance, such as pipelines in oil and gas industries, nuclear plants, pressure vessels, aeronautics, etc. However, until the rise of AHSS, the fracture toughness has not been considered to be relevant to automotive designers due to the large ductility of conventional mild steels. This fact, together with the complexity of fracture mechanics testing and the absence of affordable standard procedures for fracture toughness characterization of thin metal sheets have generated a gap of knowledge in this field. However, due to the increasingly demanding performance requirements and the frequent occurrence of fractures related to the crack initiation and propagation resistance in AHSS, the knowledge on the fracture toughness properties of high strength metal sheets has become unavoidable. As a sign of the growing interest in this topic, the number of research works related to the fracture toughness of AHSS has significantly increased in the last years [20, 26, 80, 84-91]. Nevertheless, there is still some confusion about which are the most appropriate testing methods for characterizing the fracture toughness of thin high strength metal sheets and how they can be used to understand the cracking resistance of AHSS. Some of the basic fracture mechanics concepts and the main experimental techniques for fracture toughness characterization of ductile sheet materials are reviewed below. 1.4 Fracture toughness of advanced high strength steels 20 1.4.1. Introduction to fracture mechanics Fracture mechanics is the discipline that evaluates the conditions under which a structural component can break due to the existence and growth of a crack in the component. Fracture mechanics covers a broad field of disciplines, from materials science to engineering applications, passing through applied mechanics (Figure 1.26). Materials science addresses the fracture itself at atomistic level and the evolution of the fracture process considering grains, impurities, etc., what is key to understand the behaviour of a crack in a determined stress-strain field. Applied mechanics is found one step beyond and is responsible to characterize and quantify the parameters that describe the crack resistance of the material. In order to successfully use fracture mechanics in engineering applications, it is important to have some knowledge on all these disciplines. Figure 1.26. Science and engineering fields covered by fracture mechanics [92]. In materials with high sensitivity to the presence of cracks or defects, like the case of high strength materials with high yield strength and moderated ductility, engineering fracture mechanics is a useful tool to complement the conventional design criteria based on tensile strength, yield strength or buckling stress. It was developed mainly in the 20th century from the works of Griffith [93] and Inglis [94]. Such authors set the bases for the Linear Elastic Fracture Mechanics (LEFM), which was firmly established in the 1950s. Later, at the end of the 1960s, LEFM was extended to nonlinear problems from the works of Rice [95, 96] and Hutchinson [97], giving rise to the development of the Elastic-Plastic Fracture Mechanics (EPFM). 1.4.2. Linear Elastic Fracture Mechanics The application of LEFM is only valid when there is no significant deformation before the fracture, like is the case of very brittle materials such as glass or ceramics. Griffith established the link between the fracture and the size of the defects, analysing the stresses in an elliptical crack [93]. He developed an energetic approach based on the first law of thermodynamics, using a simple energy balance. According to his theory, a crack is unstable and the fracture takes place when the energetic change associated with the crack growth is sufficient to overcome the material surface energy. However, Griffith’s model is only applicable to ideally brittle solids, since the energy of fracture comes exclusively from the surface energy of the material. The theory assumed that the fracture strength was limited by the existence of initial cracks and that brittle materials contain elliptical microcracks, which introduce high stress concentrations near their tips. He Chapter 1. Introduction 21 developed a relationship between crack length (a), surface energy connected with traction-free crack surfaces (2γ), and applied stress (σ), which is given by Equation (10). 𝜎2= 2𝛾𝐸 𝜋𝑎 (10) Where E is the Young’s modulus. Nevertheless, it was found that the energy required for fracture was much greater for most engineering materials. In 1948, Irwin [98] and Orowan [99] independently presented an extension to Griffith’s theory, proposing that the total energy required for crack growth comes from surface energy and an irreversible plastic work close to the crack tip. This approach allowed to extend Griffith’s model to ductile materials. The criterion was that the strain energy release rate, G, must be larger than the critical work, Gc, which is required to create a new unit crack area. Using Westergaards’ method to analyse stresses and displacements at the crack tip, Irwin showed that the stress field in the area of the crack tip is completely determined by a quantity K, called the stress intensity factor, as follows [100]: 𝜎𝑖𝑗 = 𝐾𝑓𝑖𝑗(𝜃) √2𝜋𝑟 (11) The relation between G and K is described in Equation (12): 𝐺=𝐾2 𝐸′ (12) Where E’=E for plane stress and E’ = E/(1- ν2) for plane strain and ν being the Poisson’s ratio. 1.4.3. Loading modes There are three primary modes of crack loading (Figure 1.27), Mode I, Mode II and Mode III. In Mode I loading (opening), the principal load is applied normal to the crack plane and tends to open the crack. Mode II (in-plane shear) involves a shear loading that tends to slide the crack faces in the direction parallel to the primary crack dimension. Mode III (out-of-plane shear) loading or tearing mode involves a shear load sliding the crack faces in the direction perpendicular to the primary crack dimension. A cracked body can experience any of these loading modes or a combination of two or three modes. However, loading Mode I is the most used one for stress analysis of cracks since it is the predominant loading mode in engineering applications and most crack failures occur under this opening mode. 1.4 Fracture toughness of advanced high strength steels 22 Figure 1.27. Loading modes that can be applied to a crack 1.4.4. Crack tip triaxiality: plane stress vs plane strain As illustrated in Figure 1.28, the stress triaxiality at the crack tip varies along the thickness direction. In the mid-thickness of the cracked plate, the material at the crack tip tries to deform in the x (crack advance direction) and z (thickness direction) axes but it is constrained by the surrounding material. This generates a high stress triaxiality (plane strain) in the mid-thickness. The stress in the thickness direction, σzz, is gradually reduced to zero towards the outer freesurfaces (Figure 1.28 right), resulting in a biaxial stress state (plane stress). The difference in plastic constraint between the mid-thickness and the outer surface has a direct influence on the size of the plastic zone surrounding the crack tip, which gradually decreases from the plane stress region (outer surface) to the mid-thickness in plane strain (Figure 1.29). The relative size of the plastic zone, rp, respect to the plate dimensions determines the dominating stress state at the crack tip. When the plastic zone is large compared to the plate thickness (rp/B > 0.5), such as the case of thin plates, the material at the crack tip can deform in the thickness direction. In this case, plane stress conditions prevail. On the other hand, if the plastic zone is relatively small (rp/B < 0.02), the deformation in the thickness direction is constrained and the crack tip is in a plane strain condition. In the intermediate range between the two conditions, a mixed plane stress/plane strain mode occurs at the crack tip [103]. Figure 1.28. Three-dimensional deformation at the crack tip of a cracked plate subjected to in-plane loading (left) and schematic variation of transverse stress through the thickness at a point near the crack tip (right) [101]. Chapter 1. Introduction 23 Figure 1.29. Three-dimensional plastic zone at the crack tip in a finite plate [102]. 1.4.5. Influence of thickness on fracture toughness The stress intensity factor, K, allows determining the conditions under which the material fractures. The condition for fracture is that the stress intensity factor reaches a critical level, Kc, which represents the fracture toughness of the material. As observed in Figure 1.30a, Kc depends on specimen thickness until a plateau is reached for a determined thickness (tc). Above this thickness value, fracture toughness becomes insensitive to the specimen thickness and it is designated with the symbol KIc. KIc is referred to as the plane strain fracture toughness in loading mode I and it is a size-independent material property. On the other hand, in the thickness-dependent region, Kc reaches a maximum value for a relatively low thickness (to). Below t0, plane stress conditions prevail and toughness tends to decrease with decreasing specimen thickness. Above t0, a mixed plane stress/plane strain mode occurs and toughness decreases as thickness increase until it reaches the thickness tc. The influence of thickness on fracture toughness is related to the relative portions of flat and shear fracture (Figure 1.30b). Very thin sheets typically exhibit a 45º shear fracture (slant fracture). When increasing thickness, the fracture shows a combination of shear fracture in the outer surfaces and flat fracture in the central region (tunnelling effect). For further thickness increase, flat fracture mechanisms, associated with plane strain conditions, dominate and thickness has no significant influence on toughness. Figure 1.30. a) Thickness-dependence of fracture toughness. b) Effect of specimen thickness on ductile fracture surface morphology. Redrawn from [101]. 1.4.6. Fracture toughness evaluation in the frame of linear elastic fracture mechanics Many efforts have been dedicated to the standardization of experimental methodologies to properly evaluate the plane strain critical stress intensity factor, KIc, through reproducible and 1.4 Fracture toughness of advanced high strength steels 24 reliable tests. The most extended and applied standard is the ASTM E399 [104]. Different specimen geometries are proposed in this standard, being the most used the Compact Tension (CT) and the Single Edge Notched Bending (SENB) geometries (Figure 1.31). Figure 1.31. Standardized specimens for fracture toughness determination: CT specimen (left) and SENB specimen (right) [105] In order to achieve a sharp crack with the minimum radius possible the standard recommends introducing fatigue pre-cracks in the notched specimens. After that, the specimens are tested and the load (P) as a function of the Crack Opening Displacement (COD) is recorded. The stress intensity factors for such geometries is given by Equation (13) 𝐾𝐼=𝑃 𝐵√𝑊 𝑓(𝑐 𝑊) (13) Where 𝑓(𝑐 𝑊) is a geometrical factor that depends on specimen geometry. Some constraints are required to guarantee the plane strain conditions. It is proposed that the crack length, c, the ligament, b, and the specimen thickness, B, must be, at least, fifty times larger than the radius of the plastic zone in plane strain. 𝑐,𝑏,𝐵≥2.5(𝐾𝐼𝐶 𝜎𝑦𝑠)2 (14) Table 1.1 shows characteristic fracture toughness values for different materials. Table 1.1. Fracture toughness values for different materials [105]. Material 𝝈𝒚𝒔 𝑴𝑷𝒂 𝑲𝑰𝑪 𝑴𝑷𝒂√𝒎 𝑴𝒊𝒏𝒊𝒎𝒖𝒎 𝒕𝒉𝒊𝒄𝒌𝒏𝒆𝒔𝒔,𝒎𝒎 Maraging steel 1500 100 12 Low carbon steel 240 200 1750 304 300 180 900 Al 7075-T651 550 31 8 Al 2024-T3 400 34 18 Ti alloy 6Al-4V 1100 40 4 Ti alloy 4Al-4Mo-2Sn-0.5Si 950 70 13,5 Chapter 1. Introduction 25 1.4.7. Elastic-Plastic Fracture Mechanics When considerable plastic deformation occurs before fracture and the size of the crack-tip plastic zone is comparable to the crack length or specimen dimensions, LEFM is no longer valid. In such case, elastic-plastic fracture mechanics (EPFM) allows characterizing the stress state at the crack tip using plastic properties such as yield strength and hardening coefficient. Thus, EPFM can be considered an extension of LEFM. The parameters used to describe the crack-tip conditions in elastic-plastic materials are described below. 1.4.7.1. Crack Tip Opening Displacement During fracture toughness tests with high-toughness structural steels, Wells [106] observed that plastic deformation had blunted the crack tip before the fracture. The degree of crack blunting increased with increasing the toughness of the material. Consequently, Wells proposed the opening at the crack tip as a parameter to characterize the fracture toughness when LEFM is no longer applicable. Such parameter is known as the crack tip opening displacement (CTOD) and it is represented by the symbol δ (Figure 1.32). Figure 1.32. Initially sharp crack blunts before fracture due to plastic deformation, causing the displacement δ in the crack tip (CTOD) [101]. The CTOD criterion establishes that the Mode I crack initiation occurs when the CTOD reaches a critical value, δ= δc. δc can be used as a design parameter to predict the failure of a cracked structure. Wells [106] performed an approximate analysis that related CTOD to the stress intensity factor in the limit of small-scale yielding. Using Irwin’s plastic zone model, the CTOD can be calculated as follows [107]: 𝛿= 4 𝜋𝐸 𝐾𝐼 𝜎𝑦𝑠 2 (15) where E is the Young’s modulus, KI is the stress intensity factor in mode I and σYS is the yield stress. By definition, the CTOD is the opening displacement at the sharp crack tip (Figure 1.32). However, alternative definitions have been proposed. The most common alternative definition was proposed by Rice [95], which defined the CTOD as the displacement at the intersection of a 90° vertex originated at the tip of the blunted crack with the crack faces (Figure 1.33). This definition is commonly used to determine the CTOD in finite element simulations. 1.4 Fracture toughness of advanced high strength steels 26 Figure 1.33. CTOD defined as the displacement at the intersection of a 90° vertex with the crack faces [101]. 1.4.7.2. Crack Tip Opening Angle Together with the CTOD, the Crack Tip Opening Angle (CTOA) is often used to describe the stable crack growth of ductile materials. The CTOA (ψ) is defined as the relative angle of the crack surfaces at a given distance (r0) behind the current crack tip (Figure 1.34). It can be calculated as follows: 𝐶𝑇𝑂𝐴=𝜓=2𝑡𝑎𝑛−1(𝛿0 2𝑟0) (16) where δ0 is the crack opening displacement at the distance ro. The CTOA reaches a steady value during the stable crack propagation. This critical CTOA (ψc) is used to define the stable crack propagation resistance of ductile materials. Figure 1.34. CTOA (ψ) definition. 1.4.7.3. J-integral The J-integral concept was introduced by Rice [95] to characterize the crack tip strain fields in nonlinear elastic materials. Since, in most cases, the monotonic loading behaviour of a nonlinear elastic and an elastic-plastic material is identical, Rice idealized the plastic deformation as nonlinear elastic behaviour and applied the deformation plasticity theory to the analysis of a crack in an elastic-plastic material. Then, he showed that the nonlinear energy release rate J could be written as a path independent contour integral: 𝐽=∫(𝑤𝑑𝑦−𝑇𝑖𝛿𝑢 𝛿𝑥 Γ𝑑𝑠) (17) Chapter 1. Introduction 27 where  is an arbitrary contraclockwise path around the crack tip (Figure 1.35), w is the strain energy density, Ti are the components of the traction vector, defined according to the outward normal along  Ti=σijnj), u is the displacement vector and ds is an element of arc length along  . The strain energy density w is defined according to Equation (18): 𝑤=∫𝜎𝑖𝑗𝜀𝑖𝑗 𝜀𝑖𝑗 0 (18) where σij and εij are the stress and strain tensors, respectively. The works of Hutchinson [97] and Rice and Rosengren [96] related the J-integral parameter to the crack tip stress fields in nonlinear materials (HRR singularity). Thus, it can be considered as both an energy parameter and a stress intensity parameter. The parameter J is a more general version of energy release rate, G. For the case of a linear elastic material J=G. 𝐽=𝐾2 𝐸′ (19) This energy release rate definition was used in the early works of Begley and Landes [108, 109] to experimentally measure J in elastic-plastic materials. Since these first experimental works, J , in particular the critical J value at the onset of stable crack propagation, Jc, has been the most extended parameter to evaluate the fracture toughness of ductile materials. Figure 1.35. Arbitrary contour around the tip of a crack [101]. 1.4.7.4. Relationship between J and CTOD The relationship between J and the CTOD was proposed by Rice [95] and later revised by Shih [110], among others. Shih evaluated the displacements at the crack tip implied by the HRR solution and related the displacement at the crack tip to J and flow properties: 𝛿=𝑑𝑛𝐽 𝜎𝑌𝑆 (20) where dn is a constant depending on σYS /E and the strain hardening exponent, n [110]. This relation between J and δ shows that these two parameters are both valid for characterizing the crack-tip conditions of elastic-plastic materials. 1.4.8. Fracture toughness evaluation in the frame of elastic-plastic fracture mechanics Different methodologies are available to characterize the fracture toughness of ductile materials within the frame of EPFM. The most extended methodology is the measurement of the J-integral 1.4 Fracture toughness of advanced high strength steels 34 𝑊𝑓 𝑙0𝑡0=𝑤𝑓=𝑤𝑒+𝑤𝑝𝛽𝑙0 (31) Using Equation (31), we is determined by testing up to fracture a series of specimens with different ligament lengths (l0) and plotting wf values as a function of l0. we and wpβ can be obtained by linear regression, where we is given by the intercept and wpβ by the slope, as shown in Figure 1.41. wf values are obtained by integrating the area under load vs displacement curves (Wf) and dividing by the initial cross-section area. It must be noted that we cannot be considered an intrinsic material property since it has an important contribution from necking. Thus, it is a material constant for a given sheet thickness. we has shown to be independent of the specimen geometry and can be obtained from different geometries [135-137]. However, for thin sheets, the EWF test protocol [138] developed by the European Structural Integrity Society (ESIS) recommends the use of Double Edge Notched Tension (DENT) specimens (Figure 1.40) because of its symmetry and minimal specimen rotation and buckling during the test. Figure 1.40. DENT specimen used for the evaluation of the EWF (left) and definition of the Fracture Process Zone (right). Figure 1.41. Experimental procedure for the determination of the essential work of fracture, we and the specific work for fracture initiation, wei. Chapter 1. Introduction 35 Some restrictions must be met to use Equation (31). First, the ligament area must be completely yielded before crack initiation and ligament must be small enough to avoid the spreading of the plastic zone to the boundaries of the specimen. Moreover, ligament length must be large enough to ensure that all specimens present a global plane stress state across the ligament. If the ligament is too small, the process zones at the two crack tips may interact and crack propagation takes place under a mixed plane stress/plane strain state that varies with the ligament length (Figure 1.42). To satisfy these requirements some limitations regarding the ligament length have been proposed. The ESIS protocol [138] recommends a lower ligament length of max (3t or 5 mm), where t is the specimen thickness. This lower boundary is intended to ensure that data is obtained under plane stress conditions. However, it has been found that this transition from plane stress to a mixed stress probably depends on the material characteristics and thickness. Therefore, the recommendation of 3t or 5 mm is based on experimental observations [135-137, 140]. For the maximum ligament length, the upper bound min (W/3, 2rp) is given. Where W is the specimen width and rp is the size of the plastic zone, calculated according to Equation (32) [140]: 𝑟𝑝=1 2𝜋 𝐸𝑤𝑒 𝜎𝑦𝑠 2 (32) Where E is the Young modulus and σys the yield strength. As indicated in the EWF test protocol [138], W/3 is arbitrarily included to avoid edge effects and the condition 2rp is to ensure that the ligament is fully yielded before fracture initiation. Figure 1.42. Valid ligament lengths for the evaluation of the EWF [139] As shown by Mai and Cotterell [136], the EWF methodology also allows to separate energetic contributions from crack initiation and crack propagation and determine a crack initiation toughness value, the specific essential work for fracture initiation, wei [136]. The specific work for fracture initiation, wf i is calculated by integrating the area under the load vs displacement curve until the onset of crack propagation, as illustrated in Figure 1.41. Contrary to wf, wfi is constant and independent of the ligament length. The EWF methodology has been extensively applied to characterize the fracture toughness of polymers [135-137, 141, 142], ductile metal [118, 119, 139, 143] and high strength steel sheets [20, 26, 84-87, 91]. Table 1.6 shows some published we values for thin metal sheets. 1.4 Fracture toughness of advanced high strength steels 36 Table 1.6. Reported we values for metal sheets. Material Thickness (mm) we (KJ/m2) TRIP steel (8% austenite vol. fraction) [84] 0.9 270 TRIP steel (24% austenite vol. fraction) [84] 0.9 50 CP steel 1000 MPa UTS [20] 1.4 405 DP steel 1000 MPa UTS [20] 1.4 138 Trip-aided Bainitic Ferritic (TBF) steel [20] 1.4 150 Quenched & Partitioned (Q&P) steel [20] 1.4 194 Press Hardened steel 1500 MPa [20] 1.5 159 Press Hardened steel 1000 MPa [20] 1.5 249 Different authors have addressed the equivalence between we and Jc [119, 136, 137]. Mai et al [136, 137] proposed that, under strictly J-controlled crack growth resistance we=Jc. The JR curve is approximately a linear function of crack growth and is given by: 𝐽𝑅=𝐽𝑐+ 𝑑𝐽 𝑑𝑎∆𝑎 (33) The integral of JR to complete the fracture is the specific work of fracture, wf, mentioned in equation (31). Equation (33) is very similar in form to the obtained in equation (31) from the energy partitioning, where Jc would correspond to we and 𝑑𝐽 𝑑𝑎 to the wpβ term. The equivalence between we and Jc is further discussed in this work. 37 Chapter 2 Objectives and scope 2.1 Objectives of the work The main objectives of this thesis work are two: first, to assess the applicability of different fracture mechanics testing methods and its derived fracture toughness parameters to characterize the crack initiation and propagation resistance of AHSS sheets and, second, to investigate how fracture toughness can be used to understand and rationalize the cracking behaviour of AHSS. A secondary objective of this research is to gain fundamental knowledge on the influence of microstructure on the fracture toughness of AHSS. Aimed at reaching these primary goals, the following specific objectives are defined: 1. To define the most suitable experimental methods and specimen geometries for characterizing the fracture toughness of thin AHSS sheets: The first step was to choose the most appropriate fracture mechanics testing procedures. The fracture toughness of the investigated materials was primarily characterized following three methodologies: essential work of fracture (EWF), J-integral and Kahn-type tear tests. The EWF method (Papers I-IV) was chosen as the main experimental technique due to its simplicity and reliability. In Paper I, a comparison of the three testing methods is made. 2. To evaluate the influence of experimental variables on the measure of fracture toughness: Different testing variables such as the specimen geometry or the notch radius on fracture toughness measurements are assessed in Paper I. An alternative rapid procedure for specimen notching is also presented in the appended Paper D. 3. To identify the fracture toughness parameters that best describe the crack initiation and propagation resistance of AHSS sheets: The relationship between the different crack initiation and propagation resistance parameters (wei, we, Ji, Jc, UIE, UPE) is discussed in Paper I. 4. To investigate the relationship between fracture toughness and the cracking resistance of AHSS: The correlation of fracture toughness with edge cracking resistance and crash failure behaviour is analysed in Papers II and III, respectively. Further investigations on the correlation fracture toughness-cracking resistance are presented in Papers A-C. 5. To analyse the influence of the microstructure on fracture toughness of AHSS: The effect of the microstructure on the crack propagation resistance of AHSS is discussed in Paper II and Paper IV. Paper II briefly discusses the relation between microstructure and fracture toughness in several 1st and 3rd Generation AHSS grades. Paper IV investigates in detail the influence of the microstructural constituents on fracture resistance of high strength dual-phase steels. Additionally, synchrotron X-Ray Powder Diffraction measurements were performed to investigate the evolution of the retained austenite fraction during the crack propagation of TRIP-assisted steels. Some preliminary results are presented in Chapter 4. The outcome of this research will contribute to filling the gap of knowledge in fracture toughness properties of thin AHSS sheets. This knowledge will be useful not only to optimize material selection for automotive applications but also to design new AHSS with enhanced mechanical properties and cracking resistance. 2.2 Scope of the research 38 2.2 Scope of the research This thesis work is focused on the experimental evaluation of the mode I plane stress fracture toughness of several AHSS and PHS sheets (thicknesses from 1-1.6) mm and on the application of fracture mechanics to understand their edge formability and cracking behaviour in crash scenario. The numerical modelling of the ductile fracture process is out of the scope of this thesis. The work covers a wide range of AHSS and PHS grades with strengths ranging from 780 to 1500 MPa. Other materials are not considered. However, the findings obtained as a result of the research could be easily extrapolated to other high-strength sheet materials, such as high strength aluminium alloys, stainless steels, etc. The proposed experimental techniques and their limitations are described below: Fracture toughness characterization Fracture toughness is measured in the frame of fracture mechanics using three different experimental methodologies, namely, the EWF, the J-integral and the Kahn-type tear test. The different fracture toughness parameters obtained from these tests are compared and discussed. To avoid the thickness limitations of ASTM E1820 for J-integral evaluation, different specimen geometries are proposed. Moreover, alternative methods for direct crack growth measurement by using optical methods are used. It is important noting that all the fracture toughness parameters evaluated in this work correspond to values of plane stress fracture toughness and, therefore, depend upon specimen thickness. The influence of the sheet thickness on the measured toughness values and the determination of the thickness independent plane strain fracture toughness (JIc) are not addressed. Edge cracking resistance and crash failure behaviour Edge cracking resistance of the investigated materials is characterized by means of ISO 16630 hole expansion tests (HET) and DIC-assisted hole tension tests (HTT). To minimize the inherent scatter of the hole expansion ratio (HER) the onset and propagation of cracks are determined online by means of a digital video camera. The use of the DIC in HTT is a valuable tool to determine the fracture strains near the edge of a punched hole. However, these fracture strains are quite sensitive to mesh quality and DIC parameters (camera resolution, facet size, step size, etc.). Other of the experimental drawbacks of these optical strain analysis is the difficulty on accurately measure strains at the edge, due to the poor strain resolution. Crash fracture behaviour is assessed through laboratory-scale axial and bending impact resistance tests. All the crash tests were performed at voestalpine Stahl laboratories in Linz, Austria. Crashworthiness is determined according to the intrusion level, overall cracking behaviour and energy absorbed. It is worth emphasizing the experimental complexity of these tests, especially of axial crash tests, which can be highly influenced by different extrinsic parameters: crash box geometry, welds, etc. Microstructural analysis Basic microstructural investigations for most of the investigated steels include light optical microscopy (LOM) and scanning electron microscopy (SEM) analysis. A detailed microstructural investigation on two GigaPascal DP steels is also performed by means of LOM, SEM, Electron backscatter diffraction (EBSD) and nanoindentation measurements. Chapter 2. Objectives and scope 39 Due to the limitations of EBSD in the evaluation of retained austenite (RA) volume fraction, alternative measurements like the saturation magnetization method or synchrotron X-Ray Powder Diffraction measurements were used in some of the investigated materials. 2.2 Scope of the research 40 41 Chapter 3 Materials and experimental methods The materials investigated and the experimental procedures are described in the appended papers. The following sections provide further experimental details and describe the procedures for the evaluation of additional parameters not included in the papers, such as the critical crack opening displacement (δc) from EWF tests or the true major strain at fracture in the notch tip from Kahntype tear tests (εf KTT). A basic description of the advanced microstructural characterization techniques is also given. 3.1 Materials A wide range of 1st, 2nd and 3rd generation AHSS were investigated in this work. The mechanical properties and chemical compositions of the studied steels are summarized in Table 3.1 and Table 3.2, respectively. Table 3.1 also indicates the sheet thickness and the paper where they are investigated. The microstructural constituents are described in Table 3.3. The steels can be divided into four main microstructural groups:  Complex-Phase (CP) like microstructures with homogeneous bainite/tempered martensite matrix: CP1000, Q&P, 3rd Gen Q&P1180, CP1200, TBF/Q&P, PHS1000  Dual-Phase (DP) like microstructures with soft (ferrite/bainitic ferrite) matrix and hard (martensite) secondary phases: DP1000, TBF, DP780, TRIP780, DP1000A, 3rd Gen DP1180, 3rd Gen TBF1180  Fully martensitic microstructures obtained by press hardening: PHS1500  Fully austenitic microstructures (high Mn steel): TWIP Within these groups, a subgroup of TRIP-assisted microstructures (retained austenite: 6-16%) can be identified: TBF, Q&P, DP780, TRIP780, 3rd Gen steels, TBF/Q&P. Table 3.1. Mechanical properties of the investigated steels. YS= yield strength; UTS= ultimate tensile strength, n: strain hardening exponent, UE: uniform elongation; TE: total elongation; t=sheet thickness. Paper Steel t [mm] YS [MPa] UTS [MPa] n [-] UE [%] TE [%] Paper I, Paper III CP1000 1.4 915 1008 0.05 4.8 8.8 DP1000 1.4 775 1015 0.07 7 11.4 TBF 1.5 755 1012 0.10 10.5 15.8 Q&P 1.4 920 1202 0.05 5.3 9.1 Paper II DP780 1.5 513 823 0.2 14.2 19.9 TRIP780 1.6 542 851 0.2 20.7 25.8 Paper II/ Paper IV DP980/DP1000A* 1.35 816 1055 0.13 6.54 9.7 Paper II 3rd Gen DP1180 1.2 895 1212 0.15 10.5 14.3 3rd Gen TBF1180 1.4 987 1216 0.11 9.2 12.6 3rd Gen Q&P1180 1.5 1034 1191 0.09 9.2 13.1 Paper III CP1200 1.6 1041 1218 0.05 3.4 6 TBF/Q&P 1.4 876 1026 0.09 7.5 11.3 PHS1500 1.5 1075 1552 0.08 3.7 5.2 PHS1000 1.5 988 1007 0.05 4.9 7.3 TWIP 1.4 530 969 0.11 55 59.5 Paper IV DP1000 B 1.4 773 1040 0.09 5.4 8.7 * This steel was designated DP980 in Paper II and DP1000 A in Paper IV. Hereinafter, this steel will be referenced as DP1000A. 3.1 Materials 42 Table 3.2. Chemical compositions (in weight per cent). Steel C Si Mn Cr B Al Ti CP1000 0.11 0.34 ~2.3 0.12 0.0017 0.040 - DP1000 0.19 0.18 ~2.3 0.46 0.0003 0.048 - TBF 0.20 0.84 >2.4 0.17 0.0003 0.039 - Q&P 0.12 0.81 >2.4 0.18 0.0002 0.043 - DP780 ~0.1 <0.9 <2.0 <0.7 <0.003 ~0.05 <0.0060 TRIP780 ~0.20 ~1.60 ~1.70 ~0.02 <0.001 ~0.05 ~0.0070 DP1000A ~0.15 <0.5 ~2.3 <0.7 <0.003 ~0.05 <0.0060 3rd Gen DP1180 ~0.20 <2.0 ~2.5 <0.7 <0.003 ~0.05 <0.0060 3rd Gen TBF1180 ~0.23 <2.0 <2.9 <0.7 <0.005 ~0.04 ~0.0070 3rd Gen Q&P1180 ~0.18 <2.0 <2.9 <0.7 <0.005 ~0.03 ~0.0060 CP1200 ~0.15 <0.5 1.8-2.2 <0.7 <0.003 - - TBF/Q&P ~0.10 0.5-1.0 2.2-2.6 <0.7 <0.003 - - PHS1500 ~0.20 ~0.20 ~1.20 <0.7 ~0.003 ~1.0 - PHS1000 ~0.20 ~0.20 ~1.20 <0.7 ~0.003 ~1.0 - TWIP ~0.50 0.10 -0.15 ~15 ~0.10 - ~1.0 - DP1000 B 0.08 0.26 ~2.6 0.31 0.0018 0.16 0.0372 Table 3.3. Microstructural constituents. F: ferrite, B: Bainite, BF: Bainitic Ferrite, M: Martensite, TM: Tempered martensite, RA: retained austenite. UB: Upper bainite, LB: Lower bainite. Steel Microstructural constituents RA volume fraction, Vγ [%] CP1000 B/TM matrix 1.3 DP1000 F/B matrix, TM, M islands, RA 4.3 TBF F/B matrix, TM, M islands, M/RA 11.2 Q&P B/TM matrix, RA 6.0 DP780 F/B matrix, M/RA islands 9.8 TRIP780 F/B matrix, M/RA islands 15.6 DP1000A F/B matrix, TM, M islands, RA 5.5 3rd Gen DP1180 UB/LB matrix, M/RA islands and laths 14.8 3rd Gen TBF1180 Carbide-free B matrix, M/RA islands and laths of RA 15.5 3rd Gen Q&P1180 TM matrix, B, M/RA islands and laths of RA 12.6 CP1200 B/TM 1.4 TBF/Q&P B/TM matrix, RA 8.4 PHS1500 M 3.3 PHS1000 TM 2.3 TWIP Austenite 100 DP1000 B F/B matrix, M islands 2.0 3.2 Fracture toughness measurements 3.2.1 Essential Work of fracture EWF tests were performed using rectangular Double Edge Notched Tension (DENT) specimens of 240 x 55 mm. Initial notches were machined by electrical discharge machining (EDM). Then, fatigue pre-cracks were nucleated at the notch root following the recommendations of the ASTM E1820. The cracks were extended about 1-1.5 mm per side. All the DENT specimens were machined at 90º with respect to the rolling direction (Figure 3.1a). For fatigue pre-cracking, a Rumul resonance fatigue machine was used (Figure 3.2a). The machine has a load cell capacity of 150 kN and it is able to work at frequencies up to 250 kHz. Fatigue pre-cracking was conducted under load (P) control in a resonance fatigue machine. Tests were run at room temperature at a constant axial load ratio, R= Pmin/Pmax=0.1 (tension–tension). Chapter 3. Materials and experimental methods 43 The ΔK (Kmax-Kmin) was kept below 0.3 Kc, where Kmax and Kmin are, respectively, the maximum and minimum stress intensity factor applied and Kc is the linear elastic fracture toughness at crack initiation. This condition was verified after the test. EWF tests were performed at a 250 kN universal testing machine (INSTRON 5585H), equipped with a digital video extensometer (Figure 3.2b). The tests were carried out according to the European Structural Integrity Society (ESIS) protocol for EWF testing [138]. Generally, 5 different ligament lengths (l0) ranging from 5 to 15 mm were used and about 3 specimens per ligament length were tested. The specimens were tested up to fracture at a constant cross-head speed of 1 mm/min. The load-line displacement was measured by means of the video extensometer using initial extensometer marks separated 50 mm. For the determination of the crack initiation, a high-resolution video camera synchronized with the testing machine was used. wfi values were determined for two specimens of each ligament length. Fracture toughness at cracking initiation, wei, was evaluated from the average of wfi values. In order to check whether the ligament was fully yielded before crack initiation and ensure the validity of the EWF measurements, a full field strain analysis was performed at the surface of the ligament area (Figure 3.2c). For this task, a speckle pattern was painted in the specimen surface and a Digital Image Correlation (DIC) equipment was used (Figure 3.2d). Images were recorded at a frame rate of 5 images/second and analysed in GOM Correlate software. For the DIC analysis, a facet size and a step size of 13 and 11 pixels respectively were used. To determine the strain level from which the material is yielded, DIC-assisted uniaxial tensile tests were previously performed for each material. The DIC was used to evaluate the Equivalent Mises strain at the onset of the plastic region (after σys). For all the investigated steels, the plastic regime was found to start approximately at a Mises strain= 0.005. Figure 3.1. DENT specimen used for EWF tests and detail of the fatigue pre-crack at the notch root. 3.4 Crash behaviour characterization 50 Figure 3.12. DIC strain mapping and sections used for εf HTT determination 3.4 Crash behaviour characterization 3.4.1 Axial impact resistance tests Axial impact resistance tests were performed using an impact simulator, which works with a load mass of 283 kg and whose speed can be varied between 10 and 40 km/h. Some images of the experimental setup for axial crash tests at voestalpine Stahl are shown in Figure 1.24a. Impact tests were performed at different crash speeds and, thus, different intrusion levels. The deformation of axially crashed sample (initial length: 300 mm) was limited to 200 mm shortening For these tests, hat profile specimens were used (Figure 3.13). Profiles were spot-welded to a closing blank of the same steel type. Also at the front and back, closing blanks were welded to the profile. Impact resistance was evaluated by determination of an overall crash index (CI, definition in Table 3.4) and the energy absorbed for each deformed crash sample. Intrusions of axial impact tests include only plastic deformation and were determined by the difference in length of unloaded (300 mm) and crashed samples. The energy absorbed during axial impact tests was calculated by integrating the area under the load vs impactor displacement curves (Figure 3.13b). To evaluate the evolution of the CI as a function of the intrusion level, a new parameter is introduced, the Crash Index Decreasing Rate (CIDR). The CIDR quantifies the damage in impact resistance tests and can be understood as a crack propagation rate. Figure 3.13. a) Geometry of axial crash samples with 1.5 mm blank thickness. b) Force vs impactor displacement curves for two axial impact tests with DP1000. Chapter 3. Materials and experimental methods 51 Table 3.4. Definition of crash index for axial impact tests [47]. Crash index (CI) Damage 100 no cracks >75 crack length < 10 mm 50-75 10 mm < crack length < 25 mm 25-50 crack length > 25 mm <25 "splitting and curling"; multiple breaks 3.4.2 Bending impact resistance tests The experimental setup for bending crash tests is shown in Figure 1.24b. The geometry of the samples for bending impact tests can be seen in Figure 3.14. A microalloyed steel grade with thickness 1.5 mm was used as a closing blank of the hat profile. The welding spots joining the closing blank and the crash profile are 8 mm in diameter, with a distance of 30 mm between the welding spots for both axial and bending impact resistance tests. Bending impact tests were conducted at speeds varying from 20 to 30 km/h. The load mass was 86 kg. The maximum bending displacement was limited to 211 mm. Crash index (CI) of bending impact samples was derived from the occurring crack length and calculated as defined in [48]. The formula for calculating the crash index is shown in Figure 3.15. The deformation process during the bending impact test was recorded with a high-speed camera. Intrusions of bending impact tests in the following content include plastic and elastic deformation. Bending impact samples do not show any welding spot failure, which means that the corresponding results characterize the behaviour of the base material. Figure 3.14. Geometry of bending impact test samples with 1.5 mm blank thickness. Length: 900 mm. 3.5 Microstructural characterization 52 Figure 3.15. Crack location and definition of crash index for bending impact test. 3.5 Microstructural characterization 3.5.1 SEM and EBSD Microstructural investigations were performed by means of Light Optical Microscopy (LOM) and Scanning Electron Microscopy (SEM). The optical microscope is able to reach up to 1000 magnifications and it is equipped with the image treatment software ANALYSIS. A highresolution SEM (Figure 3.16a) equipped with a high-resolution electron backscatter diffraction (HR-EBSD) detector was used. In Paper IV, HR-EBSD measurements were performed to investigate the microstructure of two high strength dual phase steels. In EBSD measurement, a stationary electron beam interacts with a tilted (70º from horizontal) crystalline sample. The diffracted electrons form a pattern (Figure 3.16b) that can be detected with a phosphor screen coupled to a compact lens, which focuses the image from the fluorescent screen onto the CCD camera. The diffraction pattern is characteristic of the crystal structure and orientation. Hence the diffraction pattern provides information about the structure, grain size, grain boundary nature, grain orientation, texture and phase identifications. For HR-EBSD measurements, the samples were mechanically polished to mirror surface finish with a 0.05 µm colloidal silica suspension. HR-EBSD measurements were performed at 20 kV with a step size of 0.15 µm. The analysed areas were 311 x 231 µm2. From these measurements, the inverse pole figure (IPF) map, the phase map and the mean angular deviation (MAD) map were used to characterize the microstructure. Chapter 3. Materials and experimental methods 53 Figure 3.16. a) Scanning Electron Microscope (SEM). b) Formation of diffraction patterns (EBSP). 3.5.2 Nanoindentation tests Nanoindentation technique was used in Paper IV to investigate the distribution of hard secondary phases in the two investigated dual phase steels. Nanoindentation measurements were performed using an iNano nanoindenter (Figure 3.17a) and a Berkovich indenter. In each sample, three arrays of 25x25 indentations were performed. An indentation separation of 6 µm was chosen, covering an area of 150x150 µm2 (Figure 3.17b). The indentations were performed at an applied load of 20 mN. Hardness values were evaluated using the Oliver and Pharr methodology [146]. Figure 3.17. a) iNano nanoindenter. b) SEM image showing a nanoindentation array. 3.5.3 Synchrotron X-Ray Powder Diffraction X-ray powder diffraction measurements were performed in three 3rd Generation TRIP-assisted steels with different retained austenite content and mechanical stability (TBF, TBF/Q&P, Q&P). The experiments were performed at the powder diffraction endstation of the Materials Science and Powder Diffraction beamline (BL04-MSPD) at ALBA Synchrotron (Figure 3.18a). Data were collected in transmission mode in the high-angular resolution setup using the Multicrystal Analyser Detector (MAD). This detector allows the exact determination of the diffraction peaks angles with no effects due to small displacements in sample positioning. More detailed information about the BL04-MPSD beamline and the MAD detector can be found in [147,148]. The measurements were performed directly by placing the interest zones in the X-ray beam of 3 x 1.5 mm2 (width x height, Figure 3.18b). The selected energy for the incident beam was 38keV (0.32622 Å wavelength, determined from NIST640D silicon standard data), high enough to 3.5 Microstructural characterization 54 minimize the absorption due to the sample thickness (1.4-1.5 mm). To evaluate the evolution of the retained austenite transformation during crack propagation, 4 different deformation stages were investigated (Figure 3.18c). For each stage, one sample was analyzed. Three consecutive zones (separated by 3.5 mm) were measured in each of the samples, from the crack tip (“zone 1”) to the centre of the ligament (“zone 3) (Figure 3.18b). In each zone, diffraction data were collected from -1º to 44º (2-theta) by scanning the detector at 1º/min, repeating the last 15º twice to increase the statistics, with a total measurement time of 1h. By using this scanning range referenced at the detector central channel and considering the distribution of the 13 channels of the detector with a 1.5º offset, it is ensured that all the channels collected the data in the 8 to 35º (2-theta) range. Rietveld refinements were performed with the FullProf program [149] using a pseudo-Voigt function and refining the scale factor, zero offset, isotropic thermal factor of Fe, cell parameters, the pseudo-Voigt coefficients considering an anisotropic strain broadening (modelled using the quartic form expression in reciprocal space implemented in FullProf) and one March coefficient for the preferred orientation considering an [hkl]* of [111] for the retained austenite and [110] for martensite. Figure 3.18. a) Experimental setup for Synchrotron X-Ray Powder Diffraction measurements in the MSPD line of Alba Synchrotron. b) Location of the measurements in the specimen surface. c) Different stages of deformation investigated. 55 Chapter 4 Results This chapter summarizes the main results obtained in Papers I-IV. Additional findings presented in appended papers A-D and other unpublished results are also included. In order to provide a global overview of the work, the presentation of the results is divided into three categories: fracture toughness, edge cracking resistance and crash failure behaviour. The results of microstructural characterizations are not included here. They will be used in the next chapter to discuss the influence of microstructure on fracture toughness. SEM micrographs and microstructural descriptions of the different AHSS grades are available in Papers I-IV. Paper IV, also includes results of HR-EBSD and nanoindentation measurements for the two DP steels investigated (DP1000A and DP1000B). The results of synchrotron X-ray diffraction measurements will be used in Chapter 5 to evaluate the evolution of the retained austenite-to- martensite transformation during crack propagation in TRIP-assisted steels. These results are not published yet. A publication on this topic is currently being prepared. 4.1 Fracture toughness results 4.1.1 Essential Work of Fracture 4.1.1.1. Specific essential work of fracture, we The results of EWF tests are shown in Figure 4.1 to Figure 4.3. The figures show the values of wf as a function of the initial ligament length (l0). In all cases, wf scales linearly with the ligament length and the specific essential work of fracture, we, can be obtained by extrapolation to ligament zero. Numerical values of we and non-essential plastic work βwp are given in Table 4.1. Overall, the results showed a good repeatability and a good linear data fitting was obtained for most of the steels (R2=0.87-0.99, see Table 4.1). we values range from 104 kJ/m2 to 405 kJ/m2. In general, CP-like microstructures (CP1000, TBF/Q&P, PHS1000, Q&P, CP1200) show higher we than DP-like (DP780, TRIP780, DP1000, TBF) or fully martensitic microstructures (PHS1500). CP1000 shows the highest we of the investigated steels, followed by TWIP, which has a fully austenitic microstructure, PHS1000, TBF/Q&P and DP1000B. Q&P steels and CP1200 lie in an intermediate range of we values (≈ 200 kJ/m2). On the other hand, TRIP780, 3rd Gen DP1180 and 3rd Gen TBF1180 exhibits the lowest crack propagation resistance. The rest of steels (DP780, DP1000, DP1000A, TBF) show similar toughness with we values around 150 kJ/m2. Concerning the non-essential plastic work, βwp, the trend is rather the contrary than the observed in we, i.e. DP-like and TRIP-assisted steels show higher plastic work than CP-like steels. TWIP steel shows by far the highest non-essential plastic work (βwp= 65 ± 2 MJ/m3). Figure 4.4 to Figure 4.6 shows the results of the DIC analysis in the ligament area for three different steel grades. The images show the Equivalent Mises Strain just before crack initiation for the smallest and the largest ligament. In all cases, the ligament area is completely yielded (Mises strain above 0.005-0.006) before crack initiation, which is one of the requirements for the application of the EWF methodology. This analysis was performed in all the investigated materials. All the steels satisfied this requirement. 4.1 Fracture toughness results 56 Figure 4.1. wf as a function of the initial ligament length (l0). a) 780 MPa steel grades. b) 1500 MPa steel. Figure 4.2. wf as a function of the initial ligament length (l0). Results for 1000 MPa steel grades. Figure 4.3. wf as a function of the initial ligament length (l0). Results for 1200 MPa steel grades. Chapter 4. Results 57 Figure 4.4. PHS1500. Equivalent Mises strain at crack initiation. Small (left) and large (right) ligament length. Figure 4.5. CP1200. Equivalent Mises strain at crack initiation. Small (left) and large (right) ligament length. Figure 4.6. TBF. Equivalent Mises strain at crack initiation. Small (left) and large (right) ligament length 4.1.1.2. Fracture toughness at crack initiation, wei The values of fracture toughness at crack initiation, wei, are summarized in Figure 4.7. wei values are also given in Table 4.1. Initiation toughness values can be roughly divided into 4 different groups or ranges:  very high wei (>250 kJ/m2): TWIP  high wei (180-200 kJ/m2): 3rd Gen Q&P1180, DP1000B  medium wei (≈150 kJ/m2): CP1000, Q&P, CP1200, TBF/Q&P, PHS1500, PHS1000  low wei (≈100 kJ/m2): DP1000, TBF, DP780, TRIP780, DP1000A, 3rd Gen DP1180, 3rd Gen TBF1180. 4.1 Fracture toughness results 58 Similarly to the observed in we, DP-like steel grades show the lowest toughness at crack initiation. It is worth noting that CP1000, which has the highest we, shows a crack initiation resistance comparable to PHS1500, Q&P or CP1200. In this case, TWIP steel shows the highest toughness at crack initiation. Figure 4.7. wei results for all the studied steel grades. 4.1.1.3. Critical crack tip opening displacement, δc In Figure 4.8 to Figure 4.10, the displacement at fracture (df) is plotted as a function of the ligament length. It can be observed that these graphs are very similar to the plots of wf vs l0 (Figure 4.1 to Figure 4.3) and the critical crack opening displacement (δc) can be obtained by extrapolation to zero ligament length. The values of δc are summarized in Table 4.1. Figure 4.8. df as a function of the initial ligament length (l0). a) 780 MPa steel grades. b) 1500 MPa steel Chapter 4. Results 59 Figure 4.9. df as a function of the initial ligament length (l0). Results for 1000 MPa steel grades. Figure 4.10. df as a function of the initial ligament length (l0). Results for 1200 MPa steel grades. 4.1.1.4. Thickness strain at crack initiation and propagation Figure 4.11 shows optical microscope images of the fracture surface of different steels, where thickness strain (TTSDENT) measurements were performed. The values of TTSDENT are plotted as a function of the distance from the crack tip in Figure 4.12 to Figure 4.14. TTSDENT i and TTSDENT p values are given in Table 4.1. In most cases, the thickness strain gradually increases with increasing the distance from the crack tip until a constant value is reached for a distance of about 0.4-0.6 mm. TTSDENT p is the average of thickness strains in this constant region. The increase in TTSDENT during crack growth illustrates the contribution of necking to the crack propagation resistance. The steels that develop a larger degree of necking during crack 4.2 Local ductility parameters from uniaxial tensile tests 66 Table 4.5. Summary of fracture resistance parameters obtained from KTTs. Steel UIE [kJ/m2] UPE [MJ/m3] Tear strength [MPa] TYR [-] εf KTT [-] CP1000 147 ± 15 639 ± 10 1720 1.89 0.15 DP1000 87 ± 9 479 ± 15 1538 2.08 0.06 TBF 104 ± 12 579 ± 30 1557 2.15 0.05 Q&P 144 ± 13 566 ± 26 1870 2.06 0.09 CP1200 122 ± 2 382 ± 8 1849 1.78 0.05 TBF/Q&P 148 ± 25 757 ± 38 1770 2.02 0.11 PHS1500 114 ± 34 530 ± 86 2108 1.96 0.05 PHS1000 120 ± 4 477 ± 40 1651 1.67 0.14 4.2 Local ductility parameters from uniaxial tensile tests Figure 4.21 shows the values of true fracture strain (TFS) for the AHSS grades investigated in Paper I and Paper II. The values of true thickness strain (TTS) evaluated in Paper II are also shown. CP1000 shows the highest TFS amongst the investigated steels (TFS= 1.21), followed by Q&P (TFS= 1.05). The rest of the steel grades show similar values of TFS ranging from 0.5 to 0.6. In general, for the six steels investigated in Paper II, the TTS is almost identical to the TFS, except in the case of TRIP 780, where the TTS is about half of the TFS (TFS=0.49 and TTS= 0.25). It indicates that in most of the steels the major contribution to the fracture strain comes from localized necking (thickness strain) and width strain is almost negligible. On the other hand, TRIP780 has a similar contribution from both local and diffuse (width strain) necking. Figure 4.21. True fracture strain (TFS) and true thickness strain (TTS) measured from the fracture surface of uniaxial tensile specimens. Chapter 4. Results 67 4.3 Edge cracking resistance 4.3.1 Hole Expansion Tests 4.3.1.1 Hole Expansion Ratio HER values are shown in Figure 4.22 and Table 4.6. As a rule, and in line with the observations from fracture toughness tests, more homogeneous CP-like microstructures show higher HER than DP-like, TRIP-assisted and fully martensitic microstructures. CP1000 shows the highest HER, followed by TBF/Q&P, Q&P and PHS1000. The lowest HER value is shown by TRIP780. PHS1500, DP1000, TBF, DP780, 3rd Gen DP1180 and 3rd Gen TBF1180 show similar HER (≈30 %). DP1000A and 3rd Gen Q&P1180 present slightly higher HER (≈40 %). The poorer stretch flangeability of DP-like steels is associated to the hardness differences between soft (ferrite) and hard (bainite/martensite) microstructural constituents, which contribute to the rapid generation of microvoids or decohesion of the soft/hard phase interfaces [19,50]. 4.3.1.2 Thickness strain The values of true thickness strain measured in hole expansion test specimens (TTSHET) for the six steel grades investigated in Paper II are plotted in Figure 4.22. Numerical values are summarized in Table 4.6. TTSHET values follow a similar trend to TTS from uniaxial tensile tests shown in Figure 4.21. Most of the steels show similar TTSHET (0.10-0.12), except TRIP780, which presents the lowest thinning (TTSHET =0.08) Figure 4.22. HER for the studied AHSS grades. True thickness strain measured from hole expansion test specimens (TTSHET) is also indicated for the steels investigated in Paper II. 4.3.2 Hole Tension tests The values of major strain at fracture at the hole edge obtained from Hole Tension Tests (HTT) are represented in Figure 4.23. The figure shows the results for the four AHSS grades and the different punching clearances (c) investigated in Paper B. The results are also given in Table 4.6. 4.3 Edge cracking resistance 68 Due to the higher thickness of TBF (t=1.5 mm) compared to the other steels (t=1.4 mm), the resulting clearances were slightly different. The investigated clearances for DP1000, CP1000 and TBF/Q&P were 11%, 14%,18% and 21%. In the case of TBF, the clearances were 10%,13%,17% and 20%. As expected, the results are in good agreement with hole expansion tests. DP1000 and TBF have significantly lower edge formability (lower εf HTT) than CP1000 and TBF/Q&P. These results also show the great influence of cutting clearance on edge formability, especially in DP1000 and TBF. In all cases, the highest εf HTT is attained for a punching clearance of 13-14%. For DP1000 and TBF, the fracture strain decreases by up to 38% and 28% respectively for clearances above 14%. On the other hand, CP1000 and TBF/Q&P are less sensitive to punch-to-die clearance. This suggests that materials with higher damage tolerance (fracture toughness) are less sensitive to edge damage, while in the case of low toughness materials (DP-like), edge fracture is strongly affected by the edge quality. Figure 4.23. Major strain at fracture at the hole edge (εf HTT) for the four investigated steel grades. Results for different punch-to-die clearances. Table 4.6. Summary of results from edge cracking resistance tests Hole Expansion tests Hole Tension tests, εf HTT Steel HER [%] TTSHET [-] c=10-11% c=13-14% c=17-18% c=20-21% CP1000 85 ± 4 - 0.61 ± 0.02 0.70 ± 0.04 0.64 ± 0.04 0.68 ± 0.01 DP1000 35 ± 4 - 0.34 ± 0.01 0.37 ± 0.04 0.26 ± 0.00 0.23 ± 0.02 TBF 30 ± 1 - 0.36 ± 0.02 0.36 ± 0.02 0.26 ± 0.02 0.29 ± 0.01 Q&P 55 ± 8 - - - - - TBF/Q&P 66 ± 11 - 0.53 ± 0.03 0.64 ± 0.01 0.51 ± 0.01 0.54 ± 0.01 PHS1500 28 ± 2 - - - - - PHS1000 57 ± 1 - - - - - DP780 34 ± 3 0.11 ± 0.03 - - - - TRIP780 23 ± 3 0.08 ± 0.00 - - - - DP1000A 38 ± 1 0.11 ± 0.02 - - - - 3rd Gen DP1180 32 ± 1 0.10 ± 0.02 - - - - 3rd Gen TBF1180 28 ± 2 0.11 ± 0.02 - - - - 3rd Gen Q&P1180 41 ± 4 0.12 ± 0.01 - - - - Chapter 4. Results 69 4.4 Crash fracture behaviour 4.4.1 Axial impact resistance Figure 4.24 shows some examples of tested crash boxes and the location of the crack observed during axial impact resistance tests. The results of axial impact resistance tests are described in detail in Paper III. The evolution of the Crash Index (CI) as a function of the intrusion level is illustrated in Figure 4.25. The slope defined by the decrease of CI as a function of the intrusion level (dashed lines in Figure 4.25) is defined as the Crash Index Decreasing Rate (CIDR). The CIDR quantifies the growth rate of cracks through the sample under axial impact loading, independently of the crack origin. Thus, the CIDR is related to the evolution of damage and can be used to quantify the crash resistance of the material; the lower the CIDR, the better the crash resistance. The values of CIDR obtained for the nine investigated steel grades are shown in Table 4.7 together with the critical intrusion and the energy absorbed at the maximum intrusion. The critical intrusion refers to the transition from uncracked to cracked samples, i.e. the first CI<100. The best crash resistance (lower CIDR) is shown by CP1000, PHS1000, TWIP and TBF/Q&P steels. On the other hand, DP1000, TBF and PHS1500 exhibit the worst crash resistance, as indicated by their high CIDR (2.2-2.5) and their low CI for high intrusion levels. CP1200 and Q&P show slightly better crash performance (CIDR ≈1.3-1.8). The energy absorbed at maximum intrusion represents the crash energy absorption capacity of the material. This energy directly depends on the maximum intrusion allowable by the material and, therefore, it is closely related to the CIDR. The maximum intrusion was defined for a CI of 20% (specimen severely damaged). The trend observed for energy at maximum intrusion is similar to the one observed for CIDR values. The definition of a single critical intrusion level was difficult in most cases since fracture initiation in axial crash tests can be highly influenced by damage induced at the welding spots. The range of critical intrusions for the different steels is indicated in Table 4.7. Most of the steels show similar ranges of critical intrusion ranging from 20 to 40 mm. The lowest ranges of critical intrusion are shown by Q&P, CP1200 and PHS1500. Table 4.7. Results from axial impact resistance tests Steel Critical intrusion [mm] CIDR [-%/mm] Energy at maximum intrusion [kJ] CP1000 27-32 0.29 13.7 DP1000 18-27 2.18 3.6 TBF 25-33 2.54 4.9 Q&P 6-15 1.81 5.3 CP1200 9-18 1.28 6.1 TBF/Q&P 30-45 0.71 9.4 PHS1500 12-13 2.45 3.0 PHS1000 16-20 0.36 9.5 TWIP 99-100 0.11 10.7 4.4 Crash fracture behaviour 70 Figure 4.24. a) Crash boxes of CP1000 (top) and DP1000 (bottom) after axial impact resistance tests at different speeds. b) Examples and location of cracks observed in axial crash tests. Figure 4.25. Variation of Crash Index (CI) as a function of the intrusion in axial impact tests for the studied steels. Energy absorbed is also indicated in the upper x-axis. Figure from Paper III. Chapter 4. Results 71 4.4.2 Bending impact resistance The evolution of the CI as a function of the intrusion level in bending impact tests is investigated in Figure 4.26. The values of CIDR and critical intrusion are shown in Table 4.8. The energy absorbed at maximum intrusion was not considered in bending impact tests. Maximum intrusion level was limited by the test equipment configuration (211 mm) and not by the material behaviour. Therefore, the energy at maximum intrusion was the same for all the steel grades. Due to the limited maximum intrusion, it was not possible to reach the failure of the sample in all investigated steel grades. In the event that no damage was observed, which was the case for CP1000, TBF/Q&P, PHS1000 and TWIP, at least 3 samples were crashed to the maximum possible intrusion. For these steels, it was not possible to determine a CIDR, because of the absence of cracking. For the rest of steels, the point of the first decrease in CI was used to determine the CIDR. PHS1500 and TBF present the highest CIDR, which indicates a rapid increase in the degree of cracking with the intrusion level. DP1000 and Q&P show slightly better cracking performance (CIDR= 0.26-0.28). PHS1500 shows the lowest intrusion level for crack initiation. On the other hand, DP1000 and TBF exhibit the highest critical intrusion, followed by Q&P with slightly lower values. Figure 4.26. Variation of Crash Index (CI) as a function of the intrusion in bending impact tests. Table 4.8. Results from bending impact resistance tests Steel Critical intrusion [mm] CIDR [-%/mm] CP1000 - - DP1000 150-160 0.26 TBF 150-160 0.40 Q&P 125-130 0.28 TBF/Q&P - - PHS1500 45-50 0.49 PHS1000 - - TWIP - - 4.4 Crash fracture behaviour 72 73 Chapter 5 Discussion 5.1 Fracture toughness characterization of AHSS sheets 5.1.1 Essential Work of Fracture The Essential Work of Fracture (EWF) methodology has been established as one of the most interesting methods to characterize the fracture resistance of thin ductile sheets. The main advantage of this technique is its relative experimental simplicity compared to other conventional fracture mechanics procedures, since it does not require crack growth monitoring and data postprocessing is rather simple. The EWF method was originally developed by Cotterell and Redell for ductile metals [133] and was rapidly extended for the characterization of ductile polymers [135-137]. Over the years, the methodology has been widely used for evaluating the fracture resistance of polymer films [150-153] and metallic materials: low carbon steels [133, 154,155], aluminium alloys [119, 139, 140, 156], zinc [140, 143], copper [157] and brass [158]. More recently, the method has gained increasing interest to characterize the fracture resistance of high strength steel sheets. Lacroix et al. [84] used the EWF to evaluate the fracture toughness of different TRIP-assisted steels and to investigate the influence of the TRIP effect in their crack propagation resistance. Later, Muñoz et al. [159] and Gutiérrez et al. [160] studied the applicability of the method in various AHSS steel sheets. Since these works, a number of researchers have used the EWF metholodogy to characterize the fracture properties of several AHSS (DP [20, 86, 91, 161], TWIP [85, 87], Q&P [20, 85]) and PHS [20, 26] sheets. In spite of the great potential of the EWF method to readily measure the fracture resistance of thin ductile sheets under plane stress conditions, there is not a standard procedure developed yet. One of the main challenges in EWF standardisation is the sensitivity of the method to different testing variables: notch quality (sharpness, alignment), number of specimens, ligament length range, etc. [142]. Different attempts have been made to standardise the EWF methodology [138, 162-164]. The last version of the ESIS protocol for EWF testing was revised in 2001 [138]. This protocol, developed by the ESIS TC4 committee (TC04- Polymers, Polymer composites and adhesives), is based on a series of round robin tests during a seven-year period, with the participation of 23 laboratories, and it is currently the most extended for the evaluation of the EWF. The protocol provides the guidelines for the evaluation of the EWF by using DENT specimens and discuss some of the most critical points related to specimen preparation, testing and data analysis. In this thesis work, the EWF method was applied to a wide range of advanced high strength sheet steels (t =1.2-1.6 mm), following the recommendations of the ESIS protocol [138]. However, this protocol is focused on the fracture testing of polymers and composites. Therefore, no recommendations are given about the notch preparation in metallic materials. In order to obtain the sharpest possible notch and avoid the influence of notch root radius on fracture toughness measurements, fracture mechanics standard procedures for metals recommend the nucleation of fatigue pre-cracks at the notch root [104, 111]. Accordingly, in the present work, all the DENT specimens used for EWF tests were fatigue pre-cracked according to the recommendations of ASTM E1820 [111], as explained in Section 3.2. All the investigated steels showed to satisfy the requirements for the validity of the EWF methodology. One of the basic requisites for the applicability of the energy partitioning concept is that the ligament is fully yielded before fracture initiation. This was validated in all the steel 5.1 Fracture toughness characterization of AHSS sheets 74 grades by means of DIC analysis (Figure 4.4 to Figure 4.6). Such analysis also showed that the shape of the plastic zone is almost circular with a diameter of approximately the ligament length, as previously shown by other authors [133, 161]. Some of the main validation criteria described in the ESIS protocol are briefly discussed below. For the sake of brevity, only one steel per strength class (800 MPa, 1000MPa, 1200 MPa and 1500 MPa) is shown. Similarity of load-displacement curves and linearity of wf vs l0 data The self-similarity between the load-displacement curves for a set of specimens with different ligament lengths is other of the criteria used for data validation [138, 142, 153]. The AHSS grades investigated in this work accomplished this requirement, as illustrated in Figure 5.1. It can be seen that the shape of the load-displacement curves is similar for the different ligament lengths and maximum load (Pmax) and displacement at fracture (df) scale with the ligament length (Figure 4.8 to Figure 4.10). As explained before, the specific essential work of fracture (we) is obtained by linear regression of wf vs l0 data. Therefore, wf should linearly increase with increasing the ligament length, as evidenced in Figure 4.1 to Figure 4.3. That behaviour may also be an indicator of the EWF applicability. It was found that in materials not accomplishing the ligament yielding criterion (materials not investigated in this work), the wf remained almost constant independently of the ligament length and, thus, the determination of we from extrapolation to zero ligament length was not possible. The linearity of the wf vs l0 data can be used as a data qualification criterion. Marchal et al. [140] proposed a statistical procedure to detect the loss of linearity and improve the accuracy of we. The loss of linearity may indicate a change in the fracture mechanisms or the stress state. It is especially critical in the small ligament region, where the transition from plane stress to a mixed plane strain/plane stress mode is more likely to occur [139] (Figure 1.42). For the steels investigated in this work, in general, a good linearity was observed, as indicate the high regression coefficients (R2>0.90). Only some exceptions with rather low linearity were observed, such as the case of PHS1000 (R2=0.42) or CP1200 (R2=0.53). Ligament length range The definition of a valid ligament range is a key point to obtain an accurate plane stress fracture toughness value. As mentioned before, the ligament should be small enough to ensure that the ligament is fully yielded before fracture but sufficiently large to ensure a global plane stress state. If the ligament is too small, the fracture may take place under a mixed plane strain/plane stress mode and wf does not linearly scale with l0 anymore (Figure 1.42). This transition region is often given by 3t [141], where t is the sheet thickness. However, this lower boundary is material dependent. For instance, Wu and Mai [135] found that this transition occurred for l0=14t in a linear low-density polyethylene (LLDPE) film. On the other hand, Cotterell and Reddel [133], suggested that the lower ligament length should be of the order of 5t. They also show that the transition zone from plane strain to plane stress, and thus the minimum ligament length, could be determined by observation of the fracture surface after the test. The ESIS protocol also includes an arbitrary value of 5 mm, being the criterion for lower boundary the maximum of 3t or 5 mm. This criterion has been the most extensively used and it has shown to be adequate for most of the materials investigated in the literature. Regarding the upper boundary, the conditions of W/3 and 2rp are given, respectively, to avoid the spreading of the plastic zone to the edges of the specimen and to ensure that the ligament is fully Chapter 5. Discussion 75 yielded. However, the ESIS TC4 protocol points out that some research works have shown that the data is often linear beyond this limit. An arbitrary upper limit of l0=15 mm is proposed for practical purposes. Table 5.1 shows the ligament limits for the steels investigated in this work, according to the recommendations of the EWF testing protocol. rp was evaluated according to Equation (32), a Young’s modulus of 210 GPa was used for calculation. Overall, the ligament length ranges used are within the limits established in the protocol, except for the 1200 MPa (3rd Gen DP1180, 3rd Gen TBF1180, 3rd Gen Q&P 1180, CP1200) and 1500 MPa steels. For these materials, the criterion of 2rp for the upper limit seems to be too restrictive, which is in line with the mentioned in the protocol, and an upper limit of l0=15 mm is more appropriate. Figure 5.1. Typical load-displacement curves obtained from EWF tests. a)TRIP780, b) DP1000, c) 3rd Gen 1180Q&P, d) PHS1500. 5.1 Fracture toughness characterization of AHSS sheets 82 Paper D for more details about the tool and the notching procedure). This method was validated for four different AHSS grades: CP1000, DP1000A, DP1000B and 3rd Gen TBF1180. The EWF results for specimens with sheared notches are shown, and compared to the obtained with fatigue pre-cracked specimens, in Figure 5.8a (CP1000) and Figure 5.9 (DP1000A, DP1000B and 3rd Gen TBF1180). For similar ligament length, the four investigated steel grades show practically identical wf for both fatigue pre-cracked and sheared specimens, which lead to very similar we and βwp. Figure 5.8. EWF results for specimens with EDM notches (open symbols) and specimens with fatigue precracks (black squares). a) CP1000 (results of specimens with sheared notches are also given, grey triangles). b) DP1000. c) TBF and d) Q&P. One of the major concerns that can arise when preparing notches with mechanical methods, is the accumulation of plastic deformation around the crack tip, which has shown to be critical, for example, in polymer films [152, 169]. Martínez et al. [169] observed a significant variation on the we of an ethylene-propylene block copolymer (EPBC) film with DENT specimens prepared using different notch sharpening methods. The we increased from 60±7 kJ/m2 for specimens prepared using femtosecond pulsed laser ablation (Femtolaser) to 134±7 kJ/m2 for specimens with notches sharpened by a diamond edged razor blade. This large variation of we was attributed to the plastic deformation ahead of the crack tip produced during notch sharpening. Similar observations were made by León et al. [155]. However, for the AHSS grades investigated in this project, it seems that the possible plastic deformation introduced during the shearing process has no significant influence on we. In fact, the load-displacement curves obtained for specimens prepared with this method are practically identical to the ones for specimens with fatigue precracks (see Paper D, figure 6), which suggests that this effect is almost negligible. It must be noted that this notching preparation procedure is currently under study and further investigations are needed to quantify the influence of different variables affecting the process Chapter 5. Discussion 83 (cutting clearance, punch wear, etc.) and to define its limitations (material strength, thickness, etc.). However, the method has shown to be reliable enough to accurately evaluate the fracture toughness of the four AHSS investigated in this work. These results show the great potential of the method, which can be established as an attractive alternative to fatigue pre-cracking operations in high strength metal sheets. Figure 5.9. EWF results for specimens with sheared notches (grey triangles) and specimens with fatigue pre-cracks (black squares). a) DP1000A, b) DP1000B and c) 3rd GenTBF1180. 5.1.2 J-integral The J-integral has been the most extended parameter to characterize the fracture toughness of elastic-plastic materials. The J-integral procedure is used to evaluate the critical J-integral (Jc) for stationary cracks and the J-Resistance (J-R) curves for growing cracks. Note that the notation Jc has ben used in this work instead of JIc, which is used to define the thickness independent planestrain fracture toughness. As defined in ASTM E1820 [111], Jc is a measure of fracture toughness independent of the in-plane dimensions. However, it may depend on specimen thickness. To qualify Jc as a thickness independent fracture toughness parameter, JIc, the following conditions must be met: 𝐵,𝑏0>10𝐽𝑐 𝜎𝑌 (42) where B is the specimen thickness, b0 is the uncracked ligament length and σY is the effective yield strength. Jc represents the fracture toughness near the onset of stable crack propagation. However, many tough materials do not fail catastrophically at Jc but they show a rising R curve, where J 5.1 Fracture toughness characterization of AHSS sheets 84 increases with crack growth. In these materials, the J-R curve provides a more complete description of the stable crack propagation resistance. Unfortunately, no standard methods are available to evaluate the J-integral in thin gauge materials. The ASTME1820 was developed to characterize the fracture toughness of metallic materials under plane strain conditions and, therefore, specimen size requirements cannot be satisfied by relatively thin sheets. For this reason, some researchers have proposed alternative non-standard methods for J-integral measurements in thin ductile sheets. Pardoen et al. [119] used the J-integral expression proposed by Rice et al. [170] to evaluate Jc in DENT specimens: 𝐽=𝐾2 𝐸+1 𝑡0𝑙0(2∫𝑃𝑑𝑢𝑝−𝑃𝑢𝑝) (43) Where K is the stress intensity factor, which for a DENT specimen is given by Equation (44) [171]: 𝐾= 𝑃√𝜋𝑎 𝑡0·2𝑊√1−𝑎 𝑊[1.122−0.561(𝑎 𝑊)−0.205(𝑎 𝑊)2+0.471(𝑎 𝑊)3−0.190(𝑎 𝑊)4] (44) E is the elastic modulus, t0 is the specimen thickness, l0 is the uncracked ligament length, P is the actual load and up is the actual plastic displacement. In Equation (44), W is the half of the specimen width. This approach has been also used in different research works to evaluate the initiation toughness (Jc) of high strength steels [84, 87, 88, 90]. Other of the methods commonly used for J-integral evaluation in thin polymer films [152] is the Begley and Landes method [108]. In the appended Paper I, a different approach is investigated. The J-integral of the investigated AHSS grades was evaluated following the experimental procedure described in ASTM E1820 for CT specimens introducing some variations. First, the specimen geometry was machined according to the proportions described in ASTM E561 [145], which does not present any restriction regarding specimen thickness. On the other hand, instead of using the compliance technique for actual crack length estimation, crack extension was directly measured on the specimen by means of a high resolution video camera located in one of the specimen sides. For each material, a J-R curve was constructed according to ASTM E1820. The results are shown in Section 4.1.2. As observed by Zhu and Leis [117], the results show that if the thickness requirements described in the ASTM E1820 are disregarded the methodology can be applied to evaluate a J-R curve in a thin ductile material. Furthermore, the use of a digital video equipment in combination with a digital image analysis software has shown to be accurate enough to follow the evolution of crack extension during the test. In combination with a DIC software, this can be an interesting approach to investigate the crack propagation behavior of thin metal sheets. In this work, two fracture resistance parameters were defined from the J-R curves: the J value at at crack initiation, Ji , which is the value for crack growth initiation detected in the video equipment, and Jc, which is defined as the fracture toughness near the onset of stable crack propagation obtained from the intersection of the 0.2 mm offset line with the R curve as indicated in ASTM E 1820. In steels with higher toughness, such as CP1000 and Q&P, the initial part of the resistance curve shows a more pronounced slope and the difference between Ji and Jc is greater. Chapter 5. Discussion 85 This difference is lower for steels with lower toughness (DP1000 and TBF). These steels also show flatter R curves indicating lower contribution to crack propagation resistance. 5.1.2.1 Influence of specimen geometry To investigate the influence of specimen geometry on J-R curve, J-integral measurements were performed using CT and DENT specimens. For DENT specimens, the J-integral at crack initiation (Ji ) was evaluated from the load-line displacement curve according to Equation (43). To account for the crack growth the J values for the different crack extensions were evaluated using an incremental equation as detailed in Paper I. As illustrated in Figure 4.18, specimen geometry has a significant influence on the resistance curve. However, the R curves converge in the initial part of the graph, providing similar values of Ji and Jc. This is in good agreement with the findings of Xia et al. [172] and Zhu et al. [173]. They showed that constraint level, while has little influence on Jc, has a significant effect on the slope of the R curve. Since constraint level is higher in CT than in DENT specimens, they show lower tearing modulus. 5.1.2.2 Relationship between we and Jc The relationship between we and Jc was discussed in Paper I. As mentioned in the introduction (Section 1.4.8.4), this topic has been recurrently investigated in literature [119, 135-137, 166, 174]. Different authors showed the equivalence between we and Jc in ductile polymers [135,136]. Furthermore, Mai and Cotterell [174] differentiated between Ji and Jc (they called it Jp) and found that Ji and Jc were equivalent to wei and we respectively. The results of EWF and J-integral measurements performed in Paper I are summarized in Figure 5.10. Although the values of Ji are slightly lower than wei, a quite good agreement can be observed between both parameters in the four investigated steels. It must be noted that the results of J-integral measurements are obtained from one single specimen. Therefore, no standard deviation is given. In order to obtain more accurate toughness values and better statistical confidence, it is recommended to test additional specimens. Regarding we and Jc, some differences can be discerned between we and Jc. On the one hand, some steel grades like DP1000 and TBF show a very good equivalence but, on the contrary, significant deviations are observed between these two parameters in CP1000 and Q&P. The differences between we and Jc are discussed on the basis of their conceptual differences. By definition, the specific essential work of fracture (we) contains energetic contributions from crack initiation and propagation resistance since it is derived from a linear regression of wf for the complete fracture [166]. On the other hand, Jc, as defined by the ASTM E1820, is a toughness value for a small crack advance, determined by the intersection of the J-R curve and the 0.2 mm offset line parallel to the construction line. Therefore, as suggested by Pardoen et al. [166], when there is a small contribution from crack propagation after initiation (DP1000, TBF, Q&P), JC is similar to we. On the contrary, for steels having a large contribution from crack propagation resistance, such as the case of CP1000, we and Jc differ. In the case of Q&P, as discussed in Paper I , the fact that it shows greater Jc than we is probably attributed to the additional energy associated with the 0.2 mm of crack advance. The similarities in J-R curves of CP1000 and Q&P are explained by the difference in energetic contributions from necking and plastic work to the ductile fracture process. In CP1000, a large part of the crack propagation resistance energy comes from the necking developed at the crack tip, and it is well captured by we. On the other hand, in Q&P this contribution is minimum and the major part of the energetic increase observed in J-R curve 5.1 Fracture toughness characterization of AHSS sheets 86 comes from the non-essential plastic work developed in a region outside the fracture process zone, which as derived from EWF tests (Figure 4.2 and Figure 4.3), is about the double than for CP1000 (βwp= 20± 1 MJ/m3 and 12±1 MJ/m3 for Q&P and CP1000, respectively). The findings obtained in this work, suggests that when the fracture process has an important contribution from the crack propagation, we better represents the tearing resistance of thin ductile sheets than Jc, which is line with previous observations [119]. Figure 5.10. Comparison of EWF and J-integral results. J values obtained from DENT specimens. 5.1.2.3 Thickness independent fracture toughness validation criterion As mentioned above, according to the ASTM E1820 the measured Jc value can be considered as a plane-strain thickness independent fracture toughness value (JIc), when the conditions given in Equation (42) are met. The same criterion can be extrapolated to we. In this case, the plane-strain specific essential work of fracture is denoted wIe. The fulfilment of these conditions for all the AHSS studied in this work is investigated in Table 5.2. As expected, most of the steel grades do not satisfy the minimum thickness requirements suggested by ASTM E1820 to be considered as thickness independent fracture toughness values. However, for some of them (TRIP780, 3rd Gen DP1180, 3rd Gen TBF1180 and PHS1500) the minimum thickness required is below the actual sheet thickness, which suggests that the obtained specific essential work of fracture could be considered as the plane-strain essential work of fracture (we=wIe). Further investigations on the effect of specimen thickness on we must be performed to validate these assumptions. Chapter 5. Discussion 87 Table 5.2. Validation criteria for considering Jc and we as thickness independent plane-strain fracture toughness values, JIC and wIe, respectively. Values in parenthesis are the minimum thickness values corresponding to Jc values. Steel t [mm] σY [MPa] Jc [kJ/m2] we [kJ/m2] Minimum thickness [mm] Qualification of Jc and we as Jc= JIc, we=wIe CP1000 1.4 962 286 405 ± 11 (3.0) 4.2 NO DP1000 1.4 895 158 138 ± 20 (1.8) 1.5 NO TBF 1.5 884 157 149 ± 13 (1.8) 1.7 NO Q&P 1.4 1061 280 194 ± 12 (2.6) 1.8 NO DP780 1.5 668 - 151 ± 31 2.3 NO TRIP780 1.6 697 - 106 ± 24 1.5 YES DP1000A 1.35 936 - 149 ± 21 1.6 NO 3rd Gen DP1180 1.2 1054 - 115 ± 20 1.1 YES 3rd Gen TBF1180 1.4 1102 - 104 ± 30 0.9 YES 3rd Gen Q&P1180 1.5 1113 - 196 ± 31 1.8 NO CP1200 1.6 1130 - 201 ± 24 1.8 NO TBF/Q&P 1.4 951 - 302 ± 32 3.2 NO PHS1500 1.5 1314 - 159 ± 18 1.2 YES PHS1000 1.5 998 - 330 ± 21 3.3 NO TWIP 1.4 750 - 366 ± 24 4.9 NO DP1000 B 1.4 907 - 286 ± 17 3.2 NO 5.1.3 Kahn-type tear tests The application of the Kahn-type tear tests (KTT) to characterize the tearing resistance of sheet materials has been mainly focused on thin aluminium alloy sheets [12,127-130]. In fact, the standard method (ASTM B871) that describes the experimental procedure and the specimen proportions for tear testing using the KTT is specific for aluminium alloy products [126]. However, some researchers have adapted the method to readily evaluate the fracture resistance of thin PHS [131] and TWIP [132] sheets. Ying et al. [131] used the KTT to investigate the influence of austenitization temperature, soaking time and start deformation temperature on strength and toughness of 22MnB5 press hardened sheets (t=1.6 mm). Lorthios et al. [132] also applied this method to study the tearing behaviour of a high Mn TWIP steel. In this work (Paper I and Paper C), the applicability of the KTT to characterize the fracture resistance of thin AHSS sheet was further investigated. The results showed that the specifications described in the ASTM B871 [126] regarding specimen geometry and experimental testing can be directly applied to high strength metal sheets in the range of 1.4-1.6 mm thickness and 1000- 1500 MPa. This simplistic energy-based method can be an interesting approach for indexing the tearing resistance of ductile metal sheets and evaluating the effect of processing parameters and microstructural constituents on overall toughness. However, as discussed in Paper I and Paper C, the results must be taken with caution. As shown in Figure 5.11, UIE values are in good agreement with initiation toughness values from EWF tests (wei). It is worth noting that despite the larger notch radius of KTT specimens (ρ= 150 µm) compared to fatigue pre-cracked DENT specimens used for EWF tests (ρ= 0.1 µm), the results for crack initiation resistance are very similar for both test configurations. This behaviour suggests that stress concentration ahead of the notch tip in KTT specimen closely resembles that of a crack and, thus, a machined sharp notch is suitable enough to obtain representative initiation toughness values. On the other hand, large differences are observed between we and UPE. In 5.1 Fracture toughness characterization of AHSS sheets 88 general, UPE values are much higher than we, which results in an overestimation of the crack propagation resistance. As explained in Paper I, UPE, similarly to J-resistance curves [175], contains not only the energy for creating new surfaces in the front of the crack tip but also includes an additional contribution from the non-essential plastic work dissipated in the outer region. On the contrary, we only quantifies the energy dissipated in the fracture process zone since the contribution from the plastic work is removed when extrapolating wf vs l0 data to ligament zero. This explanation can be used to understand the differences between we and UPE and demonstrates that UPE is not a material property that can be only used as a relative index of toughness. Nevertheless, as the poor correlation between we and UPE indicates (Figure 5.12a), UPE values from KTT may lead to wrong toughness estimations and misleading material ranking. As described in Section 3.2.3, an alternative approach, based on the measurement of the major strain at fracture in the notch tip (εf KTT) by means of DIC, is investigated in this work. This parameter shows a much better correlation with we (Figure 5.12b) and provides a quite good estimation of fracture toughness. Figure 5.11. Comparison between KTT and EWF results. Figure 5.12. Correlation between we and: a) UPE, b) εf KTT. Chapter 5. Discussion 89 5.1.4 Relationship between fracture toughness and tensile properties The relationship between the specific essential work of fracture and uniaxial tensile properties for all the AHSS grades investigated in this work is shown in Figure 5.13. Values of true fracture strain (TFS) and true thickness strain (TTS) are only available for the steels studied in Paper I and Paper II. As discussed in these two papers, no direct relationship can be discerned between fracture toughness and any of the conventional uniaxial tensile properties. The results evidence that greater elongation, both uniform (UE) or total (TE), or greater strain hardening exponent (n), which are used to define ductility and formability, do not indicate greater fracture toughness. Rather, on the contrary, steels showing the largest strain hardening and elongation values show relatively low fracture toughness. In the same line, no link is observed between the ultimate tensile strength by total elongation (UTSxTE) product and we. The UTSxTE product is often used in literature as a toughness indicator. However, this is in contradiction to the observed in Figure 5.13, which shows that this parameter is not suitable to estimate the cracking resistance of AHSS. On the other hand, local strain measurements from uniaxial tensile tests (TFS, TTS) give a better estimation of fracture toughness. Nevertheless, as shown by Xiong et al. [90], these fracturerelated parameters often are not accurate enough to describe the fracture behaviour of high strength sheet materials in the presence of existing cracks or defects. Therefore, to better understand the fracture performance of AHSS sheets, including crack initiation and propagation resistance, fracture toughness should be properly measured in the frame of fracture mechanics. Figure 5.13. Relationship between we and different tensile parameters. 5.2 Fracture toughness to understand edge cracking resistance 90 5.2 Fracture toughness to understand edge cracking resistance 5.2.1 Correlation between fracture toughness and edge formability As mentioned in the introduction, the relationship between edge fracture resistance and fracture toughness in high strength metal sheets has been addressed by different authors [20, 23, 26, 76, 80]. These works evidenced that sheared edge formability is governed by the material’s resistance against the propagation of pre-existing crack or defects introduced in the previous forming step (punching, shearing, cutting). Accordingly, different fracture mechanics approaches, such as J- integral [76, 80] or essential work of fracture measurements [20] have been proposed to understand edge fracture resistance. Takahashi et al. [76] used the critical J-integral value (Jc) to rationalize the differences in the edge fracture behaviour of different hot rolled dual phase steels (ferrite + martensite and ferrite+bainite). They observed that the fracture mechanisms present in fracture mechanics and hole expansion tests were very similar, which explained the good correlation between Jc and the limiting hole expansion ratio (HER). Similar conclusions were reached by Yoon et al. [80]. In that case, J-integral measurements were performed in various AHSS grades by using Single Edge Notched Tension (SENT) specimens and the parameter KJc was used as a measure of fracture toughness. KJc is derived from Jc as indicated in ASTM E18120 [111]: 𝐾𝐽𝑐 =√𝐽𝑐𝐸 (1−𝜈2) (45) where E is the Young'smodulus and ν is the Poisson’s ratio. Alternatively, Casellas et al. [20] proposed the use of the specific essential work of fracture (we) to understand the stretch flangeability of several AHSS sheets. They found that steels with higher we showed higher HER and established a linear relationship between these two parameters. This approach was further investigated in Paper II and Paper B. The results from these investigations are summarized in Figure 5.14. Figure 5.14a shows the correlation between we and HER for the AHSS grades investigated in this work. Results for CP1000 and DP1000 steels (sheet thickness, t=1.2 mm) studied in Paper A are also included. The very good linear fitting for different AHSS families (R2=0.90) confirms the close relationship between fracture toughness and stretch flangeability of AHSS and demonstrate that we is a suitable property to understand edge fracture resistance. As discussed in Paper I, the values of HER are well correlated with we but not with crack initiation resistance (wei). This is because, phenomenologically, the limiting hole expansion ratio is related to the overall crack propagation resistance rather than to initiation. It must be noted that the HER is determined when a crack has propagated through the full sheet thickness, i.e. a crack extension of about 1-1.6 mm. This argumentation also explains the better correlation between HER and we instead of Jc, which, as defined above, is a toughness value for a very small crack extension (0.2- 0.3 mm) and, thus, is not suitable enough to describe the full fracture resistance when there is significant crack propagation prior to final failure. Chapter 5. Discussion 91 Figure 5.14. a) Correlation between we and HER. Figure adapted from Paper II. b) Fracture strains measured in HTT specimens (εf HTT) for the 4 steel grades investigated in Paper B with different punch to die clearance. we are also plotted. Figure from Paper B. It is important to remark again that edge fracture does not only depend on material properties but also on the hole preparation method, edge quality, etc. The influence of punching clearance on the edge formability of different DP-like (DP1000, TBF) and CP-like (CP1000, TBF/Q&P) was studied in Paper B by means of hole tension tests (HTT) and DIC analysis, as described in previous sections. Figure 5.14b shows the fracture strain values measured on the surface of HTT specimens (εf HTT) for different punch-to-die clearances. The trend observed in Figure 5.14a can be also seen here, i.e steels with higher we present better edge formability, independently of the punching clearance. The results also suggest that tougher steels (CP1000 and TBF/Q&P) are much less sensitive to the edge quality than steels with low fracture toughness (DP1000 and TBF). This phenomenon may be related to the fact that, independently on the damage induced during punching or shearing, these steel grades show greater damage tolerance. As a conclusion, and based on the good correlation observed between we and the results from edge cracking resistance tests, it can be stated that we is a key property to understand the edge fracture resistance of AHSS sheets. 5.4 Influence of microstructure on fracture toughness 98 and stability, as well as the matrix characteristics, should be carefully controlled. In this sense, the EWF is proposed as a suitable methodology to investigate the influence of these microstructural features on crack initiation and propagation resistance. 5.4.2 Microstructural investigations on DP steels In Paper IV, the influence of microstructure on the crack propagation resistance of two industrially processed high strength dual phase sheet steels (DP1000A and DP1000B) is investigated. Microstructural investigations were performed by SEM, HR-EBSD and nanoindention measurements. The microstructure of DP1000A consists of a matrix containing ferrite (α) and bainite/tempered martensite (αb) with some dispersed martensite and fresh martensite/retained austenite (M/RA) islands. The SEM images show a large presence of carbide precipitates within B/TM grains (Figure 5.22a). According to magnetization saturation measurements, the RA volume fraction for DP1000A is 6%, which is much higher than the obtained by EBSD analysis (Figure 5.22a). The reasoning to justify the large differences between the two measuring techniques is further explained in Paper IV. Figure 5.22. SEM micrographs (left) and EBSD phase maps (right) for a) DP1000A, b) DP1000B. Chapter 5. Discussion 99 DP1000B has a ferritic-bainitic matrix with a slightly lower amount of martensite islands (≈27%), which are homogeneously distributed. Contrary to the observed in DP1000-A, carbide precipitation is hardly seen in bainitic (B) areas (Figure 5.22b). For this steel, an almost negligible amount of RA was detected by EBSD (0.09%, Figure 5.22b). The RA volume fraction obtained from magnetic measurements was 2%. Regarding the proportions of different phases, the differences between the two steel grades are not very large. Both steels show a similar distribution of bainite (harder in the case of DP1000- A). Optical micrographs and EBSD analysis revealed a larger proportion of ferrite in DP1000B (see Paper IV). The grain size is also similar for DP1000A and DP1000B. It is worth noting that the identification of phases volume fraction is a complex and challenging task that required the combination of multiple characterization techniques. A comparison of results from the different techniques led to similar estimations. The distribution of soft/hard phases is also illustrated by the nanoindentation mappings (Figure 5.23). These images confirm the larger fraction of softer regions (green) in DP1000B. Also, a greater presence of very hard (H>5.5 GPa, colour red) secondary phases is observed in DP1000- A, which can be associated with M/RA islands. Figure 5.23. Nanohardness mappings. a) DP1000-A and b) DP1000-B. The tensile curves and the results from EWF tests are shown in Figure 5.24 and Figure 5.25, respectively. Figure 5.24b also shows the evolution of the RA volume fraction as a function of the true strain. This image evidences the significant austenite to martensite transformation rate in DP1000A (TRIP effect), while this effect is barely seen in DP1000B. Both steel present quite similar mechanical properties in terms of strength and elongation, being slightly superior for DP1000A. As shown in Paper IV, DP1000A also shows superior strain hardening behaviour. On the other hand, they show very different cracking behaviour. DP1000B exhibits significantly higher fracture toughness at crack initiation (wei) and the overall fracture toughness (we) is practically the double than for DP1000A. These differences between the tensile properties and the crack propagation resistance of these two steels are mainly attributed to three factors: 1) volume fraction of fresh martensite/retained austenite, 2) proportion of ferrite in the matrix and 3) connectivity of the hard secondary phases. The larger RA volume fraction in DP1000A and the TRIP effect (Figure 5.24b) improves the strain-hardening behaviour and contribute to increase strength and elongation. On the other hand, as mentioned in the previous section, the transformation of austenite to martensite in DP1000A 5.4 Influence of microstructure on fracture toughness 100 can have a detrimental effect on fracture toughness due to the formation of a continuous network of hard bainite/tempered martensite and untempered fresh martensite ahead of the crack tip. This effect increases the connectivity of the hard phases and facilitates crack propagation. In the case of DP1000B, the greater presence of soft regions which a greater ability to accommodate plastic deformation slow down the crack propagation and increase the fracture toughness. More details about the fracture mechanisms and the crack propagation path in both DP steels are available in Paper IV. A fractographic analysis is also included. Figure 5.24. a) Engineering stress-strain curves and b) evolution of the RA volume fraction with deformation. Figure 5.25. EWF results. a) wf as a function of the ligament length. b) wfi for different ligament lengths and average wei value. 5.4.2 Retained austenite transformation during crack propagation in TRIP-assisted steels Owing to their superior strength and formability, 3rd Generation TRIP-assisted steels containing RA have become excellent candidates for substituting 1st Generation AHSS. This fact has focused the research efforts in the development of new steel grades belonging to this family. However, as evidenced in previous sections, understanding the role of the martensitic transformation on cracking resistance is of high importance. To accurately evaluate the evolution of the RA volume fraction with deformation is important to select the most adequate technique. RA volume fraction can be evaluated by different techniques such as EBSD or magnetization saturation measurements. However, in EBSD measurements, specimen preparation is critical and, as observed in the previous section and reported in different works [178.179], often the RA content is underestimated. In highly deformed zones, like in the areas surrounding the crack tip, this differentiation is even more difficult. On the other hand, Chapter 5. Discussion 101 magnetic measurements provide more accurate quantification of RA volume fraction. Nevertheless, whereas they are very useful to evaluate an overall RA content in relatively large areas (such as in tensile specimens), they are not suitable enough to evaluate local variations in RA volume fraction. In this work, the RA to martensite transformation during crack propagation has been investigated in three 3rd Generation TRIP-assisted steels (TBF, TBF/Q&P and Q&P) by means of ex-situ X- ray powder diffraction measurements performed at ALBA Synchrotron. As explained in Section 3.5.3, the measurements were performed in DENT specimens tested to different load-line displacements (1 specimen per deformation level). 4 deformation stages were defined as indicated in Table 5.3 (a schematic representation is given in Figure 3.18). Table 5.3 also shows the specimen designations used and the corresponding load-line displacement and load. Table 5.3. Definition of the different deformation stages studied. Material Specimen designation Stage definition Load-line displacement [mm] Load [N] TBF TBF St 0 No deformation 0.00 0 TBF St 1 Before initiation 0.09 14444 TBF St 2 Initiation 0.195 20826 TBF St 3 Before fracture 0.513 17730 TBF/Q&P TBF/Q&P St 0 No deformation 0.00 0 TBF/Q&P St 1 Before initiation 0.098 16592 TBF/Q&P St 2 Initiation 0.251 25090 TBF/Q&P St 3 Before fracture 0.426 23130 Q&P Q&P St 0 No deformation 0.00 0 Q&P St 1 Before initiation 0.117 14781 Q&P St 2 Initiation 0.250 21923 Q&P St 3 Before fracture 0.408 21383 For each specimen, 3 consecutive measurements separated 3.5 mm were performed close to the crack tip. Each measurement corresponds to a region identified as Zone I (crack tip), Zone II (≈ 5 mm ahead of the crack tip) and Zone III (centre of the ligament). These consecutive measurements allow detecting small variations in the martensitic transformation rate due to the strain gradients generated at the crack tip (Figure 5.26). Since the stress is higher at the crack tip and decreases with increasing the distance from the crack tip, the transformation rate should also increase in the near tip region. 5.4 Influence of microstructure on fracture toughness 102 Figure 5.26. Schematic representation of the stress field at crack tip (left) and strain gradients near the crack tip (right). The results obtained in this study are part of a publication currently under preparation. Some of the main results are shown and discussed below. Figure 5.27 to Figure 5.29 show the diffraction patterns obtained for the TBF, TBF/Q&P and Q&P at the different stages of deformation. The Miller indices (hkl) of planes corresponding to the austenite phase are indicated. For better visualization, an offset has been applied to the diffractograms of the different stages. The austenite phase fraction was derived from Rietveld refinements of the diffraction data, as described in Section 3.5.3. Figure 5.27. X-ray diffraction patterns for TBF. Results for Zone I. Figure 5.28. X-ray diffraction patterns for TBF/Q&P. Results for Zone I. Chapter 5. Discussion 103 Figure 5.29. X-ray diffraction patterns for Q&P. Results for Zone I. The RA volume fractions obtained for the different stages and for the different positions respect to the crack tip are shown in Figure 5.30 to Figure 5.32. At stage 1, only Q&P shows a slight decrease in RA volume fraction respect to the initial RA content (Vγ0). The Q&P steel shows the highest transformation rate (i.e. lower RA stability) and most of the RA is consumed before crack initiation (Stage 2). In TBF/Q&P, also a great part of the RA is transformed at Stage 2. However, TBF is still showing a significant amount of untransformed RA (7.5%), which indicates greater mechanical stability of RA. At the end of the propagation (Stage 3), most of the RA has been transformed in the 3 steels. As mentioned before, it can be observed that the transformation rate decreases with increasing the distance from the crack tip (from Zone 1 to Zone 3). Figure 5.30. TBF.RA volume fraction obtained for the different stages. The positions 1,2 and 3 correspond to the different measurement Zones (see the text for details). 5.4 Influence of microstructure on fracture toughness 104 Figure 5.31. TBF/Q&P.RA volume fraction obtained for the different stages. The position 1,2 and 3 correspond to the different measurement Zones (see the text for details). Figure 5.32. Q&P.RA volume fraction obtained for the different stages. The positions 1,2 and 3 correspond to the different measurement Zones (see the text for details). The different RA to martensite transformation rates observed in the three steels are in good agreement with the RA mechanical stability determined from uniaxial tensile tests (Figure 5.33a), i.e. TBF has the highest stability, Q&P the lowest and TBF/Q&P is an intermediate range. However, comparing the RA volume fraction as a function of the deformation in tensile and DENT specimens (Figure 5.33 b-d), it can be observed that the transformation rate is much higher in DENT specimens. This is caused by the large stress triaxiality at the crack tip, which promotes the mechanically-induced transformation at much lower levels of deformation [176]. From the analysis of the contribution of RA transformation to crack propagation resistance (Figure 5.34), no relation is observed between the amount of RA consumed during crack propagation and the difference between wei and we. The TBF steel shows a significant amount of Chapter 5. Discussion 105 untransformed RA at crack initiation (stage 2) that progressively transforms during propagation (stage 3). However, looking at fracture toughness results, no significant contribution from crack propagation resistance after crack initiation is observed. On the other hand, TBF/Q&P shows the largest differences between we and wei but the amount of RA transformed during the propagation is much lower than in TBF. Figure 5.33. a) Evolution of RA volume fraction (Vγ) with deformation in tensile tests (results from magnetic measurements). b-d) Comparison of the evolution of RA content with deformation in tensile and DENT specimens (Zone I). The arrows indicate the strain corresponding to crack initiation. Figure 5.34. RA volume fraction transformed during crack propagation (Vγ st3- Vγ st2) and difference between crack initiation (wei) and crack propagation resistance (we). 5.5 AHSS classification according to cracking resistance 106 5.5 AHSS classification according to cracking resistance The findings obtained in this work evidence that new failure criteria accounting for crack propagation resistance are increasingly necessary to better understand the overall fracture performance of AHSS sheets. According to this, and analogously to the global/local formability maps proposed by Hance and Davenport [52] or Heibel et al. [66], an alternative classification map is proposed in Paper II (Figure 5.35a). In this diagram, uniform elongation (UE) is plotted in the horizontal axis and the specific essential work of fracture (we) in the vertical axis. UE and we are used, respectively, as global formability and cracking resistance indices. On that basis, the diagram is divided into different quadrants according to the global formability and cracking resistance level. The more to the right the greater the global formability while upper quadrants indicate superior fracture resistance and damage tolerance. Additionally, an alternative diagram to the traditional “banana” plot (Figure 1.3) is also shown in Figure 5.35b. This classification system provides a more precise description of the fracture resistance of AHSS as a function of their strength level and can serve as a guide for future steel development and material selection. Figure 5.35. a) AHSS classification based on global formability (UE) and fracture resistance (we). LGF: low global formability, LCR: low cracking resistance, HGF: high global formability, HCR: high cracking resistance. b) Proposed diagram for classification of AHSS according to their strength level (UTS) and fracture resistance (we). 107 Chapter 6 Conclusions and future work 6.1 Conclusions In this thesis, the crack initiation and propagation resistance of a wide range of advanced high strength steel sheets has been studied in the frame of fracture mechanics. The relationship between fracture toughness results and fracture behaviour in cold forming (edge cracking) and during crash has been investigated. The conclusions reached from these investigations are listed below. Plane stress fracture toughness characterization methods Different fracture toughness testing methods including the Essential Work of Fracture (EWF) methodology, J-integral measurements according to ASTM E1820 and Kahn-type tear tests (KTT) have been used to evaluate the crack initiation and propagation resistance of AHSS sheets. From the comparison of the different fracture toughness methodologies and the obtained parameters, it is concluded that:  The EWF method is applicable to evaluate the crack initiation and propagation resistance of AHSS. The fulfilment of the conditions for obtaining a valid specific essential work of fracture (we) has been validated by means of DIC and stress analysis.  The estimation of stress levels in DENT specimens and the relationship between we and δc are significantly improved using an effective yield strength (σY) as defined in ASTM E1820 instead of the conventionally defined yield stress (σYS).  The ASTM E1820 can be applied to evaluate the J-integral of thin AHSS sheets if the thickness restrictions described in the standard are dismissed. An alternative method for online crack growth measurement based on high-resolution video extensometry is proposed. With this method, crack length estimation from compliance technique is avoided.  The ASTM B871, developed for tear testing of aluminium alloys by using Kahn-type tear tests, has shown to be applicable to characterize the tearing resistance of thin AHSS sheets.  The values of fracture toughness at crack initiation are shown to be independent of the testing methodology and the specimen geometry used. The identified crack initiation resistance parameters wei, Ji and UIE are found to be equivalent.  we has been shown to better quantify the crack propagation resistance of thin high strength sheets than Jc. we contains both the energy dissipated during crack initiation and crack propagation to failure, whereas Jc only has an energetic contribution from small crack extension and does not account for the complete fracture resistance. J-R curves are also unsuitable to represent the intrinsic fracture toughness of the material because they also include an extrinsic contribution from the plastic work dissipated out of the FPZ.  The additional contribution to the essential work of fracture developed in the FPZ after crack initiation comes mainly from necking. Thus, materials developing remarkable necking during crack propagation have greater crack propagation resistance. This has been verified by means of thickness strain measurements in the fracture surface of DENT specimens. References 114 Fracture Toughness to Understand Stretch-Flangeability and Edge Cracking Resistance in AHSS. Metall Mater Trans A 48 (2017) 86-94. [21] K. Sugimoto, M. Mukherjee. TRIP aided and complex phase steels. In: R.Rana, S.B Singh (Eds.), Automotive steels, chapter 8. Woodhead publishing (2017) 217-257. [22] T.B.Hilditch, T.de Souza, P.D.Hodgson. Properties and automotive applications of advanced high-strength steels (AHSS). In: M. Shome, M. Tumuluru (Eds.), Welding and Joining of Advanced High Strength Steels (AHSS), chapter 2. Woodhead Publishing (2015) 9-28. [23] N. Fonstein, H.J. Jun, G. Huang, S. Sriram, B. Yan. Effect of Bainite on Mechanical Properties of Multiphase Ferrite-Bainite-Martensite Steels. Proceedings from the Materials Science & Technology Conference, October 16-20, 2011, Columbus, Ohio [24] A. Nagathan, L. Penter. Hot stamping. In: T.Altan, A. Tekkaya (Eds.), Sheet Metal Forming Processes and Applications, chapter 7. ASM International (2012) 133-142. [25] L. Penter, N. Pierschel. Thermo-mechanical interactions in hot stamping. Conference on Thermal Issues in Machine Tools (Dresden, Germany, March 21-23, 2018). [26] S. Golling, D. Frómeta, D. Casellas, P. Jonsén. Influence of microstructure on the fracture toughness of hot stamped boron steel. Mat. Sci. and Eng. A 743 (2019) 529-539. [27] B. C. De Cooman, Kwang-geun Chin, Jinkyung Kim. High Mn TWIP Steels for Automotive Applications New Trends and Developments in Automotive System Engineering. Prof.Marcello Chiaberge, InTech (2011). https://doi.org/10.5772/14086 [28] R.E. Schramm, R.P. Reed. Stacking fault energies of seven commercial austenitic stainless steels. Metall Mater Trans A 6 (1975) 1345-1351. [29] J.Speer, D.K. Matlock, B.C. De Cooman, J.G. Schroth. Carbon partitioning into austenite after martensite transformation. Acta Mater. 51 (2003) 2611–2622. [30] D.V. Edmonds, K. He, F.C. Rizzo, B.C. De Cooman, D.K. Matlock, J.G. Speer. Quenching and partitioning martensite- A novel steel heat treatment. Mat. Sci. and Eng. A 438–440 (2006) 25–34. [31] R.L. Miller. Ultrafine-grained microstructures and mechanical properties of alloy steels. Metall Mater Trans B 3 (1972) 905–912. [32] H. Luo, J. Shi, C. Wang, W. Cao, X. Sun, H. Dong. Experimental and numerical analysis on formation of stable austenite during the intercritical annealing of 5Mn steel. Acta Mater. 59 (2011) 4002. [33] M.J. Merwin. Low-Carbon Manganese TRIP Steels. Mater. Sci. Forum. 539–543 (2007) 4327. [34] J. Shi, X. Sun, M. Wang, W. Hui, H. Dong, W. Cao. Enhanced work-hardening behavior and mechanical properties in ultrafine-grained steels with large-fractioned metastable austenite. Scripta Mater. 63 (2010) 815. [35] S. Lee, S.J. Lee, B.C. De Cooman. Austenite stability of ultrafine-grained transformationinduced plasticity steel with Mn partitioning. Scripta Mater. 65 (2011) 225. [36] Z.H. Cai, H. Ding, R.D.K. Misra, Z.Y. Ying. Austenite stability and deformation behavior in a cold-rolled transformation-induced plasticity steel with medium manganese content. Acta Materialia 84 (2015) 229-236. [37] C. Liu, Q. Peng, Z. Xue, S. Wang, C. Yang. Microstructure and Mechanical Properties of Hot-Rolled and Cold-Rolled Medium-Mn TRIP Steels. Materials 11(11) (2018) 2242. [38] L. Liu, B. He, M. Huang. The Role of Transformation-Induced Plasticity in the Development of Advanced High Strength Steels. Adv. Eng. Mater. (2018) 1701083. References 115 [39] S. Kaar, D. Krizan, J. Schwabe, H. Hofmann, T. Hebesberger, C. Commenda, L. Samek. Influence of the Al and Mn content on the structure-property relationship in density reduced TRIP-assisted sheet steels. Mat. Sci. and Eng. A 735 (2018). [40] H.L.Yi. Review on δ-Transformation-Induced Plasticity (TRIP) Steels with Low Density: The Concept and Current Progress. JOM: the journal of the Minerals, Metals & Materials Society 66 (9) 1759-1769. [41] H. L. Yi, K. Y. Lee, H. K. D. H. Bhadeshia. Extraordinary ductility in Al-bearing δ-TRIP steel. Proc. R. Soc. A (2011) 467, 234–243. [42] S. Chatterjee, M. Murugananth, H. K. D. H. Bhadeshiaet. δ-TRIP steel. Materials Science and Technology 23 (20007) No. 7 819-827. [43] A. Konieczny, T. Henderson. On Formability Limitations in Stamping Involving Sheared Edge Stretching. SAE Technical Paper 2007-01-0340, 2007. [44] K. Mori, Y. Abe, Y. Suzui. Improvement of stretch flangeability of ultra high strength steel sheet by smoothing of sheared edge. J. Mater. Process. Tech. 210 (2010) 653–659. [45] C.R.M Silva, F.J.G. Silva, R.M. Gouveia. Investigations on the edge crack defect in Dual Phase steel stamping process. Procedia manufacturing 17 (2018) 737-745. [46] M.S. Walp MS. Impact Dependent Properties of Advanced and Ultra High Strength Steel. SAE Technical Paper 2007-01-0342, 2007. [47] P. Larour, H. Pauli, T. Kurz, T. Hebseberger. Influence of post uniform tensile and bending properties on the crash behaviour of AHSS and press-hardening steel grades. Proceedings of the IDDRG2010 (Graz, Austria) 2010. [48] P. Larour, J. Naito, A. Pichler, T. Kurz, T. Murakami. Side impact crash behavior of presshardened steels-correlation with mechanical properties. 5th Int. Conf. Hot sheet metal forming of high performance steel (CHS2) (Toronto, Canada, May 31- June 3 2015) pp 281-289. [49] T.M. Link, B.M. Hance. Axial and Bending Crash Performance of Advanced High- Strength Steels. Int. Symp. on New Developments in Advanced High-Strength Steels. Keystone, Colorado, USA. 2017. [50] F. Hisker, R. Thiessen, T. Heller. Influence of Microstructure on Damage in Advanced High Strength Steels. Materials Science Forum 706–709 (2012) 925–930. [51] P. Efthymiadis, S. Hazra, A. Clough, R. Lakshmi, A. Alamoudi, R. Dashwood, B. Shollock. Revealing the mechanical and microstructural performance of multiphase steels during tensile, forming and flanging operations. Mat. Sci. and Eng. A 701 (2017) 174-186. [52] B. Hance, M.D. Davenport. Advanced High Strength Steel: Deciphering Local and Global Formability. Int. Automotive Body Congress 2016, September 28-29, Dearborn, Michigan, USA. [53] S. Kalpakjian, S. Schmid. Manufacturing Engineering and Technology, 6th edition. Prentice Hall, 2010. [54] Y. Bao. Dependence of ductile crack formation in tensile tests on stress triaxiality, stress and strain ratios. Eng. Frac. Mech. 72 (2005) 505–522. [55] Z. Marciniak, J.L. Duncan, S.J. Hu. Mechanics of sheet metal forming, 2nd edition. Butterworth-Heinemann, Oxford 2002. [56] J. Hu, Z. Guo, N. Saunders. A computational approach for forming limit diagram (FLD) evaluation at elevated temperatures for aluminium alloys. Proceedings of the 16th International Aluminum Alloys Conference (ICAA16) (Montréal, Canada) 2018. [57] ISO 12004-2:2008. Metallic Materials—Sheet and Strip–Determination of Forming-Limit Curves—Part 2: Determination of Forming-Limit Curves in the Laboratory. International References 116 Standardization Organization (ISO). Geneva, Switzerland, 2008. [58] G. Chinouilh, F. Toscan, P. Santacreu, J. Leseux. Forming Limit Diagram Prediction of Stainless Steels Sheets. SAE Technical Paper 2007-01-0338, 2007. [59] M. Abspoel, M. E. Scholting, J. M. M. Droog. A new method for predicting Forming Limit Curves from mechanical properties. J Mater Process Tech 213 (2013) 759– 769. [60] ISO 16630:2017. Metallic materials — Sheet and strip — Hole expanding test. International Organization for Standardization. [61] E. Atzema, M. Borsutzki, M. Braun, S. Brockmann, M. Bülter, B. Carlsson, P. Larour, A. Richter. A European round robin test for the hole expansion test according to ISO 16630. New Development in Sheet Metal Forming 2012 International Conference. (Fellbach, Germany, May 23-24, 2012), pp. 171-184. [62] M. Schneider, U. Eggers. Investigation on punched edge formability. Proceedings of International Deep Drawing Research Group 2011 conference. (Bilbao, Spain, June 5-8, 2014). [63] VDA 238-100, Plate bending test for metallic materials. Verband der Automobilindustrie e.V.(VDA): Berlin, Germany, 2017. [64] ASTM E8/E8M-15a (2015) Standard Test Methods for Tension Testing of Metallic Materials. ASTM International, West Conshohocken, PA, USA. [65] P. Larour, J. Freudenthaler, T.Weissböck. Reduction of cross section area at fracture in tensile test: measurement and applications for flat sheet steels. Journal of Physics: Conf. Series 896 (2017) 012073. [66] S. Heibel, T. Dettinger, W. Nester, T. Clausmeyer, A.E. Tekkaya. Damage Mechanisms and Mechanical Properties of High-Strength Multiphase Steels. Materials 2018, 11, 761. [67] B. Hance. Practical Application of the Hole Expansion Test. SAE Int. J. Engines 10(2):2017 [68] X. Chen, J. Sun, X. Zhu. Simulation Technique for Pre-forming of AHSS Edge Stretching 12th International LS-DYNA® Users Conference (Michigan, USA, June 03-05) 2012. [69] P. Larour, J. Freudenthaler, A. Grünsteidl, K.Wang. Evaluation of alternative stretch flangeability testing methods to ISO 16630 standard. Proceedings of International Deep Drawing Research Group 2014 conference. (Paris, France, June 1-4, 2014), pp.188-193. [70] P. Larour, H. Schauer, J. Lackner, E.T. Till. Edge crack simulation with the modular “smiley” forming tool. Proceedings of International Deep Drawing Research Group 2016 conference. (Linz, Austria, June 12-15, 2016), pp.457-459. [71] J. Dykeman, S. Malcolm, B. Yan, J. Chintaman, G.Huang, N. Ramisett, H. Zhu. Characterization of Edge Fracture in Various Types of Advanced High Strength Steel. SAE Technical Paper 2011-01-1058, 2011. [72] K. Watanabe, M. Tachibana, K. Koyanagi and K. Motomura. Simple Prediction Method for the Edge Fracture of Steel Sheet During Vehicle Collision – Evaluation of Fracture Limit from the Edge Using Small-Sized Test Pieces. 5th LS-Dyna User Forum (Ulm, Germany, October 12, 2006). [73] M. Feistle, M. Krinninger, I. Pätzold,W.Volk. Edge-Fracture-Tensile-Test. In: Tekkaya A., Homberg W., Brosius A. (Eds.) 60 Excellent Inventions in Metal Forming. Springer Vieweg, Berlin, Heidelberg (2015) 193-198. [74] X. Wu, H. Bahmanpour, K. Schmid. Characterization of mechanically sheared edges of dual phase steels. J. Mat. Proc. Tech. 212 (2012) 1209-1224 [75] D. Gutiérrez, Ll. Pérez, A. Lara, D. Casellas, J.M. Prado. The influence of fracture toughness in stretch-flangeability in advanced high strength steels sheets. Proceedings of the IDDRG 2013 (Zurich, Switzerland, June 02-05) References 117 [76] Y. Takahashi, O. Kawano, Y. Tanaka. Fracture Mechanical Study on Edge Flange-ability of High Tensile-strength Steel Sheets. MS&T 2009: Proceedings from the Materials Science & Technology Conference (October 25-29, 2009, Pittsburgh, Pennsylvania) pp 1317-1328 [77] P. Larour, H. Pauli, J. Freudenthaler, A. Grünsteidl. Alternative stretch flangeability characterisation methods for AHSS steel grades. Proceedings of International Deep Drawing Research Group 2011 conference. (Bilbao, Spain, June 5-8, 2014). [78] M. Feistle, R. Golle, W. Volk. Determining the influence of shear cutting parameters on the edge cracking susceptibility of high-strength-steels using the edge-fracture-tensile-test. Procedia CIRP 41 ( 2016 ) 1078 – 1083. [79] S. Nasheralahkami, S. Golovashchenko, K. Pan, L. Brown, B. Gugnani. Characterization of Trimmed Edge of Advanced High Strength Steel. SAE Technical Paper 2016-01-0358, 2016 [80] J.I. Yoon, J. Jung, S.H. Joo, T.J. Song, K.G. Chin, M.H. Seo, S.J. Kim, S. Lee, H.S. Kim. Correlation between fracture toughness and stretch-flangeability of advanced high strength steels. Matter.Lett. 180 (2016) 322-326 [81] P. Larour, J. Lackner, L. Wagner. Influence of single hat crash box flange triggering and impactor top plate welding strategy on axial crash foldability of AHSS & UHSS sheets. IOP Conf. Ser.: Mater. Sci. Eng. 651 012023 [82] T. Kurz, P. Larour, J. Lackner, T. Steck, G. Jesner. Press-hardening of zinc coated steel – characterization of a new material for a new process. Proceedings of International Deep Drawing Research Group 2016 Conference (Linz, Austria, June 12-15, 2016). [83] P. Dietsch, K. Tihay, S. Cobo, S. Sarkar, D. Hasenpouth, D. Cornette. Predictive approach for crash performance of Press Hardened Steels and its application on new product developments. 6th Int conf hot sheet metal forming of high performance steel (CHS2) 2017. [84] G. Lacroix, T. Pardoen, P.J. Jacques. The fracture toughness of TRIP-assisted multiphase steels. Acta Mater. 56 (2008) 3900–13. [85] R. Wu, J. Li, W. Li , X.C. Wu, X. Jin, S. Zhou, L. Wang. Effect of metastable austenite on fracture resistance of quenched and partitioned (Q&P) sheet steels. Mater. Sci. Eng. A 657 (2016) 57–63. [86] S. Sahoo, N. Padmapriya, P. S. De, P.C. Chakraborti, S.K. Ray. Ductile tearing resistance indexing of automotive grade DP590 steel sheets: EWF testing using DENT specimens. J. Mater. Eng. Perform. 27 (2018) 2018–23. [87] Z.C. Luo, R.D. Liu, X. Wang, M.X. Huang. The effect of deformation twins on the quasicleavage crack propagation in twinning-induced plasticity steels. Acta Mater. 150 (2018) 59–68. [88] M. Faccoli, G. Cornacchia, M. Gelfi, A. Panvini, R. Roberti. Notch ductility of steels for automotive components. Eng. Frac. Mech. 127 (2014) 181–193 [89] Z. Xiong, P.J. Jacques, A. Perlade, T. Pardoen. Ductile and intergranular brittle fracture in a two-step quenching and partitioning steel. Scripta Mater. 157 (2018) 6–9. [90] Z. Xiong, P.J. Jacques, A. Perlade, T. Pardoen. Characterization and Control of the Compromise Between Tensile Properties and Fracture Toughness in a Quenched and Partitioned Steel. Metall. and Mat. Trans. A 50 (2019) 3502-3513. [91] R. Sarkara, S.K. Chandra, P.S. De, P.C. Chakraborti, S.K. Ray. Evaluation of ductile tearing resistance of dual-phase DP 780 grade automotive steel sheet from essential work of fracture (EWF) tests. Theor Appl Fract Mec 103 (2019) 102278. [92] D. Broek. Elementary engineering fracture mechanics. Springer Netherlands 1982. DOI 10.1007/978-94-009-4333-9. References 118 [93] A. Griffith. The Phenomena of Rupture and Flow in Solids. Philosophical Transactions of Royal Society A 221 (1921) 163. [94] C.E. Inglis. Stress in a Plate Due to the Presence of Cracks and Sharp Corners. Transactions of the Institute of Naval Architects 55 (1913) 219. [95] J.R. Rice. A path independent Integral and the Approximate Analysis of Strain Concentrations by Notches and Cracks. Journal of Applied Mechanics 34 (1968) 379. [96] J.R. Rice, G.F. Rosengren. Plane Strain Deformation Near a Crack Tip in a Power-law Hardening Material. Journal of the Mechanics and Physics of Solids 16 (1968) 1-12. [97] J.W. Hutchinson. Singular Behavior at the End of a Tensile Crack in a Hardening Material Journal of the Mechanics and Physics of Solids 16 (1968) 13-31. [98] G.R. Irwin. Fracture Dynamics. In: Fracturing of Metals; American Society of Metals: Cleveland, 1948. [99] E. Orowan. Fracture strength of solids. In: Report on Progress in Physics, Vol. 12; Physical Society of London: London, 1949 185–232. [100] H.M. Westergaard. Bearing pressures and cracks. Trans. ASME J. Appl. Mech. 6 (1939) 49–53 [101] T.L. Anderson. Fracture mechanics: fundamentals and applications. 3rd ed. CRC Press; 2005 [102] C.H.Wang. Introduction to Fracture Mechanics. (1996) DOI:10.13140/RG.2.1.1444.2408 [103] Fracture toughness properties of aerospace materials. In: A. P. Mouritz (Ed.), Introduction to Aerospace Materials, chapter 19. Woodhead publishing (2012) 454-468. [104] ASTM E399-90. Standard test method for fracture toughness of metallic materials. American Society for Testing and Materials. Philadelphia, EEUU 1990. [105] J.L. Arana, J.J. González. Mecánica de fractura. Servicio editorial Universidad del País Vasco. [106] A.A. Wells. Unstable Crack Propagation in Metals: Cleavage and Fast Fracture. Proceedings of the Crack Propagation Symposium, Vol. 1, Paper 84, Cranfield, UK, 1961. [107] C.T. Sun, Z.-H. Jin. Elastic-Plastic Fracture Criteria. In: C.T. Sun, Z.-H. Jin (Eds.), Fracture Mechanics, chapter 7. Academic Press (2012) 171-187. [108] J.A. Begley, J.D. Landes. The J-Integral as a Fracture Criterion. ASTM STP 514, American Society for Testing and Materials, Philadelphia, PA, (1972) 1–20. [109] J.D. Landes, J.A. Begley. The Effect of Specimen Geometry on JIc. ASTM STP 514, American Society for Testing and Materials, Philadelphia, PA, (1972) 24–29. [110] C.F. Shih. Relationship between the J-Integral and the Crack Opening Displacement for Stationary and Extending Cracks. Journal of the Mechanics and Physics of Solids, 29 (1981) 305–326. [111] ASTM E1820. Standard test method for measurement of fracture toughness. American Society for Testing and Materials. [112] ASTM E2472. Standard test method for determination of resistance to stable crack extension under low-constraint conditions. American Society for Testing and Materials. [113] ISO 22889. Metallic materials — Method of test for the determination of resistance to stable crack extension using specimens of low constraint. International Organization for Standardization. [114] J. Chen, Y. Verreman, J. Lanteigne. On fracture toughness JIC testing of martensitic stainless steels. 13th International Conference on Fracture June 16–21, 2013, Beijing, China. References 119 [115] S. Zhang, S. Zhou, M. Li, B. Fu. Calculation and comparison on fracture toughness of specific reliability between ASTM and ISO standards. Mater. Res. Express 7 (2020) 026529. [116] N.M. Abd-Allah, M.S. El-Fadaly, M.M. Megahed, A.M. Eleiche. Fracture toughness properties of high-strength martensitic steel within a wide hardness range. J. of Materi Eng and Perform 10 (2001) 576–585. [117] X.K. Zhu, B.N. Leis. Revisit of ASTM round robin test data for determining R curves of thin sheet materials. J ASTM Int 2009 [paper ID JAI102510]. [118] Marchal Y, Schmidt K, Pardoen T, Knockaert R, Doghri I and Delannay F 1996 Comparison methods for the measurement of fracture toughness of thin sheets: (ECF 11) Mechanisms and Mechanics of Damage and Failure 2259-2265. [119] T. Pardoen, Y. Marchal, F. Delannay. Thickness dependence of cracking resistance in thin aluminium plates. J. of Mech. Sol. 47 (1999) 2093-2123. [120] J.C. Newman Jr., M.A. James, U. Zerbst. A review of the CTOA/CTOD fracture criterion. Eng. Frac. Mech. 70 (2003) 371–385. [121] J. Heerens, M.Schödel. On the determination of crack tip opening angle, CTOA, using light microscopy and δ5 measurement technique. Eng. Frac. Mech. 70 (2003) 417–426. [122] S. Koley, S. Chatterjee, M. Shome. Evaluation of fracture toughness of thin sheet of interstitial free high strength steel through critical crack tip opening angle (CTOAc) measurement. Int J Fract 194 (2015) 187–195. [123] S. Xu, N. Petri, W.R. Tyson. Evaluation of CTOA from load vs. load-line displacement for C(T) specimen. Eng. Frac. Mech. 76 (13) (2009) 2126-2134. [124] M. Ben Amara, G. Pluvinage, J. Capelle, Z. Azari. Crack Tip Opening Angle as a Fracture Resistance Parameter to Describe Ductile Crack Extension and Arrest in Steel Pipes under Service Pressure. Phys Mesomech 18 (2015) 355–369. [125] J.G. Kaufman, A.H. Knoll. Kahn-Type Tear Tests and Crack Toughness of Aluminum Sheet. Metals Research and Standards (1964) 151–155. [126] ASTM B871. Standard Test Method for Tear Testing of Aluminum Alloy Products. American Society for Testing and Materials. [127] J.G. Kaufman, H. Hunsicker. Fracture Toughness Testing at Alcoa Research Laboratories. Fracture Toughness Testing and its Applications (1965) 290-308 [128] G.G. Garret, J.F. Knott. The influence of compositional and microstructural variations on the mechanism of static fracture in aluminum alloys. Metal. Trans. A 9 (1978) 1187-1201. [129] D. Dumont, A. Deschamps, Y. Brechet. On the relationship between microstructure, strength and toughness in AA7050 aluminum alloy. Mat. Sci. and Eng. A 356 (2003) 326- 336. [130] P. Henn, M. Liewald, M. Sindel. Investigation on crashworthiness characterisation of 6xxx series aluminium sheet alloys based on local ductility criteria and edge compression tests. IOP Conf. Ser.: Mater. Sci. Eng. (2018) 418 012125. [131] L. Ying, J. Lu, Y. Chang, X. Tang, P. Hu, K. Zhao. Optimization evaluation test of strength and toughness parameters for hot-stamped high strength steels. J. of Iron and Steel Research Int. 20 (2013) 51-56. [132] J. Lorthios, A.F. Gourgues, P. Cugy, C. Scott. Damage of TWIP steels for automotive application. ICF12 Int. Conference on Fracture, Ottawa (2009). [133] B. Cotterell, J.K. Reddell. The Essential Work of Plane Stress Ductile Fracture. Int. J. Fracture (1977) 267-277. References 120 [134] K.B. Broberg. On stable crack growth. J. Mech. Phys. Solids. 23 (1975) 215-237. [135] J. Wu, Y.W. Mai. The essential fracture work concept for toughness measurement of ductile polymers. Polym. Eng. Sci. 36 (1996) 2275–88. [136] Y.W. Mai, B. Cotterell. On the essential work of ductile fracture in polymers. Int. J. Fract. 32 (1986) 105–25. [137] Y.W. Mai, P. Powell. Essential work of fracture and J-integral measurements for ductile polymers. J. Polym. Sci. 29 (1991) 785–93. [138] E. Clutton. Essential work of fracture. Moore DR, Pavan A, Williams JG, editors. Fracture mechanics testing methods for polymers, adhesives and composites, vol.28. ESIS Publ.; (2001) 177–95. [139] B. Cotterell, T. Pardoen, A.G. Atkins. Measuring toughness and the cohesive stress– displacement relationship by the essential work of fracture concept. Eng. Fract. Mech. 72 (2005) 827–48. [140] Y. Marchal, J.F. Walhin, F. Delannay. Statistical procedure for improving the precision of the measurement of the essential work of fracture of thin sheets. Int. J. Fract. 87 (1997) 189–99 [141] A.S. Saleemi A S, J.A. Nairn. The plane-strain essential work of fracture as a measure of the fracture toughness of ductile polymers Polym.Eng. Sci. 30(4) (1990) 211. [142] A.B. Martínez, J. Gámez-Pérez, M. Sanchez-Soto, J.I.Velasco, O.O. Santana, M.LL. Maspoch. The Essential Work of Fracture method- Analyzing the Post-yielding Fracture Mechanics of polymers. Eng. Fail. Anal. 16 2604-2617. [143] Y. Marchal, F. Delannay. Influence of test parameters on the measurement of the essential work of fracture of zinc sheets. Int Fract J 80 (1996) 295–310. [144] S. Hashemi, D. O'brien. The essential work of plane-stress ductile fracture of poly(etherether ketone) thermoplastic. Journal of Materials Science 28 (1993) 3977-3982. [145] ASTM E561. Standard practice for R-curve determination. American Society for Testing and Materials. [146] W.C. Oliver, G.M. Pharr. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 7 (6) (1992) 1564-1583. [147] F. Fauth, I. Peral, C. Popescu and M. Knapp. The new Materials Science and Powder diffraction beamline at ALBA Synchrotron. Powder Diffraction 28 (S2) (2013) pp. S360- S370. [148] F. Fauth, R. Boer, F. Gil-Ortiz, C. Popescu, O. Vallcorba, I. Peral, D. Fulla, J. Benach, J. Juanhuix. The crystallography stations at the Alba synchrotron. Eur. Phys. J. Plus (2015) 130: 160. [149] J. Rodríguez-Carvajal. Recent advances in magnetic structure determination by neutron powder diffraction. Physica B: Condensed Matter 192 (1993) 55-69. [150] S. Hashemi. Fracture toughness evaluation of ductile polymeric films. Journal of Materials Science 32, 1563–1573 (1997). [151] W.Y.F. Chan, J.G. Williams. Determination of the fracture toughness of polymeric films by the essential work method. Polymer 35 (8) (1994) 1666-1672. [152] N. León, A.B. Martínez, P. Castejón, D. Arencón, P.P. Martínez. The fracture testing of ductile polymer films: Effect of the specimen notching. Polymer Testing 63 (2017) 180- 193. [153] A.B. Martínez, N. León, A. Segovia, J. Cailloux, P.P. Martínez. Effect of specimen notch References 121 quality on the essential work of fracture of ductile polymer films. Engineering Fracture Mechanics 180 (2017) 296-314. [154] Y.W. Mai, K.M. Pilko. The essential work of plane stress ductile fracture of a strain-aged steel. J. Mater. Sci. 14 (1979) 386 - 394. [155] Y. Marchal, F. Delannay. Comparison of methods for fracture toughness testing of thin low carbon steel plates. Mater. Sci. Tech. 14 (1998) 1163–1168. [156] T. Pardoen, F. Hachez, B. Marchioni, P.H. Blyth, A.G. Atkins. Mode I fracture of sheet metal. J. of the Mech. and Phys. of Sol. 52 (2004) 423-452. [157] F. Tuba, L. Oláh, P. Nagy. The role of ultimate elongation in the determination of valid ligament range of essential work of fracture tests. J. Mater. Sci. 47 (2012) 2228–2233. [158] Y.W. Mai, B. Cotterell. Effects of pre-strain on plane-stress ductile fracture in α -brass, J.Mater.Sci.13 (1980) 2296-2306. [159] R. Muñoz, A. Lara, D. Casellas. Fracture toughness characterization of advanced high strength steels. Int. Deep Drawing Research Group (IDDRG) Conference 2011 (Bilbao, Spain, June 5-8, 2011) [160] D. Gutiérrez, Ll. Pérez, A. Lara, D. Casellas, J.M. Prado. Toughness evaluation of high strength steels sheets by means of the essential work of fracture. 19th European conference on fracture: fracture mechanics for durability, reliability and safety, ECF 2012. [161] S.K. M R, E. Schmidova, P. Konopík, D, Melzer, F. Bozkurt, NV Londe. Fracture Toughness Analysis of Automotive-Grade Dual-Phase Steel Using Essential Work of Fracture (EWF) Method. Metals 2020, 10, 1019. [162] A. Gray. Testing protocol for essential work of fracture. ESIS, editor. 1993, European Structural Integrity Society (ESIS) – TC4. [163] E. Clutton. Testing protocol for essential work of fracture. ESIS, editor. 1997, European Structural Integrity Society (ESIS) – TC4. [164] J.G. Williams, M. Rink. The standardisation of the EWF test. Eng Fract Mech 74(7) (2007) 1009–1017. [165] R. Hill. On discontinuous plastic states, with special reference to localized necking in thin sheets. Journal of Mechanics and Physics of Solids 1952; 1: 19-30. [166] T. Pardoen, Y. Marchal, F. Delannay. Essential work of fracture compared to fracture mechanics—towards a thickness independent plane stress toughness. Eng Fract Mech 69 (2002) 617–631. [167] O. Akourri, M. Louah, A. Kifani, G. Gilgert, G. Pluvinage. The effect of notch radius on fracture toughness JIC. Eng. Frac. Mech. 65 (2000) 491-505. [168] V.V. Chaudhari, D.M. Kulkarni, R. Prakash. Study of influence of notch root radius on fracture behaviour of extra deep drawn steel sheets. Fatigue Fract. Engng. Mater. Struct. 32 (2009) 975–986. [169] A.B. Martínez, A. Segovia, J. Gamez-Perez, M.Ll. Maspoch. Influence of femtolaser notch sharpening technique in the determination of essential work of fracture (EWF) parameters, Eng. Fract. Mech. 76 (9) (2009) 1247-1254. [170] J.R. Rice, P.C. Paris, J.G. Merkle. Some Further Results of J-Integral Analysis and Estimates. Progress in Flaw Growth and Fracture Toughness Testing. STP 536 (1973) 231- 245. [171] H. Tada, P.C. Paris, G.R. Irwin. The stress analysis of cracks handbook. Third edtition. American Society of Mechanical Engineers, New York, 2000. [172] L. Xia, C.F. Shih, J.W. Hutchinson. A computational approach to ductile crack growth References 122 under large scale yielding conditions. J. Mech. Phys. Solids 43 (1995) 389-413. [173] X.K. Zhu, S.K. Jang. J-R curves corrected by load-independent constraint parameter in dcutile crack growth. Eng. Frac. Mech. 68 (2001) 285-301. [174] Y. Mai, B. Cotterell. Effect of specimen geometry on the essential work of plane stress ductile fracture. Eng. Frac.Mech. 21 (1985) 123-128. [175] B. Cotterell, A.G. Atkins. A review of the J and I integrals and their implications for crack growth resistance and toughness in ductile fracture. Int. J. Fract. 81 (1996) 357-372. [176] P. Jacques, Q. Furnémont, T. Pardoen, F. Delannay. On the role of martensitic transformation on damage and cracking resistance in trip-assisted multiphase steels. Acta Mater 49 (2001) 139–152. [177] I. de Diego-Calderón, I. Sabirov, J.M. Molina-Aldareguia, C. Föjer, R. Thiessen, R.H. Petrov. Microstructural design in quenched and partitioned (Q&P) steels to improve their fracture properties. Mat. Sci. and Eng. A 657 (2016) 136. [178] R. Petrov, L. Kestens, A. Wasilkowska, Y. Houbaert. Microstructure and texture of a lightly deformed TRIP-assisted steel characterized by means of the EBSD technique. Mat Sci Eng A 447 (2007) 285–297. [179] T. Bhattacharyya, S.B. Singh, S. Das, A. Haldar, D. Bhattacharjee. Development and characterisation of C–Mn–Al–Si–Nb TRIP aided steel. Mat Sci Eng A 528 (2011) 2394– 2400. 123 Appendix A Appended publications 233 Paper C. D. Frómeta, A. Lara, B. Casas, D. Casellas. Fracture toughness measurements to understand local ductility of advanced high strength steels. IOP Conf. Ser.: Mater. Sci. Eng. 651 (2019) 012071 234 Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Published under licence by IOP Publishing Ltd International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 1 Fracture toughness measurements to understand local ductility of advanced high strength steels D. Frómeta1*, A. Lara1, B. Casas1 and D. Casellas1,2 1 Eurecat, Centre Tecnològic de Catalunya, Unit of Metallic and Ceramic Materials, Plaça de la Ciència, 2, Manresa 08243, Spain 2 Division of Mechanics of Solid Materials, Luleå University of Technology, 971 87 Luleå, Sweden *Corresponding author: [email protected] Abstract. The determination of the material parameters that best predict the local ductility of high strength sheet materials has become the focus of active research. Even though several correlations have been proposed, they can sometimes be not accurate enough and discussion is still open on this topic. This paper investigates the suitability of different fracture toughness measurements for local ductility prediction in multiple advanced high strength steels (AHSS). Fracture toughness is characterized by means of essential work of fracture and Khan tear tests. The results show that the essential work of fracture, we, correlates well with different local formability (HER, critical bending angle from V-bending tests and local strain at fracture from uniaxial tensile tests) and crash resistance parameters (energy absorbed in axial impact tests). It confirms that fracture toughness, measured in the frame of fracture mechanics, is a relevant material property to rationalize cracking issues associated to the local ductility of AHSS. On the other hand, it is also shown that Khan tear tests, which are conventionally used to evaluate the fracture resistance of thin metal sheets, can overestimate crack propagation resistance and offer a poor prediction ability for local formability and crash performance. 1. Introduction A wide variety of new high strength sheet materials have been developed in the last years for automotive lightweight applications. The limited ductility of these materials has posed new forming challenges that cannot be rationalized through conventional fracture characterization criteria. This fact has motivated the development of alternative characterization methodologies and improved formability mappings accounting for global and local formability. Global formability refers to the material resistance against necking instability and is well described by traditional tensile parameters (true uniform strain, elongation at fracture, n-value) and forming limit diagrams. Nevertheless, these tests provide little information regarding local formability issues (edge cracking, fractures occurring during bending on tight radius, crash folding behaviour). Thus, new experimental approaches are necessary to predict this kind of fractures associated to the local ductility of the material. In this sense, recent research works have demonstrated that fracture toughness, measured within the frame of fracture mechanics, is a relevant material property to describe such cracking related problems in AHSS [1-5]. Many works have shown the suitability of fracture toughness to rationalize and classify the stretch flangeability of high strength steels [1-4]. More recently, the measurement of fracture toughness has been also used in [5] to understand the crash failure behaviour of different AHSS grades. Therefore, it is evident that there is a close relationship between the crack propagation resistance of high International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 2 strength metal sheets and its local ductility. Nevertheless, the measurement of fracture toughness according to elastic plastic fracture mechanics (EPFM) standards [6] is rather complex and involves exhaustive specimen preparation and test monitoring, which hamper its implementation as a routine testing in automotive industry. There exist alternative simpler methods to characterize the fracture toughness of thin metal sheets, such as the essential work of fracture (EWF) methodology [7]. The EWF methodology is easier than standard methods since it permits to obtain the material crack initiation and propagation resistance without measuring the crack advance during the test, which is one of the main experimental challenges in EPFM procedures. Toughness values obtained from the EWF methodology have shown to be suitable to predict cracking related phenomena in AHSS sheets, such as edge cracking [1, 2] and crack propagation under crash loading [5]. Another method frequently used to characterize the fracture resistance of thin metal sheets is the Khan Tear Test (KTT). It has been extensively used to characterize the notch resistance of precipitation hardening aluminum alloys [8-10] and to evaluate toughness in different microstructures obtained by hot stamping of 22MnB5 steels [11] and in TWIP steels [12]. The main advantage of KTTs is that they are very simple tests and provide an estimation of the crack propagation resistance of the material. The aim of this work is twofold; firstly, to determine the fracture resistance of several AHSS grades by means of these two methodologies, the EWF and the KTT; and secondly to assess the correlation between fracture toughness and local ductility in AHSS. The ability of the proposed methodologies to predict local ductility, as well as their main advantages and drawbacks, are discussed. Fracture toughness results are compared with different local formability and crash resistance parameters widely applied in the automotive sector: • HER according to ISO 16630 • Bending angle from V-bending tests according to VDA 238-100 • Local strain at fracture from uniaxial tests obtained by means of Digital Image Correlation (DIC) • Energy absorbed in axial impact tests 2. Fracture toughness measurements 2.1. Essential Work of Fracture The Essential Work of Fracture (EWF) methodology was developed as an alternative method to quantify the ductile tearing resistance of thin ductile metal sheets [7]. The methodology permits to partition the total work of ductile fracture (Wf) in two energetic contributions: an essential work of fracture (we), spent in the fracture process zone and necessary to create new surfaces in the front of the crack tip and a nonessential plastic work (wp) surrounding the fracture area. The first term is proportional to the fracture surface and the second is proportional to the plastic volume, according to: 𝑊𝑊 𝑓𝑓=𝑤𝑤𝑒𝑒𝑙𝑙0𝑡𝑡0+𝑤𝑤𝑝𝑝𝛽𝛽𝑙𝑙0 2𝑡𝑡0 (1) where l0 is the ligament length (unfractured area ahead of the crack tip), t0 is the specimen thickness and β is a shape factor that depends on the shape of the plastic zone. Wf is obtained by testing a Double Edge Notched (DENT) specimen (Figure 1) at a constant displacement rate and integrating the area under the load vs displacement curve. The specific work of fracture (wf) is obtained by dividing Wf by the initial ligament area l0t0. Thus, equation (1) can be rewritten as: 𝑊𝑊𝑓𝑓 𝑙𝑙0𝑡𝑡0 =𝑤𝑤𝑓𝑓=𝑤𝑤𝑒𝑒+𝑤𝑤𝑝𝑝𝛽𝛽𝑙𝑙0 (2) If DENT specimens with different ligament lengths are tested and wf is plotted against the ligament length l0, a straight line with a positive intercept, which is the specific essential work of fracture (we), is obtained (Figure 1). International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 3 Figure 1. DENT specimen and experimental determination of the EWF: Wf for different ligament lengths and plot of wf against l0, the intercept indicates the specific essential work of fracture, we [5]. The obtained toughness value, we, quantifies the energy dissipated within the fracture process zone during the ductile tearing process and it is a suitable parameter to describe the crack propagation resistance of thin ductile sheets [1, 2, 5, 7]. we contains energetic contributions from both crack initiation and propagation since is an average value obtained from a linear regression of wf values for the complete separation. However, as shown by Mai and Cotterell [13], the EWF methodology also permits to separate both contributions and determine a cracking initiation toughness value, the specific work for fracture initiation. wei. The specific work for fracture initiation, wfi is calculated by integrating the area under load vs displacement curve until the onset of crack propagation (Figure 2, left). As observed in Figure 2 right, wfi is independent of the ligament length. Therefore, wei is calculated from the average of wfi values. Figure 2. Left: Determination of specific work of fracture at initiation of propagation (wf i). Right: Variation of wfi in function of ligament length and determination of the specific essential work of fracture at cracking initiation, wei [5]. For the evaluation of the EWF, rectangular DENT specimens of 240 x 55 mm (machined at 90º respect to the rolling direction) with ligament lengths ranging from 6 to 16 mm were tested up to fracture at a constant speed of 1 mm/min. In order to obtain toughness values independent of the notch radius fatigue pre-cracks were nucleated on the notch root (notch radius, ρ≈0,1 µm). More detailed information about the specimen geometry and test conditions is given in [5]. 2.2. Khan Tear Tests Khan Tear Tests (KTT) were originally developed to characterize the notch resistance of thin aluminum sheets [8]. The experimental procedure for KTT is described in ASTM B871 [14]. It consists in pulling at constant speed a single edge notched tensile (SENT) specimen with no prior fatigue pre-crack but a sharp notch. In this case, the notch radius was of 150 µm, obtained by electrical discharge machining (EDM). The specimens were machined at transverse direction and KTTs were conducted at a constant displacement rate of 1 mm/min. An initial gauge length of 10 mm was used for load-line displacement International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 4 measurement. The specimen geometry and the characteristic load vs displacement curve are shown in Figure 3. The notch resistance is characterized by the unit initiation energy (UIE) and the unit propagation energy (UPE). UIE represents the notch resistance to nucleate a crack and is calculated from the area under the load-displacement curve at maximum load. UPE is the primary result of the tear test and it is calculated from the area after the maximum load. It provides a measure of the combination of strength and ductility that permits a material to resist crack growth and it has significance as a relative index of fracture toughness. As indicated in the standard ASTM B871 [14], the method does not provide an absolute measure of the material resistance against crack propagation but a comparative measure of resistance to unstable fracture in the presence of crack-like stress concentrators. Figure 3. SENT specimen for tear tests (left). Load–displacement curve for a Kahn Tear Test (right). The UIE is calculated from the area under the curve before maximum load, and the UPE after maximum load. 2.3. Comparison between EWF and KTT Figure 4 compares fracture toughness results obtained by means of the EWF methodology and KTT. On the one hand, it is interesting to note that, even though the difference in notch radius between the two test configurations, there is a good correspondence between the values of crack initiation resistance, wei and UIE. It means that for the SENT specimen the machined sharp notch (ρ= 150 µm) closely represents the stress singularity of a crack. It supposes an advantage respect to the EWF tests since, it avoids the propagation of fatigue pre-cracks in the notch root, which is the most time consuming part in fracture mechanical characterizations. However, whereas results for crack initiation resistance are very similar, large differences are observed in the values associated to the crack propagation resistance, we and UPE. For example, DP1000 and TBF show UPE values (464 and 493 kJ/m2 respectively) comparable to PHS1000 (494 kJ/m2), which contrasts with the large differences observed in we: DP1000 and TBF shows the lowest we (138 ± 20 and 149 ± 13 kJ/m2), whereas PHS1000 exhibits a much greater we value (330 ± 21 kJ/m2). It is also observed that, according to UPE values, TBF/Q&P presents the greatest crack propagation resistance (755 kJ/m2). On the contrary, CP1000 shows the higher we (405 ± 11 kJ/m2). These differences can be associated to both the effect of the specimen geometry during crack propagation in KTT and to the conceptual dissimilarities between the two methodologies. It must be noted that after crack initiation, the load rapidly evolves from uniaxial tensile to bending and, therefore, crack propagates under a complex mixed loading mode. It give rise to UPE values that cannot be directly compared with pure Mode I fracture resistance, as given by we. Moreover, the energy calculated in the UPE is not only an energy for new fracture surface creation but it also contains the energetic contribution of the dissipated plastic work, which depends on the specimen geometry. In this regard, the EWF methodology separates both contributions and the toughness value we only quantifies the work spent in the fracture process zone to create new surfaces at the crack tip. Therefore, we better represents the steady state crack propagation resistance and it can be considered a material property, equivalent to the elastic plastic fracture International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 5 mechanics toughness value JC [13]. On the contrary, as mentioned before, UPE can be only used as a comparative value for crack propagation resistance for a given material. It is not a material property and it should be only used to rank materials crack propagation resistance. However, as observed in Figure 4, UPE can significantly overestimate such property. Figure 4. Results of EWF and KTT for different AHSS grades. EWF results taken from reference [5]. 3. Local formability and crash behaviour 3.1. Stretch flangeability The results of Hole Expansion Tests (HET) according to ISO 16630 for the investigated AHSS grades are summarized in Table 1. Initial punched holes of 10 mm in diameter were used for the expansion tests (punch to die clearance of 12 ± 2 %). Hole Expansion Ratio (HER) values are taken from reference [1]. 3.2. Bendability 3-point V-bending tests according to VDA 238-100 were performed at voestalpine Stahl following the procedure described in the work of Suppan et al. [15]. Specimens were bent with a sharp punch (r=0.4 mm) at a speed of 20 mm/min with the bending line lying parallel to rolling direction (bending strain in transverse direction). Free rotating rollers with a radius R=15 mm were used as shoulders and were separated according to: 𝑑𝑑= 2𝑡𝑡+ 0.5 (3) where d is the free space between the rolls and t is the specimen thickness. All values are in mm. The punch force and displacement was recorded and the test was stopped when the maximum punch stroke at around 160° bending angle was reached. Bending angle was indirectly calculated from the punch displacement as indicated in [16]. In the present work, the bendability was characterized by means of the critical angle (αCrit), defined as the angle at which the first visible crack was detected. In most cases, the critical angle coincided with the bending angle at maximum force (αCrit = αFmax), except for CP1000, where first visible cracks were detected up to 30° after the maximum load. Values of αCrit are reported in Table 1. 3.3. Local strain at fracture from uniaxial tensile tests assisted by digital image correlation Uniaxial tensile tests according to EN-ISO 6892 were performed in transverse direction. A Digital Image Correlation (DIC) equipment was used to monitor and determine the strain during the whole test. The International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 6 DIC equipment permits to measure local strains within the necking area. The local strain level after necking is much greater than the obtained by conventional extensometry with much larger 80 mm gage length and it better defines the local ductility potential of the material. Images were recorded at a frame rate of 10 images/second. A facet size and a step size of 11 and 9 pixels respectively were used. The local strain at fracture (Local εf) was determined from the point of maximum deformation (major logarithmic strain) at the stage before fracture (Figure 5a). Local εf values are summarized in Table 1. 3.4. Maximum energy absorbed in axial impact tests Crash resistance of AHSS is usually evaluated according to the energy absorbed, deformation, cracking and global appearance of the specimens after crash testing. In this work, the maximum energy absorbed in axial crash tests was used to characterize the impact resistance of the investigated steel grades. The energy absorbed during crash loading was calculated by integrating the area under the force vs impactor displacement (Figure 5b). To avoid the influence of the specimen thickness, the energy values from [5] were normalized by the cross-section area of the crashed sample. The values of impact energy absorbed per unit area are shown in Table 1. Details about crash specimen geometry and the experimental procedure followed for crash characterization can be found in reference [5]. a) b) Figure 5. a) Determination of local strain at fracture (Local εf) from uniaxial tensile tests with DIC. b) Force vs impactor displacement curves obtained from axial impact tests [5]. Table 1. Local formability measurements and crash behaviour for the investigated AHSS grades. Standard deviation is indicated when available. HER values and energy absorbed in axial impact tests are extracted from references [1] and [5] respectively. Mechanical properties for the transverse direction and sheet thickness are also given. Thickness, t [mm] HER [%] αCrit [º] Local εf [log.] Axial impact energy per unit area [kJ/m 2 ] Yield stress, σys [MPa] Ultimate tensile strength, σUTS [MPa] Elongation at fracture, A80 [%] CP1200 1.6 45 ± 10 77 0.48 14229 1041 1218 6.0 PHS1500 1.5 28 ± 2 55 0.42 7461 1075 1552 5.2 DP1000 1.4 35 ± 8 62 0.45 9592 738 1027 10.3 TBF 1.5 30 ± 1 80 0.45 12186 725 1019 14.7 PHS1000 1.5 57 ± 1 90 0.48 23626 988 1007 7.3 Q&P 1.4 55 ± 8 71 0.52 14122 909 1209 7.4 CP1000 1.4 85 ± 4 120 0.57 36504 908 1002 8.1 TBF/Q&P 1.4 66 ± 11 95 0.57 25047 876 1026 11.3 4. Fracture toughness vs local ductility Figure 6 plots fracture resistance results (we and UPE) against local formability and crash resistance parameters. It is observed that we shows a good correlation with all the different local ductility measurements, especially with ISO 16630 HER and maximum energy absorbed in axial impact tests International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 7 (R2=0.86 and 0.95 respectively), as previously reported and discussed by Casellas et al. [1] and Frómeta et al. [5]. It is also found a quite good correlation with the critical bending angle, αCrit (R2=0.83) and the local strain at fracture from uniaxial tensile tests (R2=0.60). Such results confirm the straight relationship between fracture toughness and the local ductility of AHSS and pose the essential work of fracture as a suitable material property to predict local formability and crash resistance. On the other hand, crack propagation resistance results from KTT (UPE), overall, show a low prediction ability for local formability and crash performance assessment. UPE shows a poor correlation with HER (R2=0.44), bending angle (R2=0.37) and impact energy per unit area (R2=0.42). Such correlation is improved for local strain at fracture (R2=0.64), which is comparable to the observed with we. Therefore, it is shown that, even though the UPE can give an estimation of the crack propagation resistance of the material, it is not a reliable parameter to predict cracking phenomena related to the material’s fracture toughness. a) b) c) d) Figure 6. Fracture toughness results (we and UPE) against different local ductility parameters: a) HER according to ISO 16630 [1]. b) Critical bending angle (αCrit) from V-bending tests. c) local strain at fracture (Local εf) from uniaxial tensile tests with DIC. d) Maximum energy absorbed in axial impact tests per unit area. 5. Summary and conclusions The experimental investigations carried out in this work allow pointing out fracture toughness, in terms of essential work of fracture, as a suitable material property to estimate the local ductility of AHSS sheets. This conclusion is based on the good correlation observed between we and the different local ductility parameters for a wide range of AHSS grades. This work compared the ability of fracture resistance values from the EWF methodology and KTT to understand crack-related problems in AHSS, as local formability or crashworthiness. Even though crack initiation values from KTT are quite reliable and very similar to the obtained by means of the EWF methodology, the crack propagation resistance in such tests is strongly influenced by the changing load mode during the test. UPE values show a poor correlation with the evaluated local ductility and International Deep Drawing Research Group 38th Annual Conference IOP Conf. Series: Materials Science and Engineering 651 (2019) 012071 IOP Publishing doi:10.1088/1757-899X/651/1/012071 8 crash resistance parameters. The good prediction capability of we compared with UPE can be understood considering their intrinsic differences, i.e. we accounts for the dissipated energy to create new surfaces in Mode I during crack propagation whereas UPE contains the contribution from plastic work during crack propagation in a mixed loading mode. Local ductility and crashworthiness are more related to crack propagation than to first crack nucleation, which explains their good correspondence with essential work of fracture values, we. Acknowledgments The authors would like to thank Dr. Clemens Suppan, Dr. Johannes Rehrl and Dr. Patrick Larour from voestalpine Stahl GmbH for providing the results of V-bending tests and for the support in the analysis of results and useful discussions. References [1] Casellas D, Lara A, Frómeta D, Gutiérrez D, Molas S, Pérez Ll , Rehrl J and Suppan C 2017 Fracture Toughness to Understand Stretch-Flangeability and Edge Cracking Resistance in AHSS Metall and Mat Trans A 48 86-94. [2] Frómeta D, Tedesco M, Calvo J, Lara A, Molas S and Casellas D 2017 Assessing edge cracking resistance in AHSS automotive parts by the Essential Work of Fracture methodology J Phys: Conf Ser 2017 896 012102. [3] Yoon J I, Jung J , Joo S H, Song T J, Chin K G, Seo M H, Kim S J, Lee S and Kim H S 2016 Correlation between fracture toughness and stretch-flangeability of advanced high strength steels Matter.Lett. 180 322-326 [4] Takahashi Y, Kawano O, Ushioda K and Aihara S 2012 Fracture Mechanical Study on Stretch Flange-Ability of Hot-Rolled High Tensile Strength Steel Sheets Proceedings of Asia Steel International Conference 2012 [5] Frómeta D, Lara A, Molas S, Casellas D, Rehrl J, Suppan C, Larour P and Calvo J 2019 On the correlation between fracture toughness and crash resistance of advanced high strength steels Eng. Frac. Mech. 205 319-332. [6] ASTM E1820. Standard Test Method for measurement of fracture toughness. [7] Cotterell B and Reddel JK 1977 The essential work of plane stress ductile fracture Int. J. Fracture 267-277. [8] Kaufman JG and Knoll AH 1964 Kahn-type tear tests and crack toughness of aluminium alloy sheets Mater. Res. Std. 4 151. [9] Garret GG and Knott JF 1978 The influence of compositional and microstructural variations on the mechanism of static fracture in aluminum alloys. Metal. Trans. A 9 1187-1201. [10] Dumont D, Deschamps A and Brechet Y 2003 On the relationship between microstructure, strength and toughness in AA7050 aluminum alloy. Mat. Sci. and Eng. A 356 326-336. [11] Ying L, Lu J, Chang Y, Tang X, Hu P and Zhao K 2013 Optimization evaluation test of strength and toughness parameters for hot-stamped high strength steels J. of Iron and Steel Research Int. 20 51 [12] Lorthios J, Gourgues A, Cugy P, Scott CP 2009 Damage of TWIP steels for automotive application, In ICF12 Int. Conf. Fracture Ottawa [13] Mai YW and Cotterell B 1986 On the essential work of ductile fracture in polymers Int. J. Fract. 32 105-125. [14] ASTM B871. Standard Test Method for Tear Testing of Aluminum Alloy Products. [15] Suppan C, Hebesberger T, Pichler A, Rehrl J and Kolednik O 2018 On the microstructure control of the bendability of advanced high strength steels Mat. Sci.and Eng. A 735 89–98 [16] Larour P, Hackl B, Leomann F and Benedyk K 2012 Bending angle calculation in the instrumented three-point bending test Proceedings IDDRG 2012, Mumbai, India. International Deep-Drawing Research Group (IDDRG 2020) IOP Conf. Series: Materials Science and Engineering 967 (2020) 012088 IOP Publishing doi:10.1088/1757-899X/967/1/012088 5 Figure 4. Images of the experimental setup for the notching process. a) Setup of the tool in the testing machine. b) Detail of the cutting tool. c) Specimen before (left) and after (right) the notching process. Figure 5. Schematization of the experimental procedure for the preparation of sheared notches in sheet specimens. 2.4. EWF tests After notch preparation, the DENT specimens were tested up to fracture according to the European Structural Integrity Society (ESIS) protocol for EWF testing [10]. The tests were conducted at a constant displacement rate of 1 mm/min. The load-line displacement was measured by means of the video extensometer using initial extensometer marks separated 25 mm. 3. Results and discussion 3.1. EWF results Figure 6 shows the load vs load-line displacement curves for the two specimen configurations. Figure 7 plots the values of wf as a function of the ligament length. Numerical values of we and βwp are given in Table 2 and Figure 8. As observed, for the same ligament length, both fatigue pre-cracked and sheared specimens show very similar load vs displacement curves (similar maximum load and displacement at fracture) in the four investigated materials. It explains the good agreement between wf values for the two notch conditions (Figure 7) and the practically identical specific essential work of fracture, we and plastic work, βwp (Figure 8). In general, very good repeatability is observed for sheared specimens, which International Deep-Drawing Research Group (IDDRG 2020) IOP Conf. Series: Materials Science and Engineering 967 (2020) 012088 IOP Publishing doi:10.1088/1757-899X/967/1/012088 6 enhance the reliability of the obtained toughness values and confirms the robustness of the new process. It is worth noting that the similarity between specimens of the same ligament length is improved in sheared specimens. This is because the notch length is precisely defined by the punch displacement and, therefore, is easier to obtain multiple specimens with the same ligament. On the other hand, the ligament size in fatigue pre-cracked specimens is determined by the length of the fatigue cracks, which makes difficult to have two specimens with identical ligament. a) b) c) International Deep-Drawing Research Group (IDDRG 2020) IOP Conf. Series: Materials Science and Engineering 967 (2020) 012088 IOP Publishing doi:10.1088/1757-899X/967/1/012088 7 d) Figure 6. Load-displacement curves obtained from EWF tests with fatigue pre-cracked specimens (left) and specimens with sheared notches (right). a) CP, b) DP-A, c) DP-B and d) 3rd Gen AHSS. a) b) c) d) Figure 7. wf values against ligament length for the two investigated specimen configurations. a) CP, b) DP-A, c) DP-B and d) 3rd Gen TRIP-assisted. International Deep-Drawing Research Group (IDDRG 2020) IOP Conf. Series: Materials Science and Engineering 967 (2020) 012088 IOP Publishing doi:10.1088/1757-899X/967/1/012088 8 Table 2. EWF results obtained with fatigue pre-cracked specimens and specimens with sheared notches Fatigue pre-cracked Sheared notches Material we [kJ/m2]  wp [MJ/m3] we [kJ/m2]  wp [MJ/m3] CP 405 ± 11 12 ± 1 404 ± 19 12 ± 1 DP-A 149 ± 21 24 ± 2 163 ± 27 21 ± 2 DP-B 286 ± 17 23 ± 1 298 ± 32 21 ± 2 3rd GEN 104 ± 30 34 ± 3 113 ± 21 30 ± 2 Figure 8. Results from EWF tests with fatigue pre-cracked (blue) and sheared (orange) specimens. Left: specific essential work of fracture, we. Right: non-essential plastic work, βwp. 3.2. Fracture surface of DENT specimens Figure 9 shows the fracture surfaces of different sheared and fatigue pre-cracked DENT specimens. It can be observed that for both specimen configurations the fracture aspect is quite similar and the ligament is well defined between the two notches. The major difference between the two notch types is the shape of the crack front (concave for the sheared notch and convex for the fatigue pre-crack). Overall, it was found that the morphology of the sheared notches is similar in the four investigated AHSS grades and independent of the ligament length. International Deep-Drawing Research Group (IDDRG 2020) IOP Conf. Series: Materials Science and Engineering 967 (2020) 012088 IOP Publishing doi:10.1088/1757-899X/967/1/012088 9 a) b) c) d) Figure 9. Fracture surfaces of DENT specimens with sheared notches (left) and fatigue pre-cracks (right). a) CP, b) DP-A, c) DP-B and d) 3rd GEN. The different areas (fractured ligament, sheared notch and fatigue pre-crack) are indicated in the first images. 4. Applications As mentioned before, fracture toughness has shown to be a useful material property to predict crackrelated problems in high strength metal sheets, such as edge fractures [1-3], crack formation during crash [4] or other fractures related to their local ductility [5]. It is illustrated in Figure 10, where we values of several AHSS grades are plotted against HER [2-3] and the maximum energy absorbed in axial crash tests [4]. As already discussed in [2-5], we shows a very good correlation with HER and crash resistance. International Deep-Drawing Research Group (IDDRG 2020) IOP Conf. Series: Materials Science and Engineering 967 (2020) 012088 IOP Publishing doi:10.1088/1757-899X/967/1/012088 10 Therefore, we can be used to estimate the fracture resistance of AHSSs and predict their cracking behaviour during forming or crash. According to the linear data fittings depicted in Figure 10, expected values of HER and impact energy for the steel grades investigated in this work are plotted as a function of the obtained specific essential work of fracture values (results from EWF tests with sheared specimens). In the case of DP-A and 3rd GEN steel, the HER was experimentally evaluated according to ISO16630. As observed, the measured HER fits very well in the linear we vs HER data fitting, which validates the suitability of we for edge cracking resistance prediction. Therefore, the new testing procedure presented, can be used as a fast and cost-effective tool to readily assess the fracture performance of AHSS sheets. Figure 10. Correlation of we with HER (left) and axial impact energy (right). The data represented by black squares is extracted from references [2-4]. Blue symbols correspond to we values obtained in this work. Solid symbols represent experimentally evaluated values of HER and axial impact energy. Open symbols are expected values according to the linear data fitting (dashed line). 5. Conclusions In the present work, an innovative device to prepare high strength metal sheet specimens for fracture toughness characterization has been presented. The tool can be easily mounted in a universal testing machine and offers an easy and cheap alternative to fatigue pre-cracking procedures. The new process has shown to be robust and reliable enough to evaluate the fracture toughness of four different AHSS sheets. It supposes a great time-saving in specimen preparation and it can be very useful to boost the use of fracture toughness measurements as routine testing for coil quality determination or for the selection of high strength sheet materials with enhanced cracking resistance. 6. References [1] Yoon J I, Jung J, Joo S H, Song T J, Chin K G, Seo M H, Kim S J, Lee S and Kim H S Correlation between fracture toughness and stretch-flangeability of advanced high strength steels Matter.Lett. 180 (2016) 322-326. [2] Casellas D, Lara A, Frómeta D, Gutiérrez D, Molas S, Pérez Ll, Rehrl J and Suppan C. Fracture Toughness to Understand Stretch-Flangeability and Edge Cracking Resistance in AHSS. Metall. and Mat. Trans. A 48 (2017) 86-94. [3] Frómeta D, Lara A, Parareda S and Casellas D. Evaluation of Edge Formability in High Strength Sheets Through a Fracture Mechanics Approach. AIP Conference Proceedings 2113, 160007 (2019). [4] Frómeta D, Lara A, Molas S, Casellas D, Rehrl J, Suppan C, Larour P and Calvo J. On the correlation between fracture toughness and crash resistance of advanced high strength steels. Eng. International Deep-Drawing Research Group (IDDRG 2020) IOP Conf. Series: Materials Science and Engineering 967 (2020) 012088 IOP Publishing doi:10.1088/1757-899X/967/1/012088 11 Frac. Mech. 205 (2019) 319-332. [5] Frómeta D, Lara A, Casas B, and Casellas D. Fracture toughness measurements to understand local ductility of advanced high strength steels. IOP Conf. Ser.: Mater. Sci. Eng. (2019) 651 012071 [6] ASTM E1820. Standard test method for measurement of fracture toughness. American Society for Testing and Materials [7] ASTM E399. Standard test method for plane-strain fracture toughness of metallic materials. American Society for Testing and Materials [8] Cotterell B and Reddel JK. The essential work of plane stress ductile fracture. Int. J. Fract. (1977) 267-277. [9] Muñoz R, Lara A and Casellas D. Fracture toughness characterization of advanced high strength steels. Int. Deep Drawing Research Group (IDDRG) Conference 2011 (Bilbao, Spain, June 5-8, 2011) [10] Clutton E. Essential work of fracture. Moore DR, Pavan A, Williams JG, editors. Fracture mechanics testing methods for polymers, adhesives and composites, vol.28. ESIS Publ.; (2001) 177–95. Acknowledgements The present research work has been partially funded by the EU Horizon 2020 programme under grant agreement H2020-EU.2.1.3. – 814517 (FormPlanet).