Fatigue Properties of Ti-6Al-4V Processed by SEBM

Transcription

1 Institut für Werkstofftechnik Metallische Werkstoffe Fatigue Properties of Ti-6Al-4V Processed by SEBM T. Niendorf, J. Günther, D. Krewerth, A. Weidner and H. Biermann EBAM 2016 Nuremberg April 2016 EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 1

2 Introduction Additive Manufacturing Assessment criteria: Cost Light weight potential Mechanical properties Fatigue Corrosion Source: DMRC EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 2

3 Introduction Additive Manufacturing Assessment criteria: Cost Light weight potential Mechanical properties Fatigue Corrosion Requisite Correlation of Process Microstructure Properties Damage evolution Source: DMRC EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 3

4 Experimental details Source: SLM Solutions a) b) Source: ARCAM SLM 250 HL (SLM Solutions GmbH, Lübeck, Germany) Yttrium laser, 400 W Layer thickness 30 µm Argon atmosphere ARCAM A2X (Arcam AB, Mölndal, Sweden) Electron Beam, 3000 W Layer thickness 50 µm Vacuum atmosphere EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 4

5 Introduction Ti6Al4V features high specific strength, good corrosion resistance and biocompatibility Additive manufactured (AM) Ti6Al4V is employed in medical and aerospace industries Arcam AB 25 µm Investigation of the fatigue behavior of additive manufactured Ti6Al4V in the high cycle fatigue (HCF) and the very high cycle fatigue (VHCF) regime Comparison of two processing routes (SLM vs. EBM) Influence of the crack initiating defects (e.g. porosity, lack of fusion, α-platelets) Changing failure modes in the VHCF internal fatigue crack initiation EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 5

6 Selective laser melting 4 TiAl6V4 manufactured on a SLM 250 HL 400 W fibre laser layer thickness of 30 μm mean particle size of 40 μm platform temperature of 200 C 4 different conditions condition temperature / pressure (1) as-built -/ - (2) stress relieving 800 C / - (3) HIP 920 C / 1000 bar (4) modified microstructure 1050 C / - EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 6

7 Selective laser melting 4 microstructure (1) (2) EBSD phase maps for Ti-6-4 following as-built (1), 800 C (2), HIP (3), 1050 C (4) (3) (4) (1) & (2) elongated grains nearly same grain size and morphology (3) intermediate state (4) large grains of α Ti β Ti at grain boundaries Building direction EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 7

8 Selective laser melting 4 residual stress Pronounced stresses in build direction Highest absolute values at 100 µm below surface Significant reduction by heat treatment Similar trends in stainless steel 316L EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 8

9 Selective laser melting 4 monotonic properties condition UTS [MPa] σ γ [Mpa] ε f [%] (1) as-built (2) 800 C (3) HIP (4) 1050 C samples per condition max. deviation of ±30 MPa for UTS and ±2 % for ε f highly brittle behaviour for the as-built condition higher temperature is associated with higher ductility EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 9

10 Selective laser melting 1 Low impact of heat treatments Residual stresses Pronounced impact of HIP Porosity EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 10

11 Monotonic properties 4 Residual stresses SLM EBM Specimen Position σ ES [MPa] in x- direction σ ES [MPa] in y- direction Specimen Position σ ES [MPa] in x- direction σ ES [MPa] in y- direction As-built/Point 1 Surface As-built/Point 2 Surface As-built/Point 3 0.1mm C/Point 1 Surface C/Point 2 Surface +5-5 As-built/Point 1 Surface -15 ± 5-30 ± 10 As-built/Point 2 0.1mm -10 ± 5-10 ± C/Point 1 Surface -20 ± 5-35 ± C/Point 2 0.1mm -10 ± 5-5 ± 5 EBM - built at 700 C Almost no residual stresses SLM - built at 200 C Extremely high residual stresses Difference between melting point and pre-heat temperature EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 11

12 Fatigue properties Additive Manufacturing Technologies Powder Bed Selective Laser Melting (SLM) Build plate heated to 200 C Electron Beam Melting (EBM) Build chamber temperature > 700 C Batch 1 Annealing 800 C, 2 h, Ar Batch 2 Hot-isostatic pressing (HIP) 920 C, 2 h, 1000 bar, Ar Batch 3 As-built condition During EBM: Building chamber heated to 700 C residual stresses significantly lower 1 EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 12

13 Fatigue properties 5 Specimens were machined from cylinders built by SLM and EBM Fatigue specimens were ground and polished to minimize surface roughness a) Raw cylinders b) Specimen geometry USFT/VHCF c) Specimen geometry HCF 1 z y x EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 13 13

14 Fatigue properties 5 Microstructure analysis by Electron Backscatter Diffraction (EBSD) SLM (annealed) SLM (HIPed) EBM EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 14

15 Fatigue properties 5 Ultrasonic fatigue testing machine (Type BOKU Vienna, Austria) Testing frequency f = 19,5 khz, load ratio R = -1 Pulse/ pause testing mode and compressive air cooling (b) Temperature monitoring by infrared camera (c) b) c) a) EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 15 15

16 Fatigue properties 5 Fracture surface analysis for every specimen after fatigue failure using SEM SLM Determination of the type, size, position and morphology of the failure relevant defect EBM Lack of fusion Circular pores a-phase EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 16

17 Fatigue properties 5 Surface porosity/ lack of fusion b) 100 µm Internal porosity Results obtained at varying testing frequencies: Ultrasonic testing = low frequency testing Key aspect location of initiating defect c) 200 µm EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 17

18 Fatigue properties 5 Porosity b) 25 µm Lack of fusion a) Results only obtained by ultrasonics: EBM as built = SLM heat treated c) 25 µm EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 18

19 Fatigue properties 5 Non-HIP HIP Typical appearance of fracture surfaces EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 19

20 Fatigue properties 5 Surface α-platelet b) 25 µm Internal α-platelet a) Results only obtained by ultrasonics: EBM as built = SLM heat treated Superior properties upon HIP c) 50 µm EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 20

21 Fatigue properties 5 Conventionally processed material 2 Ti6Al4V upon AM + HIP almost able to meet properties of conventionally processed alloy EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 21

22 Fatigue properties 5 Determination of average defect sizes and stress intensity factors Weibull propability plot a-phase as large as pores DK I,max relatively high for a-phase still, highest fatigue lives EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 22

23 Fatigue properties 5 Murakami approach can be employed for determination of fatigue properties EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 23

24 Fatigue properties Two failure mechanisms in both conditions (non-hiped vs. HIPed condition) Data evaluated by use of two separate S-N curves Surface fatigue crack initiation vs. internal fatigue crack initiation Surface fatigue crack initiation causes premature failure Internal fatigue crack initiation is observed only in the VHCF regime (>2x10 6 ) EBM as-built condition is similar to the annealed SLM condition in terms of fatigue properties Defect size and location determine fatigue properties Subsequent HIP processing improves fatigue properties Application of Murakamis area model is possible Impact of microstructure will be addressed in future work EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 24

25 Introduction Additive Manufacturing New Materials Prospects Graded Structures Microstructure Geometry Composition Wear resistance Design of chemical potential EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 25

26 Introduction Additive Manufacturing New Materials Prospects Graded Structures Microstructure Geometry Composition Wear resistance Design of chemical potential There are (almost) no limits! Think additive! EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 26

27 Introduction Additive Manufacturing New Materials Prospects Graded Structures Microstructure Geometry Composition Wear resistance Design of chemical potential Requisite Correlation of Process Microstructure Properties Damage evolution EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 27

28 New materials 200 µm Ti-6Al-4V Co-Cr-Alloy Ti EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 28

29 Hardness, HV 30 New materials Monotonic/Hardness + >20 % EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 29

30 Hardness, HV 30 New materials Monotonic/Hardness Fatigue + >20 % EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 30

31 New materials New materials can be processed by SLM and EBM Hardness of Ti-6Al-4V can be increased by adding small amounts of Co-Cr BUT: Fatigue properties are deteriorated Optimization with respect to loading conditions! EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 31

32 Thank you for your attention! Financial support by DFG is gratefully acknowledged References: 1 Leuders et al., J. Mater. Res., Vol. 29, No. 17, Heinz et al., Ultrasonics, Vol. 53, Murakami, Y., Metal Fatigue: Effect of small defects and nonmetallic inclusions, Elsevier Ltd., Leuders et al., IJF, Vol. 48, Günther et al., IJF, accepted for publication, 2016 EBAM 2016 Prof. Dr.-Ing. Thomas Niendorf 32