- Email: [email protected]
Effect of martensite content and geometry of inclusions on the VHCF properties of predeformed metastable austenitic stainless steels
Available online at www.sciencedirect.com Available online at www.sciencedirect.com
ScienceDirect ScienceDirect
Available online www.sciencedirect.com Available online at at www.sciencedirect.com Structural Integrity Procedia 00 (2016) 000–000 Structural Integrity Procedia 00 (2016) 000–000
ScienceDirect ScienceDirect
Procedia Structural 2 (2016) 1093–1100 Structural IntegrityIntegrity Procedia 00 (2016) 000–000
www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia
21st European Conference on Fracture, ECF21, 20-24 June 2016, Catania, Italy 21st European Conference on Fracture, ECF21, 20-24 June 2016, Catania, Italy
Effect of martensite content and geometry of inclusions on the VHCF properties of predeformed metastable austenitic stainless VHCF properties of predeformed metastable austenitic stainless Thermo-mechanical modelingsteels of a high pressure turbine blade of an steels airplane gas turbine engine Andrei Grigorescua*, Anton Kolyshkina, Martina Zimmermannb, Hans-Jürgen Christa
XVEffect Portuguese Conference on Fracture, 2016, 10-12 February 2016, Paço deon Arcos, of martensite contentPCF and geometry of inclusions thePortugal
a b a Andrei Grigorescua*, Anton Kolyshkin , Martinab Zimmermann c , Hans-Jürgen Christ Institut für Siegen, Germany P.Werkstofftechnik, Brandãoa, Universität V. Infante , 57068 A.M.Siegen, Deus * a
Institut füraInstitut Werkstoffwissenschaft, TU Dresden, Helmholtzstraße 01062Germany Dresden, Germany für Werkstofftechnik, Universität Siegen, 570687,Siegen, a b Department of Mechanical Engineering, Instituto Superior Técnico, Universidade7,de01062 Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa, Institut für Werkstoffwissenschaft, TU Dresden, Helmholtzstraße Dresden, Germany Portugal b IDMEC, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa, Portugal Abstract c CeFEMA, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa, Abstract Portugal b
The effect of inclusions on the VHCF properties of a metastable austenitic stainless steel in the predeformed condition was studied. Theof material contains inhomogeneous distribution of oxide inclusions with an in elongated geometry condition in the rolling The effect inclusions on thean VHCF properties of a metastable austenitic stainless steel the predeformed was direction. Samples werecontains monotonically predeformeddistribution at -80°C start at a strain 0.1 %/s geometry resulting inina the martensite studied. The material an inhomogeneous of temperature oxide inclusions withrate an of elongated rolling Abstract content ofSamples about 60were vol-% and subsequently fatiguedat by means high frequency testing machines. Onresulting the one in hand the high direction. monotonically predeformed -80°C startoftemperature at a strain rate of 0.1 %/s a martensite loading martensite content an subsequently increase of the HCF components strength, onof the other hand to fatigue failure occurs even 107conditions, content of their about 60results vol-%inand fatigued by means frequency testing machines. On thebeyond one hand the high During operation, modern aircraft engine arehigh subjected increasingly demanding operating cycles. Thecontent higher notch sensitivity of (HPT) theofmartensite phaseconditions leads crackparts initiation fromoccurs inclusions accompanied by the martensite results in anturbine increase the HCF Such strength, on to theinternal other these hand fatigue eventypes beyond 107 loading especially the high pressure blades. cause to failure undergo different of time-dependent formation a fine granular area (FGA). A martensite direct correlation between themethod sizecrack of(FEM) theinitiation FGA-and the number of accompanied cycles failure can degradation, onenotch of which is creep. model using the finite element was developed, in order to betoable tobypredict cycles. Theofhigher sensitivity ofAthe phase leads to internal from inclusions the creep ofthe HPT blades. records (FDR) for a VHCF specific aircraft, provided byofacycles commercial aviation be the shown. In abehaviour order find optimal martensite content for between both HCFthe and the the notch sensitivity of the in formation of fine to granular area (FGA). AFlight directdata correlation size of theregime, FGA-and number to material failure can tothe obtain thermal and mechanical data three different flight cycles. In sensitivity order to create the v3D ol%model α’ different conditions wasmartensite investigated. The for results show a significantly higher notch sensitivity be company, shown.predeformed In were order used to find optimal content bothfor HCF and VHCF regime, the notch offor the30 material in neededpredeformed for the FEM analysis, a HPT bladedoscrap wasresults scanned, its chemical composition and material martensite, whereas higher martensite contents notThe show any further significant increase. Since mechanical components are were in α’ different conditions was investigated. showand a significantly higher notch sensitivity forproperties 30 vol% obtained. The data that was gathered was fed FEM differentand simulations were run, first with a simplified practice subjected tohigher complex cyclic loading situations, samples wereand extracted tested Since both parallel and components transversal to martensite, whereas martensite contents dointo notthe show anymodel further significant increase. mechanical arethe in3D rectangular block in to loading better the model,rolling and then with theand realdirection. 3D mesh obtained from bladedirection scrap. rolling direction, in order to order reproduce theestablish relation between and loading The change in the testing practice subjected toshape, complex cyclic situations, samples were extracted tested both parallel and transversal to theThe overall expected inreproduce terms results of displacement was observed, in and particular at the trailing edgechange of the the blade. Therefore such perpendicular to the rolling a larger projection area of the inclusions and reduces number of cycles to a rolling direction, inbehaviour order todirection the in relation between rolling loading direction. The in testing direction model can be useful in the goal of predicting turbine blade life, given a set of FDR data. failure due to the increased intensity factor. this case, the area thethe inclusions (but notreduces of the FGA) correlates with the perpendicular to the rollingstress direction results in a In larger projection areaofof inclusions and the number of cycles to number of to cycles to failure. These findings are In discussed basis of inclusions a detailed (but microstructural characterization of the failure due the increased stress intensity factor. this case,onthethearea of the not of the FGA) correlates with © 2016 The Authors. Published by Elsevier B.V. material of focusing effectThese of martensite the inclusion respect to the rollingcharacterization direction and theofload number cycleson to the failure. findingscontent, are discussed on themorphology basis of a with detailed microstructural the Peer-review under responsibility of the Scientific Committee of PCF 2016. axis applied. material focusing on the effect of martensite content, the inclusion morphology with respect to the rolling direction and the load axis applied. Keywords: HighThe Pressure Turbine Blade;byCreep; Finite 3D Model; Simulation. Copyright © 2016 Authors. Published Elsevier B.V.Element This is Method; an open access article under the CC BY-NC-ND license © 2016 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). © 2016 Theunder Authors. Published by Scientific Elsevier B.V. Peer-review responsibility of the Committee of ECF21.
* Corresponding author. Tel.: +49 (0)271/740-3422; fax: +49 2717402545 E-mail address:author. [email protected] * Corresponding Tel.: +49 (0)271/740-3422; fax: +49 2717402545 E-mail address: [email protected] 2452-3216 © 2016 The Authors. Published by Elsevier B.V. * Corresponding author. Tel.: +351 Peer-review underThe responsibility of218419991. theby Scientific Committee of ECF21. 2452-3216 © 2016 Authors. Published Elsevier B.V. E-mail address: [email protected] Peer-review under responsibility of the Scientific Committee of ECF21.
2452-3216 © 2016 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the Scientific Committee of PCF 2016.
Copyright © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer review under responsibility of the Scientific Committee of ECF21. 10.1016/j.prostr.2016.06.140
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1094 2
Peer-review under responsibility of the Scientific Committee of ECF21. Keywords: austenitic stainless steel, AISI 304, inclusions, very high cycle fatigue;
1. Introduction Nomenclature ΔK ΔKth area FGA
areaproj q Kt Kf
stress intensity factor range threshold of the stress intensity factor range square root of the area of the ‘fine granular area’ the projection area of inclusion measured perpendicular to the applied load notch sensitivity factor stress concentration factor ratio of fatigue limit of smooth specimens to that of notched specimens
Metastable austenitic stainless steels are widespread for the use in sheet metal forming industry. Plastic deformation during forming can lead to a deformation-induced transformation from the fcc austenite to bcc α’ martensite (Maier et al. 1993). The amount of phase transformation depends amongst others on the deformation temperature and is responsible for the pronounced hardening effect during monotonic deformation. Moreover, some authors have found that a martensitic transformation prior to fatigue testing is beneficial for the LCF (Maier et al. 1993) and HCF (Myeon et al. 1997) properties of the material. This study focuses on the very high cycle fatigue (VHCF) strength and the corresponding damage mechanisms of predeformed austenitic stainless steel. Over the last decade it was shown for a remarkable number of metals that failure occurs even beyond the classical fatigue limit of 2∙106-107 cycles. The phase of crack initiation gains importance at larger number of cycles to failure and various reports in literature propose that inhomogeneously distributed and very localized cyclic plastic deformation and a succeeding crack initiation are the dominant life controlling mechanisms (Lukas et al. 2002). Depending on the material studied, the localized plastic deformation can lead to surface as well as subsurface crack initiation, mostly occurring at slip markings or at pores or inclusions, respectively. In order to distinguish between the different damage mechanisms, Mughrabi (2002) suggested the classification of metals for the VHCF range into type I materials showing surface crack initiation, typically single phase ductile metals, and type II materials with inner defects such as inclusions, pores etc. preferentially showing internal crack initiation. Because of their high ductility, metastable austenitic stainless steels can be associated with the category of type I materials regarding the VHCF-behavior. However, due to its pronounced strain hardening behavior the material can reach high flow stresses at higher amounts of deformation induced α’ martensite. Moreover, it can contain different types of inclusions. These properties classify the metastable austenitic stainless steels for the type II category. Müller-Bollenhagen et al. (2010) showed that at high martensite contents this material can fail at very high number of cycles due to crack initiation at internal inclusions. This confirms the type II VHCF-character of the material. In the present study the dominating fatigue damage mechanisms are discussed with regard to the amount of martensite obtained through predeformation. Since mechanical components are subjected to complex loading situations, the focus of the study presented lies on the effect of the orientation of the inclusions with respect to the direction of the applied stress and the fatigue behavior of this material. 2. Experimental setup The material used for this study is the metastable austenitic stainless steel AISI304L. Metal sheets with 2 mm thickness were recrystallized after cold rolling so that any likely texture effects were reduced to a minimum. Fig. 1 schematically shows the main denomination conventions used in this paper. ND, TD and RD refer to normal,
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1095 3
transversal and rolling direction, respectively. The investigated material contains an inhomogeneous distribution of elongated oxide inclusions with their longitudinal shapes oriented parallel to the rolling direction. Consequently, a change of the load axis from RD to TD results in an increase of the projected area (areaproj) of the inclusions along the applied load direction as depicted in Fig 1. The amount of α’ martensite was adjusted by a predeformation. Specimens were cut parallel to the rolling direction (RD-samples) as well as perpendicular to the rolling direction (TD-samples) and then monotonically deformed up to a true plastic strain of εpl=0.15 at starting temperatures of -40°C and -90°C resulting in α’ martensite contents of 30 vol-% and 60 vol-%, respectively. Subsequently, fatigue specimens were machined and then mechanically and electro-chemically polished. For the investigation of the notch sensitivity, sharp U-shaped notches were introduced at the edge of flat specimens by means of spark erosion. For a notch radius of 250 µm the notch factor determined by means of FEM calculations was Kt=3.2.
Fig. 1. Schematic representation of the orientation of fatigue samples and distribution of elongated inclusions with respect to the rolling and to the transversal direction of the metal sheet
For the fatigue tests a resonance pulsating test system operating at ~90 Hz and an ultrasonic test system (BOKU, Vienna) working at a frequency of ~20 kHz were used. The fractographic investigations and the characterization of the microstructure of the fatigue samples were carried out using a scanning electron microscope (FEI, Helios). 3. Results and discussion 3.1. Fatigue results The fatigue results are depicted in figure 2a. The relationship between the martensite volume fraction and the fatigue properties of the metastable austenitic stainless steel was investigated using samples prepared from the asreceived fully austenitic condition and the two predeformed conditions (30 vol-% and 60 vol-% of α’) in the rolling direction (RD-samples, empty symbols).
Fig. 2. (a) SN-curves; (b) Fracture surface TD-sample (a=490 MPa); (c) Fracture surface RD-sample (a=485 MPa)
1096 4
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
In the fully austenitic condition the material exhibits a true durability at 260 MPa in the HCF and VHCF region. In the predeformed condition the material shows a steady increase of cyclic strength in the HCF regime with increasing volume fraction of α’ martensite at 30 vol-% and 60 vol-%. In the VHCF region quite a different picture emerges. For the specimens containing 30 vol-% martensite there is a constant fatigue limit in the VHCF regime which is equal to the cyclic strength in the HCF regime. For a volume fraction of 60% failure occurs up to 109 cycles and the fatigue limit decreases in the VHCF regime resulting in a two-step S-N-curve. In this regime, all cracks initiated at internal inclusions. The fracture surfaces show the typical “fish-eye” appearance observed for type II materials (Fig. 2b and 2c). This change in fatigue behavior can be explained by comparing the microstructures at 30 vol-% and 60 vol-% martensite (Fig. 3). For the lower volume fraction the martensite is allocated as single needles that are only about to start forming a coherent network. These needles are smaller than the typical dimensions of the inclusions, which remain surrounded by a predominantly austenitic phase. The localized cyclic hardening impede in this case the crack initiation at the inclusions. At 60 vol-% martensite the microstructure is dominated by coherent clusters of α’ martensite. This condition promotes internal crack initiation and the likely underlying mechanisms will be discussed in the next chapter.
Fig. 3. Phase distribution in the predeformed samples containing 30 vol-% (a) and 60 vol-% (b) ’ martensite phase
Li (2012) reported that the fatigue life of high strength steels containing non-metallic inclusions is predominantly determined by the position and the size (areaproj - the projection area of inclusion measured perpendicular to the applied load) of the inclusions. Due to the difference in size and shape of the inhomogeneously distributed inclusions with respect to the RD and TD direction, the material analysed in this study offers the possibility to systematically investigate these effects on the fatigue life in the VHCF regime. In order to further deepen the understanding of this aspect, additional tensile specimens were taken perpendicular to the rolling direction (TD-samples, full symbols) and then prepared and tested in the same conditions (with 60 vol-% α’ martensite) as the specimens predeformed in the rolling direction (RD-samples, empty symbols). The results of the fatigue tests are depicted also in figure 2. As expected, the change in the loading direction from RD to TD leads to a decrease of the fatigue life in the VHCF regime which is attributed to the larger areaproj of the crack initiating inclusions. A detailed investigation on the damage mechanisms and the correlation between fatigue life and inclusion geometry is presented in the subsequent chapter. 3.2. Damage mechanisms in the VHCF regime The VHCF failure of AISI 304L with a high amount of α’ martensite (e.g. 60 vol-%) is characterised by internal crack initiation at non-metallic inclusions for both RD- and TD-samples. The fractographic investigations for all the samples that failed beyond 107 cycles show typical “fish-eye” morphology. The crack initiating inclusions are surrounded by a fine granular area (FGA), which is characteristic for the VHCF failure of martensitic steels (Sakai 2009). Using the fracture-mechanically based concept proposed by Murakami (2002) the cyclic stress intensity factor of the FGA can be assessed by
K FGA 0.5 areaFGA
(1)
For the RD-specimens which failed in the VHCF regime an average ΔKFGA value of 4 MPa√m was calculated.
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1097 5
This stands in good agreement with the threshold value for fatigue crack growth of the martensitic phase ΔKth = 5 MPa√m as measured by Bowe et al. (1988). Hence, under the assumption of the applicability of the ΔK concept to the given VHCF mechanism, one can regard the FGA formation as the crack initiation stage. After the FGA reaches a critical value (dependent on the local stress amplitude) stage two fatigue crack growth can be assumed, developing a smooth fracture surface due to the neglectable microstructural influence, which together with the FGA gives the typical “fish-eye” appearance. The microstructure surrounding the inclusions was investigated on the basis of cross-sections prepared perpendicular to the crack initiating inclusions and subsequent SEM analysis using the EBSD-method. The results are presented in Fig. 4 and Fig. 5.
Fig. 4. (a) Fracture surface RD-sample with 60 vol-% α’ martensite; (b) Cross-section inclusion; (c) Phase-map around inclusion
Fig. 5. (a) Fracture surface TD-sample with 60 vol-% α’ martensite;; (b) Cross-section inclusion; (c) Phase-map around inclusion
One can observe that elongated inclusions are responsible for the crack initiation in both RD and TD-samples. For the RD-sample the FGA was not formed as expected in the largest section of the inclusion (Fig. 4b). In this case, the FGA started at a site where the surrounding microstructure is almost fully martensitic (Fig. 4c). This appears to be a condition for the formation of FGA when the elliptical inclusions have the long axis parallel to the loading direction. The high volume fractions of α’ martensite and the distribution in coherent large clusters increases the probability that some planes of the inclusions are completely surrounded by the α’ martensite phase. The higher strength and brittleness of the martensite phase become predominant at higher martensitic contents and allows for FGA formation also at very low stress intensity factor ranges (1.42 MPa√m see Fig. 7a). The TD-samples are characterized by a large projected area of the elongated inclusion on the crack surface. The starting ΔKincl is thus generally larger than in the aforementioned RD-samples. In this case the prerequisite of a completely martensitic surrounding matrix in the plane of FGA formation is less likely to be fulfilled. Nevertheless, the formation of an FGA around the inclusions can be observed for this condition, too. It can be assumed that, although the critical cross section of the inclusion was probably not fully surrounded by martensite, the high stress concentration caused by the inclusion may have induced a phase transformation in the already highly strained and hardened austenitic phase of the microstructure attaining the necessary constraint for FGA formation. Microcracks can also be initiated due to the high notch sensitivity of the duplex microstructure. Fig. 5a shows such a microcrack observed beneath the cracksurface of the sample with microstructure-controlled propagation in the austenitic phase. In this context the notch sensitivity of the duplex microstructure austenite + martensite gains importance and will be discussed in the next
1098 6
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
chapter. The micrographs in figure 5b and 5c indicate also a tendency towards fine grain formation in the vicinity of the inclusions. Sakai (2009) proposed the polygonization of very fine subgrains as a pre-stage of FGA formation which stands in good agreement with the aforementioned observation. However, the limited resolution of the SEMEBSD technique used in the presented study does not allow an unequivocal statement about the suitability of this theory in case of FGA formation for metastable austenitic stainless steels. Murakami et al. (1999) explain the FGA growth as a consequence of higher hydrogen concentrations in the martensitic phase around the inclusion which increases the mobility of dislocations. Moreover, Murakami and Matsunaga (2006) found accelerated crack propagation in the martensitic phase at the crack tip of hydrogen-charged specimens of a metastable austenitic steel. Hence, the hydrogen-assisted FGA formation appears plausible in the case of predeformed metastable austenitic stainless steels tested in the VHCF regime. Grad et al. (2012) showed by means of FIB-prepared TEM-foils that the grain refinement in the vicinity of the inclusions leads to a local decrease of the ΔKth as can be observed for fine grained materials. It was shown that this process takes the largest fraction of the total life of the samples. For the RD samples investigated in this study an increase in the area of the FGA with the number of cycles to failure was observed (Fig 6a). These results are consistent with the observations made by Sakai (2009) and demonstrate that the number of cycles to failure strongly correlates with the mechanism of FGA formation. This relationship could not be confirmed for the TD-samples. In this case the projected area of the inclusion relative to the area of FGA is larger and variations in the dimension of the inclusions have a decisive influence on the fatigue life of the specimens.
Fig. 6. Correlation between √area of FGA and number of loading cycles for: (a) RD-samples; (b) TD-samples
Sakai (2009) analysed the mechanisms of internal crack initiation from the view point of fracture mechanics using the concept proposed by Murakami (2002) and correlated ΔKincl with the number of cycles to failure. Using the same fracture-mechanics concept, the stress intensity factor range was calculated for the RD- and TD-samples in this study and plotted against the number of cycles to failure. The results depicted in figure 7b confirm the clear correlation between ΔKincl and fatigue life and demonstrate the importance of the dimensions of large inclusions (TD-samples) as a controlling factor of the fatigue life.
Fig. 7. Correlation between stress intensity factor at inclusions and number of loading cycles for: (a) RD-samples; (b) TD-samples
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1099 7
The RD-samples are characterized be smaller ΔKincl (due to smaller projarea) and in this case the FGA has to grow first at smaller rates. This process depends strongly on the surrounding microstructure as mentioned before and results in very inhomogeneous FGA growth rates explaining the randomly distributed points in figure 7a. Obviously the notch sensitivity of the microstructure surrounding the inclusions plays an important role for the fatigue behavior of the material. In this respect samples with smooth and notched surfaces in the same predeformation conditions were tested in order to determine the influence of the α’ martensite volume fraction on the notch sensitivity of the material. The notch sensitivity factor q is defined as
q
K f 1 Kt 1
(2)
where Kf represents the ratio of fatigue limit of smooth specimens to that of notched specimens and Kt is the stress concentration factor. The results of the fatigue tests are depicted in Fig. 8a. For predeformed smooth specimens containing 60 vol-% α’ martensite, the fatigue limit is defined as the stress amplitude at which the sample reaches 2·107 without failure initiated at the surface. All the specimens which failed in the VHCF regime are plotted as runouts in the Fig.8a. The calculated notch sensitivity was plotted against the martensitic content in Fig. 8b.
Fig. 8. (a) Fatigue results for smooth and notched specimens; (b) calculated notch sensitivity in relation to the martensitic content
The results show a strong increase in the notch sensitivity at 30 vol-% α’ martensite, whereas higher martensite contents do not result in further significant increase of the notch sensitivity of the material. This behavior can be explained by the strong hardening effect of the formation of a coherent martensitic network in the microstructure observed in figure 3a. Although these results cannot be interpreted in direct relation with crack initiation at small internal inclusions, they represent a plausible indicator of a negative effect of high martensite contents on the fatigue behavior of the material. Thus the martensite volume fraction obtained during cold working processes should remain beneath 30% in order to keep the notch sensitivity low and avoid fatigue failure at very high number of cycles. 4. Conclusions The fatigue behavior of the metastable austenitic stainless steel AISI304L was investigated and the following conclusions can be drawn: During the predeformation a coherent network of martensitic needles starts to form at 30 vol-% α’ martensite. For volume fractions ≤ 30 % α’ martensite, the material exhibits a true durability without failure in the VHCF regime. This behavior is explained by the local cyclic hardening of the soft austenitic phase
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1100 8
resulting in a decrease of strain amplitude. At higher amounts of α’ martensite the damage mechanisms change and failure occurs after very high number of cycles due to internal crack initiation at inclusions accompanied by the formation of a fine granular area. A direct correlation between the FGA-area and the number of cycles to failure could be found for RDsamples (with small areaproj) whereas no such relationship could be confirmed for the TD-samples (with larger areaproj). This confirms the fatigue life controlling role of the crack initiating inclusions if they exceed a certain projected area. This statement is supported by the correlation between the stress intensity factor at inclusions and the fatigue life, which was confirmed only for the TD-samples. In the case of RDsamples the lack of correlation is explained by the dependence of the FGA-growth on the surrounding microstructure for the small areaproj of the crack initiating inclusions. The notch sensitivity increases at 30 vol-% α’-martensite, whereas higher martensite contents do not show any further significant increase. In this respect, a martensite content lower than 30 vol-% is recommended for optimized HCF and VHCF properties.
Acknowledgements The authors gratefully acknowledge the financial support of this study by Deutsche Forschungsgemeinschaft in the framework of the priority programme Life (DFG-SPP1466). References Bowe, K., H., Hammerschmidt, W., Hornbogen, E., Hühner, M., 1988. Bruchmechanische Eigenschaften von metastabilen Austeniten. Materialwissenschaft und Werkstofftechnik 19, 193-201. Grad, P., Reuscher, B., Brodyanski, A., Kopnarski, M., Kerscher, E., 2012. Mechanism of fatigue crack initiation and propagation in the very high cycle regime of high-strength steels. Scripta Materialia 67, 838-841. Li, S., X., 2012. Effects of inclusions on very high cycle fatigue properties of high strength steels. International Materials Reviews 57, 92-114. Lukas, P., Kunz, L., 2002. Specific features of high-cycle and ultra-high-cycle fatigue. Fatigue&Fracture Enginnering Materials & Structures 25, 747-753 Maier, H., Donth, B., Bayerlein, M., Mughrabi, H., Maier, B., Kesten, M., 1993. Optimierte Festigkeitssteigerung eines metastabilen austenitischen Stahles durch verformungsinduzierte Martensitumwandlung bei tiefen Temperaturen. Zeitschrifft für Metallkunde 84, 820-826. Mughrabi, H., 2002. On ‘Multi-stage’ fatigue life diagrams and the relevant life-controlling mechanisms in ultrahigh-cycle fatigue. Fatigue&Fracture Enginnering Materials & Structures 25, 755-764 Murakami, Y., 2002. Metal Fatigue: Effects of Small Defects and Nonmetallic Inclusions, Elsevier, Oxford. Murakami, Y., Nomoto, T., Ueda, T., 1999. Factors influencing the mechanism of superlong fatigue failure in steels, Fatigue & Fracture of Engineering Materials & Structures 22, 581-590. Murakami, Y., Matsunaga, H., 2006. The effect of hydrogen on fatigue properties of steels used for fuel cell systems. International Journal of Fatigue 28, 1509-1520 Müller-Bollenhagen, C., Zimmermann, M., Christ, H.-J., 2010. Adjusting the very high cycle fatigue properties of a metastable austenitic stainless steel by means of the martensite content. Procedia Engineering 2, 1663-1672. Myeon, T., H., Yamabayashi, Y., Shimojo, Y., Higo, Y., 1997. A new life extension method for high cycle fatigue using micro martensitic transformation in austenitic stainless steel. International Journal of Fatigue 19, 69-73. Sakai, T., 2009. Review and prospects for current studies on very high cycle fatigue of metallic materials for machine structural use. Journal of Solid Mechanics and Materials Engineering 3, 425-439.
ScienceDirect ScienceDirect
Available online www.sciencedirect.com Available online at at www.sciencedirect.com Structural Integrity Procedia 00 (2016) 000–000 Structural Integrity Procedia 00 (2016) 000–000
ScienceDirect ScienceDirect
Procedia Structural 2 (2016) 1093–1100 Structural IntegrityIntegrity Procedia 00 (2016) 000–000
www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia
21st European Conference on Fracture, ECF21, 20-24 June 2016, Catania, Italy 21st European Conference on Fracture, ECF21, 20-24 June 2016, Catania, Italy
Effect of martensite content and geometry of inclusions on the VHCF properties of predeformed metastable austenitic stainless VHCF properties of predeformed metastable austenitic stainless Thermo-mechanical modelingsteels of a high pressure turbine blade of an steels airplane gas turbine engine Andrei Grigorescua*, Anton Kolyshkina, Martina Zimmermannb, Hans-Jürgen Christa
XVEffect Portuguese Conference on Fracture, 2016, 10-12 February 2016, Paço deon Arcos, of martensite contentPCF and geometry of inclusions thePortugal
a b a Andrei Grigorescua*, Anton Kolyshkin , Martinab Zimmermann c , Hans-Jürgen Christ Institut für Siegen, Germany P.Werkstofftechnik, Brandãoa, Universität V. Infante , 57068 A.M.Siegen, Deus * a
Institut füraInstitut Werkstoffwissenschaft, TU Dresden, Helmholtzstraße 01062Germany Dresden, Germany für Werkstofftechnik, Universität Siegen, 570687,Siegen, a b Department of Mechanical Engineering, Instituto Superior Técnico, Universidade7,de01062 Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa, Institut für Werkstoffwissenschaft, TU Dresden, Helmholtzstraße Dresden, Germany Portugal b IDMEC, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa, Portugal Abstract c CeFEMA, Department of Mechanical Engineering, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1, 1049-001 Lisboa, Abstract Portugal b
The effect of inclusions on the VHCF properties of a metastable austenitic stainless steel in the predeformed condition was studied. Theof material contains inhomogeneous distribution of oxide inclusions with an in elongated geometry condition in the rolling The effect inclusions on thean VHCF properties of a metastable austenitic stainless steel the predeformed was direction. Samples werecontains monotonically predeformeddistribution at -80°C start at a strain 0.1 %/s geometry resulting inina the martensite studied. The material an inhomogeneous of temperature oxide inclusions withrate an of elongated rolling Abstract content ofSamples about 60were vol-% and subsequently fatiguedat by means high frequency testing machines. Onresulting the one in hand the high direction. monotonically predeformed -80°C startoftemperature at a strain rate of 0.1 %/s a martensite loading martensite content an subsequently increase of the HCF components strength, onof the other hand to fatigue failure occurs even 107conditions, content of their about 60results vol-%inand fatigued by means frequency testing machines. On thebeyond one hand the high During operation, modern aircraft engine arehigh subjected increasingly demanding operating cycles. Thecontent higher notch sensitivity of (HPT) theofmartensite phaseconditions leads crackparts initiation fromoccurs inclusions accompanied by the martensite results in anturbine increase the HCF Such strength, on to theinternal other these hand fatigue eventypes beyond 107 loading especially the high pressure blades. cause to failure undergo different of time-dependent formation a fine granular area (FGA). A martensite direct correlation between themethod sizecrack of(FEM) theinitiation FGA-and the number of accompanied cycles failure can degradation, onenotch of which is creep. model using the finite element was developed, in order to betoable tobypredict cycles. Theofhigher sensitivity ofAthe phase leads to internal from inclusions the creep ofthe HPT blades. records (FDR) for a VHCF specific aircraft, provided byofacycles commercial aviation be the shown. In abehaviour order find optimal martensite content for between both HCFthe and the the notch sensitivity of the in formation of fine to granular area (FGA). AFlight directdata correlation size of theregime, FGA-and number to material failure can tothe obtain thermal and mechanical data three different flight cycles. In sensitivity order to create the v3D ol%model α’ different conditions wasmartensite investigated. The for results show a significantly higher notch sensitivity be company, shown.predeformed In were order used to find optimal content bothfor HCF and VHCF regime, the notch offor the30 material in neededpredeformed for the FEM analysis, a HPT bladedoscrap wasresults scanned, its chemical composition and material martensite, whereas higher martensite contents notThe show any further significant increase. Since mechanical components are were in α’ different conditions was investigated. showand a significantly higher notch sensitivity forproperties 30 vol% obtained. The data that was gathered was fed FEM differentand simulations were run, first with a simplified practice subjected tohigher complex cyclic loading situations, samples wereand extracted tested Since both parallel and components transversal to martensite, whereas martensite contents dointo notthe show anymodel further significant increase. mechanical arethe in3D rectangular block in to loading better the model,rolling and then with theand realdirection. 3D mesh obtained from bladedirection scrap. rolling direction, in order to order reproduce theestablish relation between and loading The change in the testing practice subjected toshape, complex cyclic situations, samples were extracted tested both parallel and transversal to theThe overall expected inreproduce terms results of displacement was observed, in and particular at the trailing edgechange of the the blade. Therefore such perpendicular to the rolling a larger projection area of the inclusions and reduces number of cycles to a rolling direction, inbehaviour order todirection the in relation between rolling loading direction. The in testing direction model can be useful in the goal of predicting turbine blade life, given a set of FDR data. failure due to the increased intensity factor. this case, the area thethe inclusions (but notreduces of the FGA) correlates with the perpendicular to the rollingstress direction results in a In larger projection areaofof inclusions and the number of cycles to number of to cycles to failure. These findings are In discussed basis of inclusions a detailed (but microstructural characterization of the failure due the increased stress intensity factor. this case,onthethearea of the not of the FGA) correlates with © 2016 The Authors. Published by Elsevier B.V. material of focusing effectThese of martensite the inclusion respect to the rollingcharacterization direction and theofload number cycleson to the failure. findingscontent, are discussed on themorphology basis of a with detailed microstructural the Peer-review under responsibility of the Scientific Committee of PCF 2016. axis applied. material focusing on the effect of martensite content, the inclusion morphology with respect to the rolling direction and the load axis applied. Keywords: HighThe Pressure Turbine Blade;byCreep; Finite 3D Model; Simulation. Copyright © 2016 Authors. Published Elsevier B.V.Element This is Method; an open access article under the CC BY-NC-ND license © 2016 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). © 2016 Theunder Authors. Published by Scientific Elsevier B.V. Peer-review responsibility of the Committee of ECF21.
* Corresponding author. Tel.: +49 (0)271/740-3422; fax: +49 2717402545 E-mail address:author. [email protected] * Corresponding Tel.: +49 (0)271/740-3422; fax: +49 2717402545 E-mail address: [email protected] 2452-3216 © 2016 The Authors. Published by Elsevier B.V. * Corresponding author. Tel.: +351 Peer-review underThe responsibility of218419991. theby Scientific Committee of ECF21. 2452-3216 © 2016 Authors. Published Elsevier B.V. E-mail address: [email protected] Peer-review under responsibility of the Scientific Committee of ECF21.
2452-3216 © 2016 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the Scientific Committee of PCF 2016.
Copyright © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer review under responsibility of the Scientific Committee of ECF21. 10.1016/j.prostr.2016.06.140
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1094 2
Peer-review under responsibility of the Scientific Committee of ECF21. Keywords: austenitic stainless steel, AISI 304, inclusions, very high cycle fatigue;
1. Introduction Nomenclature ΔK ΔKth area FGA
areaproj q Kt Kf
stress intensity factor range threshold of the stress intensity factor range square root of the area of the ‘fine granular area’ the projection area of inclusion measured perpendicular to the applied load notch sensitivity factor stress concentration factor ratio of fatigue limit of smooth specimens to that of notched specimens
Metastable austenitic stainless steels are widespread for the use in sheet metal forming industry. Plastic deformation during forming can lead to a deformation-induced transformation from the fcc austenite to bcc α’ martensite (Maier et al. 1993). The amount of phase transformation depends amongst others on the deformation temperature and is responsible for the pronounced hardening effect during monotonic deformation. Moreover, some authors have found that a martensitic transformation prior to fatigue testing is beneficial for the LCF (Maier et al. 1993) and HCF (Myeon et al. 1997) properties of the material. This study focuses on the very high cycle fatigue (VHCF) strength and the corresponding damage mechanisms of predeformed austenitic stainless steel. Over the last decade it was shown for a remarkable number of metals that failure occurs even beyond the classical fatigue limit of 2∙106-107 cycles. The phase of crack initiation gains importance at larger number of cycles to failure and various reports in literature propose that inhomogeneously distributed and very localized cyclic plastic deformation and a succeeding crack initiation are the dominant life controlling mechanisms (Lukas et al. 2002). Depending on the material studied, the localized plastic deformation can lead to surface as well as subsurface crack initiation, mostly occurring at slip markings or at pores or inclusions, respectively. In order to distinguish between the different damage mechanisms, Mughrabi (2002) suggested the classification of metals for the VHCF range into type I materials showing surface crack initiation, typically single phase ductile metals, and type II materials with inner defects such as inclusions, pores etc. preferentially showing internal crack initiation. Because of their high ductility, metastable austenitic stainless steels can be associated with the category of type I materials regarding the VHCF-behavior. However, due to its pronounced strain hardening behavior the material can reach high flow stresses at higher amounts of deformation induced α’ martensite. Moreover, it can contain different types of inclusions. These properties classify the metastable austenitic stainless steels for the type II category. Müller-Bollenhagen et al. (2010) showed that at high martensite contents this material can fail at very high number of cycles due to crack initiation at internal inclusions. This confirms the type II VHCF-character of the material. In the present study the dominating fatigue damage mechanisms are discussed with regard to the amount of martensite obtained through predeformation. Since mechanical components are subjected to complex loading situations, the focus of the study presented lies on the effect of the orientation of the inclusions with respect to the direction of the applied stress and the fatigue behavior of this material. 2. Experimental setup The material used for this study is the metastable austenitic stainless steel AISI304L. Metal sheets with 2 mm thickness were recrystallized after cold rolling so that any likely texture effects were reduced to a minimum. Fig. 1 schematically shows the main denomination conventions used in this paper. ND, TD and RD refer to normal,
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1095 3
transversal and rolling direction, respectively. The investigated material contains an inhomogeneous distribution of elongated oxide inclusions with their longitudinal shapes oriented parallel to the rolling direction. Consequently, a change of the load axis from RD to TD results in an increase of the projected area (areaproj) of the inclusions along the applied load direction as depicted in Fig 1. The amount of α’ martensite was adjusted by a predeformation. Specimens were cut parallel to the rolling direction (RD-samples) as well as perpendicular to the rolling direction (TD-samples) and then monotonically deformed up to a true plastic strain of εpl=0.15 at starting temperatures of -40°C and -90°C resulting in α’ martensite contents of 30 vol-% and 60 vol-%, respectively. Subsequently, fatigue specimens were machined and then mechanically and electro-chemically polished. For the investigation of the notch sensitivity, sharp U-shaped notches were introduced at the edge of flat specimens by means of spark erosion. For a notch radius of 250 µm the notch factor determined by means of FEM calculations was Kt=3.2.
Fig. 1. Schematic representation of the orientation of fatigue samples and distribution of elongated inclusions with respect to the rolling and to the transversal direction of the metal sheet
For the fatigue tests a resonance pulsating test system operating at ~90 Hz and an ultrasonic test system (BOKU, Vienna) working at a frequency of ~20 kHz were used. The fractographic investigations and the characterization of the microstructure of the fatigue samples were carried out using a scanning electron microscope (FEI, Helios). 3. Results and discussion 3.1. Fatigue results The fatigue results are depicted in figure 2a. The relationship between the martensite volume fraction and the fatigue properties of the metastable austenitic stainless steel was investigated using samples prepared from the asreceived fully austenitic condition and the two predeformed conditions (30 vol-% and 60 vol-% of α’) in the rolling direction (RD-samples, empty symbols).
Fig. 2. (a) SN-curves; (b) Fracture surface TD-sample (a=490 MPa); (c) Fracture surface RD-sample (a=485 MPa)
1096 4
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
In the fully austenitic condition the material exhibits a true durability at 260 MPa in the HCF and VHCF region. In the predeformed condition the material shows a steady increase of cyclic strength in the HCF regime with increasing volume fraction of α’ martensite at 30 vol-% and 60 vol-%. In the VHCF region quite a different picture emerges. For the specimens containing 30 vol-% martensite there is a constant fatigue limit in the VHCF regime which is equal to the cyclic strength in the HCF regime. For a volume fraction of 60% failure occurs up to 109 cycles and the fatigue limit decreases in the VHCF regime resulting in a two-step S-N-curve. In this regime, all cracks initiated at internal inclusions. The fracture surfaces show the typical “fish-eye” appearance observed for type II materials (Fig. 2b and 2c). This change in fatigue behavior can be explained by comparing the microstructures at 30 vol-% and 60 vol-% martensite (Fig. 3). For the lower volume fraction the martensite is allocated as single needles that are only about to start forming a coherent network. These needles are smaller than the typical dimensions of the inclusions, which remain surrounded by a predominantly austenitic phase. The localized cyclic hardening impede in this case the crack initiation at the inclusions. At 60 vol-% martensite the microstructure is dominated by coherent clusters of α’ martensite. This condition promotes internal crack initiation and the likely underlying mechanisms will be discussed in the next chapter.
Fig. 3. Phase distribution in the predeformed samples containing 30 vol-% (a) and 60 vol-% (b) ’ martensite phase
Li (2012) reported that the fatigue life of high strength steels containing non-metallic inclusions is predominantly determined by the position and the size (areaproj - the projection area of inclusion measured perpendicular to the applied load) of the inclusions. Due to the difference in size and shape of the inhomogeneously distributed inclusions with respect to the RD and TD direction, the material analysed in this study offers the possibility to systematically investigate these effects on the fatigue life in the VHCF regime. In order to further deepen the understanding of this aspect, additional tensile specimens were taken perpendicular to the rolling direction (TD-samples, full symbols) and then prepared and tested in the same conditions (with 60 vol-% α’ martensite) as the specimens predeformed in the rolling direction (RD-samples, empty symbols). The results of the fatigue tests are depicted also in figure 2. As expected, the change in the loading direction from RD to TD leads to a decrease of the fatigue life in the VHCF regime which is attributed to the larger areaproj of the crack initiating inclusions. A detailed investigation on the damage mechanisms and the correlation between fatigue life and inclusion geometry is presented in the subsequent chapter. 3.2. Damage mechanisms in the VHCF regime The VHCF failure of AISI 304L with a high amount of α’ martensite (e.g. 60 vol-%) is characterised by internal crack initiation at non-metallic inclusions for both RD- and TD-samples. The fractographic investigations for all the samples that failed beyond 107 cycles show typical “fish-eye” morphology. The crack initiating inclusions are surrounded by a fine granular area (FGA), which is characteristic for the VHCF failure of martensitic steels (Sakai 2009). Using the fracture-mechanically based concept proposed by Murakami (2002) the cyclic stress intensity factor of the FGA can be assessed by
K FGA 0.5 areaFGA
(1)
For the RD-specimens which failed in the VHCF regime an average ΔKFGA value of 4 MPa√m was calculated.
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1097 5
This stands in good agreement with the threshold value for fatigue crack growth of the martensitic phase ΔKth = 5 MPa√m as measured by Bowe et al. (1988). Hence, under the assumption of the applicability of the ΔK concept to the given VHCF mechanism, one can regard the FGA formation as the crack initiation stage. After the FGA reaches a critical value (dependent on the local stress amplitude) stage two fatigue crack growth can be assumed, developing a smooth fracture surface due to the neglectable microstructural influence, which together with the FGA gives the typical “fish-eye” appearance. The microstructure surrounding the inclusions was investigated on the basis of cross-sections prepared perpendicular to the crack initiating inclusions and subsequent SEM analysis using the EBSD-method. The results are presented in Fig. 4 and Fig. 5.
Fig. 4. (a) Fracture surface RD-sample with 60 vol-% α’ martensite; (b) Cross-section inclusion; (c) Phase-map around inclusion
Fig. 5. (a) Fracture surface TD-sample with 60 vol-% α’ martensite;; (b) Cross-section inclusion; (c) Phase-map around inclusion
One can observe that elongated inclusions are responsible for the crack initiation in both RD and TD-samples. For the RD-sample the FGA was not formed as expected in the largest section of the inclusion (Fig. 4b). In this case, the FGA started at a site where the surrounding microstructure is almost fully martensitic (Fig. 4c). This appears to be a condition for the formation of FGA when the elliptical inclusions have the long axis parallel to the loading direction. The high volume fractions of α’ martensite and the distribution in coherent large clusters increases the probability that some planes of the inclusions are completely surrounded by the α’ martensite phase. The higher strength and brittleness of the martensite phase become predominant at higher martensitic contents and allows for FGA formation also at very low stress intensity factor ranges (1.42 MPa√m see Fig. 7a). The TD-samples are characterized by a large projected area of the elongated inclusion on the crack surface. The starting ΔKincl is thus generally larger than in the aforementioned RD-samples. In this case the prerequisite of a completely martensitic surrounding matrix in the plane of FGA formation is less likely to be fulfilled. Nevertheless, the formation of an FGA around the inclusions can be observed for this condition, too. It can be assumed that, although the critical cross section of the inclusion was probably not fully surrounded by martensite, the high stress concentration caused by the inclusion may have induced a phase transformation in the already highly strained and hardened austenitic phase of the microstructure attaining the necessary constraint for FGA formation. Microcracks can also be initiated due to the high notch sensitivity of the duplex microstructure. Fig. 5a shows such a microcrack observed beneath the cracksurface of the sample with microstructure-controlled propagation in the austenitic phase. In this context the notch sensitivity of the duplex microstructure austenite + martensite gains importance and will be discussed in the next
1098 6
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
chapter. The micrographs in figure 5b and 5c indicate also a tendency towards fine grain formation in the vicinity of the inclusions. Sakai (2009) proposed the polygonization of very fine subgrains as a pre-stage of FGA formation which stands in good agreement with the aforementioned observation. However, the limited resolution of the SEMEBSD technique used in the presented study does not allow an unequivocal statement about the suitability of this theory in case of FGA formation for metastable austenitic stainless steels. Murakami et al. (1999) explain the FGA growth as a consequence of higher hydrogen concentrations in the martensitic phase around the inclusion which increases the mobility of dislocations. Moreover, Murakami and Matsunaga (2006) found accelerated crack propagation in the martensitic phase at the crack tip of hydrogen-charged specimens of a metastable austenitic steel. Hence, the hydrogen-assisted FGA formation appears plausible in the case of predeformed metastable austenitic stainless steels tested in the VHCF regime. Grad et al. (2012) showed by means of FIB-prepared TEM-foils that the grain refinement in the vicinity of the inclusions leads to a local decrease of the ΔKth as can be observed for fine grained materials. It was shown that this process takes the largest fraction of the total life of the samples. For the RD samples investigated in this study an increase in the area of the FGA with the number of cycles to failure was observed (Fig 6a). These results are consistent with the observations made by Sakai (2009) and demonstrate that the number of cycles to failure strongly correlates with the mechanism of FGA formation. This relationship could not be confirmed for the TD-samples. In this case the projected area of the inclusion relative to the area of FGA is larger and variations in the dimension of the inclusions have a decisive influence on the fatigue life of the specimens.
Fig. 6. Correlation between √area of FGA and number of loading cycles for: (a) RD-samples; (b) TD-samples
Sakai (2009) analysed the mechanisms of internal crack initiation from the view point of fracture mechanics using the concept proposed by Murakami (2002) and correlated ΔKincl with the number of cycles to failure. Using the same fracture-mechanics concept, the stress intensity factor range was calculated for the RD- and TD-samples in this study and plotted against the number of cycles to failure. The results depicted in figure 7b confirm the clear correlation between ΔKincl and fatigue life and demonstrate the importance of the dimensions of large inclusions (TD-samples) as a controlling factor of the fatigue life.
Fig. 7. Correlation between stress intensity factor at inclusions and number of loading cycles for: (a) RD-samples; (b) TD-samples
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1099 7
The RD-samples are characterized be smaller ΔKincl (due to smaller projarea) and in this case the FGA has to grow first at smaller rates. This process depends strongly on the surrounding microstructure as mentioned before and results in very inhomogeneous FGA growth rates explaining the randomly distributed points in figure 7a. Obviously the notch sensitivity of the microstructure surrounding the inclusions plays an important role for the fatigue behavior of the material. In this respect samples with smooth and notched surfaces in the same predeformation conditions were tested in order to determine the influence of the α’ martensite volume fraction on the notch sensitivity of the material. The notch sensitivity factor q is defined as
q
K f 1 Kt 1
(2)
where Kf represents the ratio of fatigue limit of smooth specimens to that of notched specimens and Kt is the stress concentration factor. The results of the fatigue tests are depicted in Fig. 8a. For predeformed smooth specimens containing 60 vol-% α’ martensite, the fatigue limit is defined as the stress amplitude at which the sample reaches 2·107 without failure initiated at the surface. All the specimens which failed in the VHCF regime are plotted as runouts in the Fig.8a. The calculated notch sensitivity was plotted against the martensitic content in Fig. 8b.
Fig. 8. (a) Fatigue results for smooth and notched specimens; (b) calculated notch sensitivity in relation to the martensitic content
The results show a strong increase in the notch sensitivity at 30 vol-% α’ martensite, whereas higher martensite contents do not result in further significant increase of the notch sensitivity of the material. This behavior can be explained by the strong hardening effect of the formation of a coherent martensitic network in the microstructure observed in figure 3a. Although these results cannot be interpreted in direct relation with crack initiation at small internal inclusions, they represent a plausible indicator of a negative effect of high martensite contents on the fatigue behavior of the material. Thus the martensite volume fraction obtained during cold working processes should remain beneath 30% in order to keep the notch sensitivity low and avoid fatigue failure at very high number of cycles. 4. Conclusions The fatigue behavior of the metastable austenitic stainless steel AISI304L was investigated and the following conclusions can be drawn: During the predeformation a coherent network of martensitic needles starts to form at 30 vol-% α’ martensite. For volume fractions ≤ 30 % α’ martensite, the material exhibits a true durability without failure in the VHCF regime. This behavior is explained by the local cyclic hardening of the soft austenitic phase
Andrei Grigorescu et al. / Procedia Structural Integrity 2 (2016) 1093–1100 Author name / Structural Integrity Procedia 00 (2016) 000–000
1100 8
resulting in a decrease of strain amplitude. At higher amounts of α’ martensite the damage mechanisms change and failure occurs after very high number of cycles due to internal crack initiation at inclusions accompanied by the formation of a fine granular area. A direct correlation between the FGA-area and the number of cycles to failure could be found for RDsamples (with small areaproj) whereas no such relationship could be confirmed for the TD-samples (with larger areaproj). This confirms the fatigue life controlling role of the crack initiating inclusions if they exceed a certain projected area. This statement is supported by the correlation between the stress intensity factor at inclusions and the fatigue life, which was confirmed only for the TD-samples. In the case of RDsamples the lack of correlation is explained by the dependence of the FGA-growth on the surrounding microstructure for the small areaproj of the crack initiating inclusions. The notch sensitivity increases at 30 vol-% α’-martensite, whereas higher martensite contents do not show any further significant increase. In this respect, a martensite content lower than 30 vol-% is recommended for optimized HCF and VHCF properties.
Acknowledgements The authors gratefully acknowledge the financial support of this study by Deutsche Forschungsgemeinschaft in the framework of the priority programme Life (DFG-SPP1466). References Bowe, K., H., Hammerschmidt, W., Hornbogen, E., Hühner, M., 1988. Bruchmechanische Eigenschaften von metastabilen Austeniten. Materialwissenschaft und Werkstofftechnik 19, 193-201. Grad, P., Reuscher, B., Brodyanski, A., Kopnarski, M., Kerscher, E., 2012. Mechanism of fatigue crack initiation and propagation in the very high cycle regime of high-strength steels. Scripta Materialia 67, 838-841. Li, S., X., 2012. Effects of inclusions on very high cycle fatigue properties of high strength steels. International Materials Reviews 57, 92-114. Lukas, P., Kunz, L., 2002. Specific features of high-cycle and ultra-high-cycle fatigue. Fatigue&Fracture Enginnering Materials & Structures 25, 747-753 Maier, H., Donth, B., Bayerlein, M., Mughrabi, H., Maier, B., Kesten, M., 1993. Optimierte Festigkeitssteigerung eines metastabilen austenitischen Stahles durch verformungsinduzierte Martensitumwandlung bei tiefen Temperaturen. Zeitschrifft für Metallkunde 84, 820-826. Mughrabi, H., 2002. On ‘Multi-stage’ fatigue life diagrams and the relevant life-controlling mechanisms in ultrahigh-cycle fatigue. Fatigue&Fracture Enginnering Materials & Structures 25, 755-764 Murakami, Y., 2002. Metal Fatigue: Effects of Small Defects and Nonmetallic Inclusions, Elsevier, Oxford. Murakami, Y., Nomoto, T., Ueda, T., 1999. Factors influencing the mechanism of superlong fatigue failure in steels, Fatigue & Fracture of Engineering Materials & Structures 22, 581-590. Murakami, Y., Matsunaga, H., 2006. The effect of hydrogen on fatigue properties of steels used for fuel cell systems. International Journal of Fatigue 28, 1509-1520 Müller-Bollenhagen, C., Zimmermann, M., Christ, H.-J., 2010. Adjusting the very high cycle fatigue properties of a metastable austenitic stainless steel by means of the martensite content. Procedia Engineering 2, 1663-1672. Myeon, T., H., Yamabayashi, Y., Shimojo, Y., Higo, Y., 1997. A new life extension method for high cycle fatigue using micro martensitic transformation in austenitic stainless steel. International Journal of Fatigue 19, 69-73. Sakai, T., 2009. Review and prospects for current studies on very high cycle fatigue of metallic materials for machine structural use. Journal of Solid Mechanics and Materials Engineering 3, 425-439.