ESIS TC 24 Workshop Politecnico di Milano, 30 March 2012

Transcription

1 ESIS TC 24 Workshop Politecnico di Milano, 30 March 2012 Effects of different heat treatment and resulting microstructure on resistance of EA4T railway axle steel of equal strength to fatigue crack growth Ivo Černý SVÚM a.s., Podnikatelská 565, Praha 9, Czech Republic Contacts:

2 CONTENT Introduction general Introduction to the paper Experiments Results and discussion Conclusions 2

3 Introduction Railway axles are commonly operated in a very high cycle fatigue regime with number of loading cycles in the order of 10 9 Safe operation is based on a combination of safe life design and regular inspections in combination with damage tolerance principles. For the application of damage tolerance principles, material properties, particularly the da/dn-dk curve describing fatigue crack propagation in combination within fracture toughness are essential data. Comprehensive data bases already exist. They, however, contain a lot of experimental points with quite a large scatter. Therefore, upper or lower bounds are usually determined for design purposes 3

4 Introduction The aim of this contribution is to review some recent data on resistance of an EA4T steel to fatigue crack growth including the threshold region and compare some further recently evaluated data from the point of view of material heat treatment, microstructure effects of type of specimens and crack closure phenomenon. Some recent results and knowledge gained during solution of research grant of the Czech Ministry of Education, Youth and Sports MSM Investigation of damage mechanisms and quantification of effects of defects on service life of heavy loaded components, chapter devoted to railway axle damage issues, will be reported and discussed. The grant has been solved at SVÚM a.s., Praha. 4

5 Introduction - general Within the mentioned grant MSM , quite a big effort has been devoted to improvement of experimental method of measurement crack growth length and resistance to fatigue crack growth (FCG), development and verification of a method for measurement of growth of physically short cracks, evaluation of FCG characteristics of different structural materials including railway axle steels including damage mechanisms. 5

6 Introduction - general Concerning the experimental methods of measurement crack growth length and resistance to FCG of long cracks, the accredited laboratory of the Strength Department of SVÚM a.s. participated in the Proficiency Testing Programme (PTP), i.e. Round Robin Tests, organised by Exova in a collaboration with General Electric Aviation company. Excellent results of the SVUM laboratory in the PTP programme provide a solid basis for exact measurement of FCG properties in general, without questions about correctness of the measurement. 6

7 Introduction - general - PTP FCG programme 7

8 Introduction - general - PTP FCG programme 8

9 Introduction - general - PTP FCG programme 9

10 Difference between calculated and actual a (mm) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Introduction - general - PTP FCG programme 10 8 y = x R 2 = a calculated from DCPD measurement and analytical formula (mm) 10

11 Introduction - general - PTP FCG programme 11

12 Introduction - general - PTP FCG programme 12

13 Potential V(a) / Vref ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Introduction - general - method for short crack - sensitivity analysis W=15.72 mm, y=5 mm W=15.72 mm, y=3 mm W=15.72 mm, y=1 mm W=5 mm, y=5 mm W=5 mm, y=3 mm W=5 mm, y=1 mm Crack length (mm) 13

14 Potential increment V(a) / Vref Introduction - general - method for short crack - sensitivity analysis W=5 mm, y=5 mm W=5 mm, y=3 mm W=5 mm, y=1 mm Crack length (mm) 14

15 Introduction - general - method for short crack - sensitivity verification. 15

16 Potential V(a)/Vref ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Introduction - general - method for short crack - sensitivity verification Time from the start (s) 0.12 mm 0.07 mm 0.37 mm

17 Experiments FCG measurement performed on EA4T steel taken from two actual railway axles manufactured at Bonatrans a.s. Bohumín, Czech Republic. Two different heat treatment processes, namely quenching in air and in oil. Equal static strength in both the cases. SEN(B3) specimens (low value of T-stress), W = 30 mm, B = 7 mm, test span mm, load frequency 40 Hz, load asymmetry R = 0.1, load shedding method. 17

18 Experiments Fatigue machine SCHENCK PHG COD measured near crack (notch) mouth using semiconductor extensometer INOVA PXA, gauge length 20 mm 18

19 Experiments 19

20 Load (V) Load (V) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Experiments - COD measurement - dynamically 0.0 impossible -0.5 CMOD, zatěžování f = 7 Hz CMOD, f = 40 Hz INOVA PXA INOVA PXA 20

21 da / dn (m / cycle) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Results 1.E-06 1.E-07 EA4T, air hardened, SEN(B3) EA4T, quenched in oil, SEN(B3) Luke-Varfolomeev, M(T) 1.E-08 1.E-09 1.E D K (MPa m 1/2 ) 21

22 da / dn (m / cycle) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Results and discussion 1.E-06 EA4T, air hardened, SEN(B3) EA4T, quenched in oil, SEN(B3) 1 Luke-Varfolomeev, M(T) 1.E-07 1.E-08 Three points to be discussed: 1. Differences between FCG. curves 2. Agreement between FCG in. quenched in oil and Luke-...Varfolomeev in Paris region 3. Differences between the two. curves in threshold region E-09 1.E K (MPa m 1/2 ) 22

23 Results and discussion - point 1 Unexpected differences between the two FCG curves - K-factor corresponding to same da/dn values is approximately by 50% higher for the steel quenched in oil. HV30 measurement: average 213 HV in case of quenching in air and 209 HV for quenching in oil - very similar values, which correspond to strength of 660 MPa and 670 MPa, respectively. These values are near the lower strength limit for the EA4T steel. According to EN strength should be between MPa. Microstructure analysis can explain the differences. 23

24 Results and discussion - point 1 Quenching in air Quenching in oil 24

25 Results and discussion - point 1 Quenching in air Quenching in oil 25

26 da / dn (m / cycle) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Results and discussion - point 2 1.E-06 EA4T, air hardened, SEN(B3) EA4T, quenched in oil, SEN(B3) 1 Luke-Varfolomeev, M(T) 1.E-07 1.E-08 Three points to be discussed: 1. Differences between FCG. curves 2. Agreement between FCG in. quenched in oil and Luke-...Varfolomeev in Paris region 3. Differences between the two. curves in threshold region E-09 1.E K (MPa m 1/2 ) 26

27 Results and discussion - point 2 The agreement between the curves in the stable Paris region is a positive result considering the fact that measurements were performed on quite different types of specimens with different T-stress. It indicates that the reproducibility of FCG measurement may be sufficient and the dependence on specimen type need not be high and crucial. Data about static strength, parameters of heat treatment and material microstructure would be interesting. There are significant differences in the threshold region - Point 3 27

28 Results and discussion - point 3 - COD measurement and K eff analysis COD measurement and analysis of DK eff was carried out to explain the differences and behaviour in the near threshold region Three static measurement of COD carried out during the FCG test of the material quenched in oil: in threshold region (da/dn = 2.1 E-09 m/cycle) in stable region of medium FCG rate (da/dn = 1.85 E-08 m/cycle) in region of quite fast FCG (da/dn = 7.45 E-08 28

29 Load (N) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Results and discussion - point 3 - COD evaluation F cl =1100 N (loading) Unloading Loading 1000 F cl =900 N (unloading) COD (mm) 29

30 Fcl (N) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Results and discussion - point 3 - COD evaluation Average Unloading mean Loading mean load 1 unld. 1 mean 2 load 2 unld. 2 mean 3 load 3 unld. 3 mean Measurement

31 Results and discussion - point 3 - COD evaluation Measurement 1 Measurement 3 Measurement 2 31

32 da/dn (m/cycle) ESIS TC 24 Workshop, Politecnico di Milano, 30 March 2011 Results and discussion - point 3 - COD measurement and K eff analysis 1.E-06 1.E-07 Zerbst et al., to be published in EFM, Delta Keff A4T, quenched in oil, SEN(B3), Delta K A4T, quenched in oil, SEN(B3), Delta Keff EA4T, air hardened, SEN(B3) U = E-08 U = 0.90 U = DK eff / DK U = E-09 1.E D K (MPa m 1/2 ) 32

33 Conclusions Though strength the EA4T steel was equal, near the lower limit according to the EN Standard, FCG in the two modifications of the steel was very different, near the lower and upper limits of the known scatter band. FCG rate in the steel quenched in air was significantly higher than in that quenched in oil. Once more pointed out: Equal hardness, equal static strength, just minor differences between microstructure of the two modifications. Good agreement between measurement in M(T) and SEN(B3) specimens in Paris region. COD could explain differences in the near threshold region. 33

34 Acknowledgement The work was supported by the grant MSM of the Czech Ministry of Education, Youth and Sports. 34

35 Accredited full-scale fatigue tests at SVÚM a.s., Praha Total view of fairly old Sincotec resonance machine with controlling computer unit and tested axle 35

36 Accredited full-scale fatigue tests at SVÚM a.s., Praha New facilities being purchased, fully certified by Deutsche Bahn 36

37 Accredited full-scale fatigue tests at SVÚM a.s., Praha... in connection with the new SVÚM a.s. technology park near Praha - in Čelákovice 37

38 Accredited full-scale testing and research at SVÚM Since nothing has been finally decided by now and there is lack of financial sources, any preliminary, not binding expression of possible interest in a collaboration with SVÚM would be very welcome and helpful for the present-day position of Strength Department and future railway research and testing activities at SVÚM. Contact address: Ivo Černý, Project Manager SVÚM a.s. Podnikatelská Praha 9 Czech Republic Ivo.Cerny@seznam.cz, 38

39 Thank you for your kind attention 39