Effect of residual stresses on the fatigue lifetime of ... - Science Direct

Available online at www.sciencedirect.com Available online at www.sciencedirect.com

ScienceDirect ScienceDirect

Structural Integrity 00 (2017) 000–000 Available online at www.sciencedirect.com Available online atProcedia www.sciencedirect.com Structural Integrity Procedia 00 (2017) 000–000

ScienceDirect ScienceDirect

Procedia Structural Integrity 00 4 (2017) Structural Integrity Procedia (2016)42–47 000–000

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

www.elsevier.com/locate/procedia

ESIS TC24 Workshop "Integrity of Railway Structures", 24-25 October 2016, Leoben, Austria ESIS TC24 Workshop "Integrity of Railway Structures", 24-25 October 2016, Leoben, Austria

Effect of residual stresses on the fatigue lifetime of railway axle Effect of residual stresses on the fatigue lifetime of railway axle

XV Portuguese Conference on Fracture, PCF 2016,a 10-12 February 2016, Paço de Arcos, a a b a Portugal a

Pavel Hutařa, Pavel Pokornýa, Jan Poduškaa, Rostislav Fajkošb, Luboš Náhlíka* Pavel Hutař , Pavel Pokorný , Jan Poduška , Rostislav Fajkoš , Luboš Náhlík *

Thermo-mechanical modeling of a high pressure turbine blade of an airplane gas turbine engine

CEITEC IPM, Institute of Physics of Materials, Academy of Sciences of the Czech Republic, v. v. i., Žižkova 22, 616 62 Brno, Czech Republic b CEITEC IPM, Institute of Physics of Materials, Academy of Sciences the Czech Republic, v. v. i., Žižkova 22, 616 62 Brno, Czech Republic BONATRANS GROUP, a. s., Revolučníof1234, 735 94 Bohumín, Czech Republic b BONATRANS GROUP, a. s., Revoluční 1234, 735 94 Bohumín, Czech Republic

a

Abstract Abstract

P. Brandãoa, V. Infanteb, A.M. Deusc*

a

Department of Mechanical Engineering, Instituto Superior Universidade Lisboa,and Av. low Rovisco Pais, 1, costs. 1049-001 Lisboa, The operation of railway axles should fulfill at leastTécnico, two main demands:desafety operation A significant Portugal The operation of given railway should of fulfill at least two main demands: safety and low operation A significant part bof operation costs is byaxles the length regular inspection intervals which should reveal potentialcosts. fatigue cracks in IDMEC, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa, part of operation costs is given by the of regularnature, inspection intervals which should reveal potential fatigue cracks in railway axle. The detection of cracks is oflength a probabilistic therefore their detection is not ensured in all cases. For the safe Portugal c railway axle. The detection of cracks is of a probabilistic nature, therefore their detection is not ensured in all cases. For the safe operation of trains, an existence of potential initial crack Superior should be considered on thedeaxle surface and residual lifetime CeFEMA, Department of Mechanical Engineering, Instituto Técnico, Universidade Lisboa, Av. Rovisco Pais, 1,fatigue 1049-001 Lisboa, operation trains, an existence of potential should be considered on the axle lifetime surface and residualshould fatiguetake lifetime Portugal should be of conservatively determined for thisinitial case. crack Reliable procedure of residual fatigue estimation into should conservatively determined this case. Reliable procedure of residual lifetime estimation should take into accountbereal axle geometry, materialforcharacteristics and loading of the railway fatigue axle. This paper shows methodology for account real axle geometry, material andtheloading the railway axle. This paper shows methodology determination of residual fatigue lifetimecharacteristics (RFL) based on fractureofmechanics approach, taking into account real spectrumfor of determination of residual fatigue (RFL)wheels based on fracture mechanics approach, account real spectrum of theAbstract loading cycles, existence oflifetime press-fitted andthe surface residual stresses given taking by theinto thermo-mechanical surface the loading cycles, existence surface by the and thermo-mechanical surface treatment of the railway axle. Itofis press-fitted demonstratedwheels that theand effect of theresidual residualstresses stresses given is significant should not be neglected their modern aircraft to increasingly demanding operating conditions, treatment of the operation, railway axle. is demonstrated thatcomponents the effect of are the subjected residual stresses is significant and should not be neglected in During the numerical estimation of It residual fatigueengine lifetime of the axle. the high pressure turbine (HPT) Such conditions in especially the numerical estimation of residual fatigueblades. lifetime of the axle. cause these parts to undergo different types of time-dependent degradation, one of which is creep. A model using the finite element method (FEM) was developed, in order to be able to predict © 2017 The Authors. Published by Elsevier B.V. Copyright © 2017. The Authors. Published Elsevier B.V. records (FDR) for a specific aircraft, provided by a commercial aviation the behaviour of HPT blades.byFlight © 2017creep The Authors. Published by B.V.data Peer-review under responsibility of Elsevier the Scientific Committee of ESIS TC24. Peer-review of the Scientific Committee of ESIS company,under wereresponsibility used to obtain thermal and mechanical dataTC24. for three different flight cycles. In order to create the 3D model Peer-review under responsibility of the Scientific Committee of ESIS TC24. needed for the FEM analysis, a HPT blade scrap was scanned, and its chemical composition and material properties were Keywords: Railway axle;that residual lifetime; residual inspection intervals; numerical simulation obtained. The data was fatigue gathered was fed into stresses; the FEM model and different simulations were run, first with a simplified 3D Keywords: Railway axle; residual fatigue lifetime; residual stresses; inspection intervals; numerical simulation rectangular block shape, in order to better establish the model, and then with the real 3D mesh obtained from the blade scrap. The overall expected behaviour in terms of displacement was observed, in particular at the trailing edge of the blade. Therefore such a 1. model Introduction can be useful in the goal of predicting turbine blade life, given a set of FDR data.

1. Introduction © 2016Ones The Authors. Published by Elsevier of B.V. of the critical components trains are railway axles, see references Zerbst et al. (2005), Zerbst et al. Peer-review under responsibility ofIttheisScientific Committee ofaxles PCFaxles, 2016. Ones of the critical components of trains are railway see references et al. or (2005), Zerbst et al. (2013) and Zerbst et al. (2013b). known that railway can include defects Zerbst like cracks surface scratches, (2013) can and lead Zerbst al. (2013b). It is known that railway axles can include defects like cracks or surface scratches, which to etfatigue crack initiation and propagation. Unfortunately, the detection of such (in certain cases Keywords: Turbine Blade; Creep; Finite Method; 3D Model; Simulation. which can High lead Pressure to fatigue crack initiation andElement propagation. Unfortunately, the detection of such (in certain cases

* Corresponding author. Tel.: +420 532 290 358 * Corresponding Tel.: +420 532 290 358 E-mail address:author. [email protected] E-mail address: [email protected] 2452-3216 © 2017 The Authors. Published by Elsevier B.V. 2452-3216 © 2017 Authors. Published Elsevier B.V. Peer-review underThe responsibility of theby Scientific Committee of ESIS TC24. Peer-review underauthor. responsibility the Scientific Committee of ESIS TC24. * Corresponding Tel.: +351of218419991. E-mail address: [email protected] 2452-3216 © 2016 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the Scientific Committee of PCF 2016. 2452-3216 Copyright  2017. The Authors. Published by Elsevier B.V. Peer-review under responsibility of the Scientific Committee of ESIS TC24 10.1016/j.prostr.2017.07.005

2

Pavel Hutař et al. / Procedia Structural Integrity 4 (2017) 42–47 Author name / Structural Integrity Procedia 00 (2017) 000–000

43

dangerous) cracks is not possible in all cases. The longer crack, the higher probability of its detection exists, see Benyon and Watson (2001). However, the detection does not depend only on; the size of the crack, but also on the position of the crack and non-destructive testing (NDT) method used. Due to this fact it is necessary to know the length of crack which is detected with required probability. By assumption of initial crack length in the axle it is possible to setup frequency of regular inspections including NDT method, see references Zerbst et al. (2011) and Zerbst et al. (2013). If the inspection intervals are shorter than the time necessary for crack propagation from initial size up to critical one, the operation of the railway axle is safe. Nevertheless, if regular inspections are performed unnecessary often, the costs for axle (train) operation will increase. Nowadays, the regular inspection intervals are often verified by numerical simulations, see references Luke et al. (2010), Luke et al. (2011), Smith (2000), Zerbst et al. (2005) and Zerbst et al. (2013). However, it seems that in many cases numerically estimated residual fatigue lifetime (RFL) of railway axles is still different in comparison to the experimentally determined one. One important factor influencing fatigue crack propagation in the railway axle is level of residual stresses. Especially compressive residual stresses close to the surface of the axle can effectively retard fatigue crack propagation from initial surface defect and consequently extend the RFL of the axle, see e.g. Regazzi et al. (2014). The aim of this contribution is to quantify the effect of the residual stresses (based on numerical simulations) on RFL of the railway axle. Nomenclature a a0 b C, n, p k KB KI Kmax Kmax,th Kmin KPF KRS K L R v (da/dN) ax EA4T SIF RFL (B) (PF) (RS)

crack length initial crack length crack width material constants of NASGRO relationship dynamic coefficient stress intensity factor corresponding to bending load stress intensity factor (general expression) maximum of stress intensity factor in load cycle threshold value in Kmax expression minimum of stress intensity factor in load cycle stress intensity factor corresponding to press-fit load stress intensity factor corresponding to residual stress stress intensity factor range distance from axle surface (depth) stress ratio fatigue crack propagation rate residual axial stress steel grade stress intensity factor residual fatigue lifetime bending press-fit residual stress

2. Estimation of residual fatigue life of railway axle with consideration of residual stresses Presented numerical estimation represents enhanced version of already published procedure (see Náhlík et al. (2017)) for determination of RFL of railway axle. The difference is in implementation of influence of residual stresses, which are induced during manufacturing process, to the procedure of RFL estimation. Fig. 1 shows considered railway axle. In this case the critical initial crack location was determined close to the railway wheel seat, see Fig.1. The manufacturing process of railway axle usually contains surface treatment of the axle. Compressive residual stresses of extreme magnitudes usually between 20-60 MPa are developed after application of thermomechanical treatment. The compressive residual stresses close to the axle surface can effectively retard fatigue crack propagation and consequently extend the RFL of the axle. Typical distribution of the axial residual stresses is shown in Fig. 2.

44

Author name / Structural IntegrityStructural Procedia 00 (2017)4000–000 Pavel Hutař et al. / Procedia Integrity (2017) 42–47

3

Fig. 1. Scheme of the railway wheelset considered.

It is visible that the highest value of residual stress is close to the surface of the axle and compressive stress field is in the depth between 0-8 mm from the axle surface. This area is very important for RFL prediction (the time of fatigue crack grow up to length 8 mm is higher than the time of fatigue crack grow from the length 8 mm to the critical crack length). Therefore, the residual stresses can significantly influence the RFL of the railway axle and should be considered in the procedure of RFL estimation.

Fig.2. Considered distribution of the axial residual stresses from the outer axle surface to the inner one.

The distribution of residual stresses shown in the Fig. 2 is implemented to the finite element model of the axle. The stress intensity factors are determined according to the methodology described in the paper Náhlík et al. (2017). The stress intensity factors are evaluated using optimised crack front shape obtained by the special procedure described in the mentioned paper (Náhlík et al. (2017)). Fig. 3a shows values of the stress intensity factor as function of the crack length (K-a dependence), which are determined only for the bending load and load caused by the press-fitted wheel. The total stress intensity factor is given by superposition of each stress intensity factor component. Consideration of residual stresses leads to the decrease of total stress intensity factor (compressive residual stresses close to the axle surface reduce the total load), see Fig. 3b. The railway axles are subjected to variable amplitude loading (trains go on straight track, to curved track, over switches, crossovers, etc.). These events commonly increase the basic bending load level caused by weight of the vehicle (Beretta et al. (2016), Pokorný et al. (2016) and Traupe et al. (2004)). Variable amplitude loading is represented by the load spectrum (load block), see histogram of load amplitudes in Fig. 4. Each load amplitude is described by dynamic coefficient k. The static load (caused by weight of the train only) corresponds to k = 1,

4


45

however some events during the train operation can lead to peaks of load amplitude with k = 2.9. The values of stress intensity factor caused by these peaks are depicted in Fig. 3a (bending load).

Fig. 3. Components of the stress intensity factor KI as function of the crack length a for the cases: a) residual stresses are not considered, b) residual stresses are included.

Fig. 4. Load spectrum of railway axle (histogram of loading amplitudes) - Pokorný et al. (2016).

The total load (maximal load) in load cycle, Kmax, is given by superposition of bending (KB), press-fit (KPF) and residual stress field (KRS) contributions:

Kmax (a)  k.K B (a)  K PF (a)  K RS (a) .

(1)

The minimum SIF in load cycle is given by relationship:

Kmin (a)  kK B (a)  K PF (a)  K RS (a) .

(2)


46

5

Then the stress ratio is defined as:

R

K min (a) . K max (a)

(3)

Based on Eq. (3) the operation stress ratio for considered axle was determined in the range from R = -2.5 (for extreme values of residual stresses) to R = 0.

Fig. 5. Experimentally obtained v-K curves of EA4T steel for stress ratios R = -2, -1, -0.5 and 0.1 with NASGRO fit: a) v-K expression, b) v-Kmax expression.

The fatigue crack propagation rates were determined experimentally, see Pokorný et al. (2016b) for details of experiments. The measured v-K curves (curves describing dependence of the crack propagation rate v = da/dN on the stress intensity factor K values) were obtained for stress ratios R = -2, -1, -0.5 and 0.1 which fit well the range of stress asymmetries obtained based on Eq. (4). Table 1. Estimated values of RFL (fatigue crack growth from initial crack length a0 up to final one) [thousands of km]

initial crack length a0 2 mm 3 mm 5 mm 97 39 18 204 63 23 628 107 31 2 943 218 41 15 717 682 55 85 230 3 249 75 inf 17 326 107

peak of axial residual stress 0 -10 -20 -30 -40 -50 -60

Measured v-K data are shown in Fig. 5. It is evident that the experimental curves fit each other well in the case of v-Kmax expression. Based on this fact, the fatigue crack propagation rate can be sufficiently described just by one v-Kmax curve. The v-Kmax interpretation of the fatigue data corresponds well with NASGRO relationship in Kmax expression, see Pokorný et al. (2016b). Therefore, NASGRO relationship can be expressed in the simplified form: p

K  da n  C  K max  1  max,th  , dN K max  

(4)

where C, n and p are material constants, Kmax is maximal value of the stress intensity factor in load cycle, Kmax,th is

6


47

threshold value in Kmax expression. In the range of measured data from R = -2 to R = 0.1 the changes of the threshold value are very small and can be neglected, see Fig.5b (Pokorný et al. (2016b)). Then the crack increment a was evaluated based on Eq. (4). Table 1 shows calculated values of RFL for different considered magnitudes of residual stresses and various initial crack lengths. The estimated RFL of the axle is 97 000 km of train operation in the case of initial crack length a0 = 2 mm, when no residual stresses are considered. The consideration of lower band of residual stresses (-20 MPa) leads to 6 times longer estimated RFL (628 000 km) in comparison to the case, which does not take into account the existence of residual stresses. If the upper band of residual stress (-60 MPa) is considered the estimated RFL is theoretically infinite. This is evident from Fig. 3b, where maximal stress SIF, Kmax, is below the threshold value of initial crack (a0 = 2 mm). The consideration of residual stress is also important in the case of longer initial cracks. However, the ratio between RFL determined for axle with and without consideration of residual stresses decreases with the crack length increase. It can be concluded, that residual stresses induced during thermo-mechanical treatment play an important role in the RFL of the railway axle and should be included in the procedure of determination of frequency of regular inspections together with a definition of required NDT method. 3. Conclusions The paper deals with effect of residual stresses induced during manufacturing process of the railway axle on residual fatigue lifetime. The residual stresses reduce the total stress intensity factor of the fatigue crack and contribute to the resistance of the axle surface to fatigue crack propagation. Exclusion of residual stress from RFL estimation leads to the strong underestimation of axle RFL. In the case of initial crack length a0 = 2 mm the RFL of the axle is about 97 000 km of train operation. The consideration of lower band of residual stresses (-20 MPa) leads to ca 6 times longer RFL (i.e. 628 000 km). For longer initial cracks the influence of residual stresses on RFL decreases, however remains still important. The results obtained show that consideration of residual stresses can explain difference in conservative RFL estimation (without residual stresses) and experimental tests. Acknowledgements The research infrastructure IPMINFRA supported by the Ministry of Education, Youth and Sports of the Czech Republic through the project No. LM2015069 was used. The co-operation between IPM and Bonatrans Group is realized in the frame of Strategy 21 “Top research in the public interest” of the Czech Academy of Sciences. References Benyon J.A., Watson A.S., 2001, The use of Monte-Carlo analysis to increase axle inspection interval, In: Procedings of the 13 th international wheelset congress, Rome, Italy, 2001. Beretta S., Carboni M., Regazzi D., 2016. Load interaction effects in propagation lifetime and inspections of railway axles, Int. J. Fatigue 91/2, 423 Luke M., Varfolomeev I., Lütkepohl K., Esderts A., 2010. Fracture mechanics assessment of railway axles: experimental characterization and computation, Eng. Fail. Anal. 17, 617. Luke M., Varfolomeev I., Lütkepohl K., Esderts A., 2011. Fatigue crack growth in railway axles: assessment concept and validation tests, Eng. Fract. Mech. 78, 714. Náhlík L., Pokorný P., Ševčík M., Fajkoš R., Matušek P., P. Hutař, 2017. Fatigue lifetime estimation of railway axles, Eng. Fail. Anal. 73, 139. Pokorny P., Hutar P., Náhlík L., 2016. Residual fatigue lifetime estimation of railway axles for various loading spectra, Theoretical and Applied Fracture Mechanics 82, 25. Pokorný P., Náhlík L., Hutař P., 2016b. Influence of Variable Stress Ratio During Train Operation on Residual Fatigue Lifetime of Railway Axles, Structural Integrity Procedia 2, 3585 Regazzi D., Beretta S., Carboni M., 2014. An investigation about the influence of deep rolling on fatigue crack growth in railway axles made of a medium strength steel, Eng. Fract. Mech. 138, 587. Smith R.A., 2000. Fatigue of Railway Axles: A Classic Problem Revisited, European Structural Integrity Society 173. Traupe M., Meinen H., Zenner H. 2004. Sichere und wirtschaftliche Auslegung von Eisenbahnfahrwerken. Zerbst U., Mädler K., Hintze H., 2005. Fracture mechanics in railway applications—an overview, Eng. Fract. Mech. 72,163. Zerbst U., Schödel M., Beier H.T., 2011. Parameters affecting the damage tolerance behaviour of railway axles, Eng. Fract. Mech. 78, 793. Zerbst U., Beretta S., Köhler G., Lawton A., Vormwald M., Beier H.T., Klinger C., Černý I., Rudlin J., Heckel T., Klingbeil D., 2013. Safe life and damage tolerance aspects of railway axles – a review, Eng. Fract. Mech. 98, 214. Zerbst U., Klinger C., Klingbeil D., 2013b. Structural assessment of railway axles - a critical review, Eng. Fail. Anal. 35, 54.